I posted a new blog posting (http://seanolive.blogspot.com/2010/06/some-new-evidence-that-generation-y.html) that summarizes some recent experimental evidence where I tested a group of teenagers' preferences in loudspeakers and MP3 versus CD music formats. This is just the beginning of a more thorough study, so the results are very preliminary. Still I thought it would be interesting to get some feedback.
I could find no evidence that these high school students preferred the "sizzling sounds of MP3" over higher quality lossless formats, as reported by Jonathan Berger. I also found they preferred the most accurate, neutral loudspeakers when given the opportunity to hear and compare them with something less accurate and neutral.
These results are not too surprising to me, but the media seems to have been reporting a different story over the past year.
Cheers
Sean
Audio Musings (http://seanolive.blogspot.com)
These results are not too surprising to me, but the media seems to have been reporting a different story over the past year.
Who would have thought? The media reporting sensationalist news without properly researching the claims behind it.
Thanks for this research as it's quite valuable. I'm curious, have you ever administered a controlled test where you asked listeners to compare a number (say 3) of speakers that are colored a similar way with one neutral speaker? I'm curious as to if this changes your results since the "normal" speakers may then seem to be colored ones and the neutral speaker seen to be the oddball or deviant speaker.
Also, I was not able to access the slideshow because a google account is needed. Could you remove this restriction?
Finally, an off topic question. I've noticed that in a few pictures (e.g. your avatar) you have the K701 headphone. Is there any particular reason for your selection of this headphone?
These results are not too surprising to me, but the media seems to have been reporting a different story over the past year.
Who would have thought? The media reporting sensationalist news without properly researching the claims behind it.
Thanks for this research as it's quite valuable. I'm curious, have you ever administered a controlled test where you asked listeners to compare a number (say 3) of speakers that are colored a similar way with one neutral speaker? I'm curious as to if this changes your results since the "normal" speakers may then seem to be colored ones and the neutral speaker seen to be the oddball or deviant speaker.
Also, I was not able to access the slideshow because a google account is needed. Could you remove this restriction?
Finally, an off topic question. I've noticed that in a few pictures (e.g. your avatar) you have the K701 headphone. Is there any particular reason for your selection of this headphone?
Thanks for the feedback. I thought I had shared the Google doc slide show but I was mistaken - it's now shared.
I've administered numerous tests where one loudspeaker is relatively neutral and the others are colored in different ways. With trained listeners, it's usually not a problem since they can recognize what is colored and what is not. I have administered some tests where there were clearly contextual effects where the accurate speaker may have sounded too bright or too dull depending on how many other loudspeakers were in the test that were too bright or too dull. Usually these contextual effects have more to do with macro spectral balance effects-- not narrow band colorations. The solution is to have 1 or 2 anchors that are known to be neutral and too bright or too dull.
The AKG K701 headphones are a Harman product, which we use for various purposes: Binaural Room Scanning playback, listener training, etc So I have quite a few pairs of them at work and at home. My Avatar shows my cat wearing a pair (not his choice)
Thanks for the answers and the slideshow. I'm curious as to why you used 128kbps MP3s instead of a higher bitrate? Higher bitrate MP3s are quite common nowadays from sources like Amazon.com. I was a surprised to see your sample preferred CDs to MP3s, but of course the lower bitrate partly explains this. 128kbps seems to be transparent for many people, so I'm still somewhat perplexed by the results of the CD vs MP3 preference.
Maybe that's the difference between between ages 16-18 and 30+
As for the K701, there is an ongoing fight on if that headphone is neutral or colored. I thought perhaps you had measured and decided it was the most neutral product out there, but it being a Harmon product is certainly a great reason to use it!
As for my colored versus neutral speaker question, there's always some debate about liking neutrality versus what you are used to if you have been acclimated to a colored speaker. It's just of ongoing interest to me about how people pick speakers and how they form judgments of what accurate or good sound is.
Nice work Sean. You have a fun job and I'll bet the kids had a fun field trip.
The results are not surprising to me either.
I am interested to know how your test subjects were selected. If this was a voluntary field trip, were they self-selected with an interest in audio?
I'm curious as to why you used 128kbps MP3s instead of a higher bitrate? Higher bitrate MP3s are quite common nowadays from sources like Amazon.com.
This is answered and acknowledged in the blog post.
This is answered and acknowledged in the blog post.
Yes, you are right. So much for the value of skimming.
Interesting, thanks for posting.
I wonder what would result from a similar trial of dynamic vs. heavily compressed and/or clipped masters.
My Avatar shows my cat wearing a pair (not his choice)
But he looks like he's having
such fun!
Interesting, thanks for posting.
I wonder what would result from a similar trial of dynamic vs. heavily compressed and/or clipped masters.
My Avatar shows my cat wearing a pair (not his choice)
But he looks like he's having such fun!
That would be an interesting experiment. If you could show that kids and adults actually prefer their music with more dynamics, etc,then perhaps the record executives would leave well alone the recording when it's mixed and mastered.
I'm sure you'd become a legend here, and in many other places, if you could accomplish that! (Stop getting our hopes up. )
Nice work Sean. You have a fun job and I'll bet the kids had a fun field trip.
The results are not surprising to me either.
I am interested to know how your test subjects were selected. If this was a voluntary field trip, were they self-selected with an interest in audio?
The test subjects were from a local high school in Los Angeles. The school is a Music/Performing Arts Magnate, so many of the students are musically trained - but not all. It made a good sample to compare against Berger's since his subjects in his Mp3 study were also music students attending at Stanford although they were slightly older by as much as 3-4 years.
The high school teacher contacted me wanting a tour of Harman to learn more about the science and engineering of audio equipment. I said fine, as long as I can run your class through some listening experiments, which he jumped at. I always take the opportunity to gather data when people come to visit me. You can't in or out of my lab unless you give me some data. There's no free lunch.
I'm sure you'd become a legend here, and in many other places, if you could accomplish that! (Stop getting our hopes up. )
I think I'm already becoming a legend over at the martin logan owners' forum (http://www.martinloganowners.com/forum/showthread.php?p=123999). The thread is called: "Gen Y prefers more accurate audio - but does not like ML"
They don't seem to like the fact that Generation Y may have better hearing and taste in sound quality than they do...It is interesting that they admit they can identify which model is ML based on how terrible the anechoic measurements are. That's like OJ admitting that was his glove, DNA, footprint and smoking gun left at the murder scene but he didn't kill her? Hmmm...
Heh, and they're already questioning your methods and integrity . . . Oh well! Does their manufacturer of choice ever do controlled, non-commercial-purpose tests like this?
Heh, and they're already questioning your methods and integrity . . . Oh well! Does their manufacturer of choice ever do controlled, non-commercial-purpose tests like this?
Not to my knowledge; if they did wouldn't they be publishing them in the J.AES, product brochures, or at least on their website?
I'd think so, but in any case I don't follow audio tech news or research. Still, I had a hunch there wouldn't be any. This seems to render their criticism rather hollow. Oh, and prepare to be sad:
Looks like his "blog" is just thinly veiled marketing hype for all the Harman branded gear. Totally biased IMHO, and of no educational value. I have removed that "bookmark" from my audio favorites.
Despite the fact you report results anonymously, use double-blind/objective methods, etc. "Science! Huh! What is it good for? Absolutely nothing!"
(And I find it useful to read a page before favouriting it.)
Sean, what a fascinating job you have. I think I might be jealous! The youth are often given a great deal less credit than they are due especially by the press. I guess this happens every generation though.
Have you conducted the same experiment by gender? I would find those results very interesting.
(One off the topic question: Why do so many people use cats in thier avatars????)
I'd think so, but in any case I don't follow audio tech news or research. Still, I had a hunch there wouldn't be any. This seems to render their criticism rather hollow. Oh, and prepare to be sad:
Looks like his "blog" is just thinly veiled marketing hype for all the Harman branded gear. Totally biased IMHO, and of no educational value. I have removed that "bookmark" from my audio favorites.
Despite the fact you report results anonymously, use double-blind/objective methods, etc. "Science! Huh! What is it good for? Absolutely nothing!" (And I find it useful to read a page before favouriting it.)
Yes, I think the Catholic Church said something like that (Science! Huh! What is it good for?)when Galileo told them that their belief system about the solar system didn't jive with reality. People seen to have trouble accepting the truth when they've already put down $3800 on something far removed from it
Sean, what a fascinating job you have. I think I might be jealous! The youth are often given a great deal less credit than they are due especially by the press. I guess this happens every generation though.
Have you conducted the same experiment by gender? I would find those results very interesting.
(One off the topic question: Why do so many people use cats in thier avatars????)
Thanks Cyberdux. .I haven't done the gender analysis yet but there were only 5 females out of the 18 listeners. I plan on running more people through this test so stay tuned
Listener 1 was a female and preferred CD 12/12 times. She said the differences were plainly obvious to her after the first trial. She wrote back to me to thank me and said she was so excited after the field trip she went home and practiced her violin for 1 hour. I wish I could get my daughter that excited so she would practice the piano 1 hour
Excellent post but I have two small nitpicks.
Yes, I think the Catholic Church said something like that (Science! Huh! What is it good for?)when Galileo told them that their belief system about the solar system didn't jive with reality.
While anti-scientism was certainly the end result, this is a vast oversimplification of the relationship between the Church and Galileo.
At the time, the distinction between secular and religious scientific research was slight; of the (few) secular academics and philosophers, many
opposed Galileo; the Pope initially
supported the publication of the
Dialogues, and some have speculated that his later opposition to it was mostly done for political reasons rather than anything else.
Galileo promoted Copernicanism while opposing Tychonic cosmology, notably omitting any reference to it in the
Dialogues, even though it was a numerically equivalent alternative which one could argue was not disproven until several hundred years later (observation of stellar parallax). Of course there are other issues with Tychonic theory related to inertial frames, but there was dreadfully little that was scientifically known at the time which could seriously disprove the scientific theory which was, in fact, being promoted in the Church.
Part of Galileo's dislike of the Tychonic theory was due to a theory of tides he was advocating, which was not merely dead wrong, but internally inconsistent (Wikipedia quotes Skeptical Inquirer as describing this part of the
Dialogues as "cockamamie"!).
Also, anybody who frames a debate with a character named "Simplicio" is an enormous douche who deserves to be cockpunched.
---
Also, what's with using underscores for horizontal rules between sections? If you can bold text, you can use <hr>, right?
The youth are often given a great deal less credit than they are due especially by the press. I guess this happens every generation though.
Damn youth and their 16khz+ hearing. Maybe that's why they can hear the difference between 128kpbs lossy and CDs.
Have you conducted the same experiment by gender? I would find those results very interesting.
Something that I would find very interesting would be the results with regards to the split between gender as far as a love of gear. Many audiophiles seem to be men and just as obsessed with gear as they are music. How would this influence the results of this type of blind test? Could you measure enjoyment of the different genders based on if the music is being reproduced transparently or not (e.g. a colored speaker on low bitrate lossy file)?
"Many audiophiles seem to be men and just as obsessed with gear as they are music. How would this influence the results of this type of blind test? Could you measure enjoyment of the different genders based on if the music is being reproduced transparently or not (e.g. a colored speaker on low bitrate lossy file)?"
Sometimes one wonders whether some of these men actually like music at all after many years of bowing down and worshipping at the throne of brushed aluminium, perspex, and glass.
To do this test at all, the gear might indeed have to be hidden.
PS Please forgive me, I do not know how to "quote" correctly so that the poster quoted is correctly attributed.
To do this test at all, the gear might indeed have to be hidden.
Yes, that's exactly the type of test I was talking about. From what I've seen of women and men, there are different ideas of "value," "High End" and so forth and so on. When if the gear is hidden (as is done in all of Sean Olive's tests) there may be certain psychological factors that play into what different genders consider to be "High Fidelity" sound and what makes for "Good" gear since people will have to guess at what gear is being used.
Have you conducted the same experiment by gender? I would find those results very interesting.
Don't women have a higher upper limit anyway? That one woman's 100% CD preference seems to support that.
Damn youth and their 16khz+ hearing. Maybe that's why they can hear the difference between 128kpbs lossy and CDs.
Huh, my hearing cuts off at ~17+KHz*, at least on my current setup. Is 28 still "youth"?
*) measured with the audiocheck.net downloaded mp3 and subsequently with foobar's tone:// and Audacity's Chirp. Not that I can hear any of it when there's tons of music below 17, naturally.
Please don't say "Generation Y". It makes one look like a media language parrot.
Your project itself is well done, as all the stuff you post here. I hope that it will get the attention it deserves.
I have some doubts (http://forums.audioholics.com/forums/showthread.php?t=66341) about concluding that because a speaker was preferred under very unusual circumstances - listened to by a cluster of people, with a single speaker in the middle of a large room, directly in front of the cluster, playing a mono signal - that an accurate prediction for stereo performance in a real room can be made.
Perhaps if the tests were repeated in stereo, with the speakers positioned as the would be in a real, reverberant, typically sized/furnished room and the same preferences remained, then I suppose such conclusions could be drawn.
cheers,
AJ
p.s. I don't own, nor would I own any ML speakers
Multichannel vs. mono was explored in separate research (http://www.aes.org/e-lib/browse.cfm?elib=14622). Listener sensitivity increased as the number of speakers was reduced.
Multichannel vs. mono was explored in separate research (http://www.aes.org/e-lib/browse.cfm?elib=14622). Listener sensitivity increased as the number of speakers was reduced.
Sean provided me that link on another forum.
The differences in preference ratings increased as the number of playback channels was reduced
That would seem to support my contention. I don't see how the current test set up - mono with the speakers polar field placed in the middle of room directly in front of listener (s!), can be extrapolated to predict preferences in stereo, with 2 speakers placed closer to reverberant room corners and the listener between.
It's certainly an excellent way to focus on certain aspects - on axis linearity, resonances, etc....but the whole picture??
cheers,
AJ
You probably should let Sean straighten you out on this. But as I understand it, they asked listeners to compare speakers. They tried various configurations for the comparisons. The configuration in which listeners were most able to distinguish one speaker from another was mono. In a stereo configuration, different speakers sounded more similar presumably because of what you've said, you're listening more to the room in that scenario and the room doesn't change when you swap out the speakers.
I think it may be fair to say that mono listening may unduly exaggerate differences between speakers but it does appear to be valid to apply findings in mono testing to stereo.
The configuration in which listeners were most able to distinguish one speaker from another was mono.
That's interesting. I read it the opposite way but I can now see yours as well.
The differences in preference ratings increased as the number of playback channels was reduced
To me that means that the listeners agreed less on their preference ratings (= more variety) when less speakers were used. IOW, listeners quite agreed about best and worst speakers in a multi-channel setup, but agreed less (=more confusion) in a mono setup.
Hopefully Sean can tell us if the "
differences in preference ratings" are those between subjects or between speakers.
[I posted too soon. Reading the paper's abstract clears it up ]
Also, anybody who frames a debate with a character named "Simplicio" is an enormous douche who deserves to be cockpunched.
Hah, I always thought I was the only one who found that an act of monumental arrogance (to the point of being admirable!). You have to give it to him: he really was confident!
I have some doubts (http://forums.audioholics.com/forums/showthread.php?t=66341) about concluding that because a speaker was preferred under very unusual circumstances - listened to by a cluster of people, with a single speaker in the middle of a large room, directly in front of the cluster, playing a mono signal - that an accurate prediction for stereo performance in a real room can be made.
Perhaps if the tests were repeated in stereo, with the speakers positioned as the would be in a real, reverberant, typically sized/furnished room and the same preferences remained, then I suppose such conclusions could be drawn.
cheers,
AJ
p.s. I don't own, nor would I own any ML speakers
Let me address your points:
1) Listening to a single speaker the middle of the room pretty well approximates a center channel setup. If you analyze most surround recordings/movies/TV broadcasts there is a substantial mono component coming out of the front channels, with maybe 80-90% coming from the center channel speaker. We can argue what the percentages are but hopefully you get my point.
2) A "cluster of people" around the reference axis of a loudspeaker is not too uncommon in my home when our family watches movies/TV/ plays Wii games. While we are a close family, we don't sit on each others' laps on the sweet spot. When listening to music, we are seldom on the reference axis of the loudspeakers but rather walking around the room doing other things. Are you arguing that electrostatic loudspeaker owners are a lonely, anti-social, solitary demographic who have no friends or family, and only listen to stereo reproductions on the tiny sweet spot they produce?
3) While the students were tested in small cluster around the sweet spot of the electrostatic speaker (within +- 30 degrees of the reference axis), the trained listeners all sat in the same seat directly on the reference axis. The trained listeners actually rated the electrostatic loudspeaker significantly lower when sitting in the optimal reference axis.
4) Our listening room in terms of reverberation is not atypical of a domestic listening rooms. The walls are all reflective with some diffusion, there is furniture, and the floor is carpeted. Most of the speakers that perform well in this room generally get good reviews and do well in the real world based on feedback I hear.
5) If you read our recent paper on mono/stereo/surround equalization comparisons and Floyd Toole's recent book, there is good correlation between loudspeaker preference ratings in mono and stereo. When you move from mono to stereo to surround, there is more noise in the ratings, as listeners seem to be less able to formulate reliable and discriminating preference ratings. The best explanation I have so far, is that the off-axis reflected sound of the loudspeaker may play a perceptually less important role as you add more speakers because the direct (and mostly on-axis) sound coming from the multiple sources tends to dominate (mask) the reflected sounds. Of course, you can't ignore the off-axis performance of the loudspeaker because there are many programs where you go from multiple channels being active to just a single channel (vocal or dialog).
My last point, should be viewed as good news for manufacturers who make crappy loudspeakers with poor off-axis performance: the solution for the customer is to simply buy more of them. If you fill the room with enough of them, you may not be might not be able to tell how truly bad they are -- at least until the solo instrument or dialog appears in the center channel -
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
While I understand the rationale for using CBR 128 encoding - more-or-less equivalence with Berger's undocumented test - the almost complete lack of any mention (let alone discussion) of the importance of bitrate ultimately leaves us with yet another "MP3 is worse than CD" Internet article (that "Now versus Then" slide touches on the subject, but then you never revisit it).
MP3 @ CBR 128 is not exactly robust - simply changing to VBR encoding at that same target bitrate improves things dramatically, to say nothing of raising the target bitrate to present-day "industry standards", i.e. ~256 kb.
You do say in your conclusion that this study is in an "early phase", and you note in a couple places that Berger's findings are perhaps outdated - as well as "name-checking" the iTunes AAC encoder @ 256 kb - but with statements such as "the high school students preferred the most accurate option, preferring CD over MP3", you seem to simply be providing yet more fuel to the "other fire", i.e. lossy encoding inherently "sounds bad".
I can only hope that you'll be doing a study in the not-too-distant future where you repeat this test - or one very similar to it - utilizing more "modern" bitrates and encoders.
Perhaps releasing the results of your current study without having the results of a higher-bitrate test to compare them to was a bit premature?
The configuration in which listeners were most able to distinguish one speaker from another was mono.
That's interesting. I read it the opposite way but I can now see yours as well.
The differences in preference ratings increased as the number of playback channels was reduced
To me that means that the listeners agreed less on their preference ratings (= more variety) when less speakers were used. IOW, listeners quite agreed about best and worst speakers in a multi-channel setup, but agreed less (=more confusion) in a mono setup.
Hopefully Sean can tell us if the "differences in preference ratings" are those between subjects or between speakers.
[I posted too soon. Reading the paper's abstract clears it up ]
No, it was the opposite effect of what your describe: in that study we found that in stereo and surround (5.1) playback conditions, the differences in listener preferences among the loudspeaker equalizations were smaller than in mono playback condition.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
While I understand the rationale for using CBR 128 encoding - more-or-less equivalence with Berger's undocumented test - the almost complete lack of any mention (let alone discussion) of the importance of bitrate ultimately leaves us with yet another "MP3 is worse than CD" Internet article (that "Now versus Then" slide touches on the subject, but then you never revisit it).
MP3 @ CBR 128 is not exactly robust - simply changing to VBR encoding at that same target bitrate improves things dramatically, to say nothing of raising the target bitrate to present-day "industry standards", i.e. ~256 kb.
You do say in your conclusion that this study is in an "early phase", and you note in a couple places that Berger's findings are perhaps outdated - as well as "name-checking" the iTunes AAC encoder @ 256 kb - but with statements such as "the high school students preferred the most accurate option, preferring CD over MP3", you seem to simply be providing yet more fuel to the "other fire", i.e. lossy encoding inherently "sounds bad".
I can only hope that you'll be doing a study in the not-too-distant future where you repeat this test - or one very similar to it - utilizing more "modern" bitrates and encoders.
Perhaps releasing the results of your current study without having the results of a higher-bitrate test to compare them to was a bit premature?
I would be the first person to admit that increasing the bit-rate, moving to VBR or using a more transparent CODEC will produce different results, and I will certainly correct any misconception that a good lossy codec is not capable of producing a near transparent reproduction: there is already substantial scientific data from prior listening tests to prove that point.
The main point here was to illustrate that when music is being coded or streamed at less-than-optimal quality rate, younger people may prefer the higher quality option, contrary to what a previous informal study found.
The results should be considered preliminary. More variables (test subjects, music samples, playback conditions,etc ) will be manipulated and tested in the upcoming months.
Regarding your "premature release" comment, I wanted to talk about the study now so I can useful get feedback - like yours - to help improve the design of future experiments. I don't think talking about this study is any less "premature" than many of the presentations and preprints given at Audio Engineering Society and other scientific conferences where the research discussed is often in the very early developmental stages.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Hi Sean,
Thanks for your reply.
1) Listening to a single speaker the middle of the room pretty well approximates a center channel setup. If you analyze....
But these are
not center channel speakers. As a matter of fact, it is clear that all are intended to be LR "main" speakers. The only conceivable application as a center would be with some form of perforated screen, which has to be a (very) small percentage of the intended market. I would have no objection if your test were for center channels, it would be a reasonable one.
2) A "cluster of people" around the reference axis of a loudspeaker is not too uncommon in my home when our family watches movies/TV/ plays Wii games. While we are a close family, we don't sit on each others' laps on the sweet spot. When listening to music, we are seldom on the reference axis of the loudspeakers but rather walking around the room doing other things.
But once again, the speaker under test is not being placed or listened to on it's intended "reference axis" relative to the cluster. It is not coupling to modes as it would be normally placed relative to the cluster. It's polar response is not reflecting of the room surfaces as it would be normally placed relative to the cluster. Yet you are judging it's perceived performance as if it were.
Are you arguing that electrostatic loudspeaker owners are a lonely, anti-social, solitary demographic who have no friends or family, and only listen to stereo reproductions on the tiny sweet spot they produce?
That could well be the situation in the case of "Audiophiles", but that is not what I'm arguing. I'm saying judge the ML main LR's vs the 362 main LR's as they were intended...placed and listened to as LR's...in stereo (we'll save MCH for another day).
4) Our listening room in terms of reverberation is not atypical of a domestic listening rooms.
Yet this is how the ML, with its dipolar radiation is being judged. Atypically, in the center of a less reverberant room.
5) If you read our recent paper on mono/stereo/surround equalization comparisons and Floyd Toole's recent book, there is good correlation between loudspeaker preference ratings in mono and stereo. When you move from mono to stereo to surround, there is more noise in the ratings, as listeners seem to be less able to formulate reliable and discriminating preference ratings. The best explanation I have so far, is that the off-axis reflected sound of the loudspeaker may play a perceptually less important role as you add more speakers because the direct (and mostly on-axis) sound coming from the multiple sources tends to dominate (mask) the reflected sounds.
Well, the sound field certainly becomes more complex as the number of sources increase. When you say good correlation, do you mean to speakers (placed) and tested in stereo?
My last point, should be viewed as good news for manufacturers who make crappy loudspeakers with poor off-axis performance: the solution for the customer is to simply buy more of them. If you fill the room with enough of them, you may not be might not be able to tell how truly bad they are -- at least until the solo instrument or dialog appears in the center channel -
Let me finish with this. Do you consider the performance of the Primus 360 to be similar enough to the 362, where the results would not have changed much?
The 360's measured very well by Stereophile and as the 362's predecessor, seems very similar, so I'm hoping your answer is yes.
cheers,
AJ
Hi Sean,
Thanks for your reply.
1) Listening to a single speaker the middle of the room pretty well approximates a center channel setup. If you analyze....
But these are not center channel speakers. As a matter of fact, it is clear that all are intended to be LR "main" speakers. The only conceivable application as a center would be with some form of perforated screen, which has to be a (very) small percentage of the intended market. I would have no objection if your test were for center channels, it would be a reasonable one.
My center channel speaker is identical to my L/R speaker. And to my surround speakers.
I would have no objection if your test were for center channels, it would be a reasonable one.
Well we agree on that much. While our speaker mover can accommodate testing of four sets of left, center or right channels, on what basis are you arguing that testing them in center locations somehow invalidates the results of the test, or prevents us from extrapolating to how they might perform as left/right, or surround channels for that matter. In all cases, the listener is hearing the exact same direct sound (albeit at different angles of incidence if they don't look at the speaker). So are you arguing that by moving the speaker into the left position the change in off-axis reflected sounds will significantly change the rank ordering of the loudspeakers?
But once again, the speaker under test is not being placed or listened to on it's intended "reference axis" relative to the cluster. It is not coupling to modes as it would be normally placed relative to the cluster. It's polar response is not reflecting of the room surfaces as it would be normally placed relative to the cluster. Yet you are judging it's perceived performance as if it were.
So you are arguing that the electrostatic loudspeaker can only accommodate 1 listener or 1 listening seat in the room? To me that is a gross admission of failure in the design of the loudspeaker. Nonetheless, the results from the trained listeners who were sitting on it's intended reference axis indicate the speaker is significantly flawed. Moreover,if you compare the results of the high school students sitting on versus off the reference axis there is no significant difference. No matter where you sit the speaker sounds imbalanced and colored. The anechoic measurements explain why this is so.
That could well be the situation in the case of "Audiophiles", but that is not what I'm arguing. I'm saying judge the ML main LR's vs the 362 main LR's as they were intended...placed and listened to as LR's...in stereo (we'll save MCH for another day)
.
I've done enough mono versus stereo loudspeaker comparisons (as has Floyd Toole) over the past 20 years to convince myself that the results track each other. If a speaker has resonances,sounds colored and spectrally imbalanced in mono those characteristics will not magically disappear in stereo. Do you have an explanation why this would not be true?
Yet this is how the ML, with its dipolar radiation is being judged. Atypically, in the center of a less reverberant room
I didn't say our room was atypically less reverberant. I said it (acoustically) was not atypical [that means typical) of an average domestic listening room based on a study that John Bradley did of some 600 Canadian homes.
The 360's measured very well by Stereophile and as the 362's predecessor, seems very similar, so I'm hoping your answer is yes.
Yes, the only differences should be cosmetic ones.
While our speaker mover can accommodate testing of four sets of left, center or right channels, on what basis are you arguing that testing them in center locations somehow invalidates the results of the test
I am not contending that the
results are invalid. I'm saying the conclusions from those results and the predictions about the speakers perceived performance, when placed closer to corners...and listened to in stereo...is debatable. Had you done such a (stereo) test and found the same/similar preferences as mono, I would not be typing right now.
or prevents us from extrapolating to how they might perform as left/right
Comparison of Loudspeaker-Room Equalization Preferences for Multichannel, Stereo, and Mono Reproductions....
The differences in preference ratings increased as the number of playback channels was reduced
IOW, the increase in the number of channels possibly
decreases the perceived problems in mono. And who buys an ML Vista, or a 362, to place them in the center of a room and listen in mono?
So are you arguing that by moving the speaker into the left position the change in off-axis reflected sounds will significantly change the rank ordering of the loudspeakers?
Absolutely....especially when integrated into the fact that the complete
stereophonic soundfield, with each of the two sources placed towards the corners, will be perceived completely differently by a centered, or center cluster of listeners. Your own paper seems to support this.
I've done enough mono versus stereo loudspeaker comparisons (as has Floyd Toole) over the past 20 years to convince myself that the results track each other. If a speaker has resonances,sounds colored and spectrally imbalanced in mono those characteristics will not magically disappear in stereo. Do you have an explanation why this would not be true?
I have never said they would "magically" disappear. I question whether our perception of them is the same, under very different conditions - generating a stereophonic soundfield, placed near corners of a reverberant room. The exact same issues may be there, but will we perceive them the same way.
Did you measure a single ML (corner positioned) at the listening area and compare it to the anechoic measurement? Presented as an overlay would be beneficial. Thanks.
I didn't say our room was atypically less reverberant. I said it (acoustically) was not atypical [that means typical) of an average domestic listening room based on a study that John Bradley did of some 600 Canadian homes.
I should read slower and more carefully late at work . My apologies.
Yes, the only differences (360 vs 362) should be cosmetic ones.
No matter where you sit the (ML) speaker sounds imbalanced and colored. The anechoic measurements explain why this is so.
Thank you. I'll try my best to test the veracity of that prediction.
cheers,
AJ
The 360's measured very well by Stereophile and as the 362's predecessor, seems very similar, so I'm hoping your answer is yes.
Yes, the only differences should be cosmetic ones.
The 360s have one woofer per box, while the 362s have two.
I have a pair of 360s and they are very smooth.
The 360s have one woofer per box, while the 362s have two.
This is the 360 (http://www.stereophile.com/floorloudspeakers/605infinity/) we're referring to Arnold, the predecessor to the 362 (http://www.infinitysystems.com/home/products/product_detail.aspx?prod=P362BK&ser=PRI&Language=ENG&Region=USA&Country=US)
cheers,
AJ
Heh, and they're already questioning your methods and integrity . . . Oh well! Does their manufacturer of choice ever do controlled, non-commercial-purpose tests like this?
Not to my knowledge; if they did wouldn't they be publishing them in the J.AES, product brochures, or at least on their website?
Not neccessarily. Some research is just commercial proprietary research never shared with anyone outside the vault.
So you are arguing that the electrostatic loudspeaker can only accommodate 1 listener or 1 listening seat in the room? To me that is a gross admission of failure in the design of the loudspeaker.
As an owner of electrostatic speakers I would certainly say yes, they accomodate 1 listener and/or one listening seat in the room. How is that a gross admission of failure of the design oof the loudspeaker? That is like saying a formula one racing car is an example of a failed design of an automobile because it only seats one person. One can't judge success or failure of a design without considering the purpose for which it is designed.
Nonetheless, the results from the trained listeners who were sitting on it's intended reference axis indicate the speaker is significantly flawed. Moreover,if you compare the results of the high school students sitting on versus off the reference axis there is no significant difference. No matter where you sit the speaker sounds imbalanced and colored. The anechoic measurements explain why this is so.
How about a comparison using the speakers as they were intended to be used? If I put black pepper in my mouth directly it tastes pretty awful. I think you would get a pretty solid concensus of that result were you to test a wide range of people. But what does that tell us about the taste of pepper on food? You won't get as much of a concensus on that test but at least it would be a test of the product as it is intended to be used. It seems to me that the fact that such comparisons leads to less agreement doesn't follow that it is an inferior test. How do you know your quest for greater consolidation of results simply isn't leading you to test under a woefully limited context, kind of like my pepper analogy? IMO it just isn't a fair test unless you use the competition's speakers as they were designed to be used in a room they were designed to be used in with a wide range of source material as they would used in the field.
As an owner of electrostatic speakers I would certainly say yes, they accomodate 1 listener and/or one listening seat in the room. How is that a gross admission of failure of the design oof the loudspeaker? That is like saying a formula one racing car is an example of a failed design of an automobile because it only seats one person. One can't judge success or failure of a design without considering the purpose for which it is designed.
When you purchased it, did the manufacturer of the speaker or dealer explicitly state this was how the speaker was designed to be used? Did they talk to you about the sweet spot, ideal speaker positioning, ideal seating position, and state all of this was very particular and important to this particular speaker? Did they tell you this behavior or design diverged quite a bit from the typical dynamic speaker, which most people are used to?
I've visited many audio dealers and NONE have ever made such statement when trying to sell me an electrostatic speaker. I don't recall seeing any of this in the marketing literature either.
If they are assuming that the buyer will know or setup their room to use the speaker that way, that is a faulty assumption. Many people do not have the luxury of changing the whole room for a pair of speakers. Beyond that, even people will dedicated listening rooms may have multiple couches, chairs, seating surfaces, etc. In my home I sometimes listen to music and read while lying down or sitting in a recliner that is not in the sweet spot. I would certainly be irritated to learn that the manufacturer of an allegedly superior speaker designed them with the assumption that I would not perform the common activity of listening outside the sweet spot.
Other than consumers like Richard Branson or the CEO of Oracle, few people purchase formula one cars. Many "average joes" with no knowledge outside of what their dealers and marketing literature has told them will buy electrostatics.
As for the formula car, it's pretty obvious from looking at the car that it only seats one person. And everybody knows the cars are designed for a specific purpose and not one professional associated with Formula one would say otherwise. The Formula one car is not advertised as a car for daily use. Is it obvious from looking at an electrostatic that it's off axis performance is different from a dynamic speaker? I think it's safe to assume speakers sold to consumers will be used as "daily use" items, meaning they have to accommodate to the different ways consumers use them and the differences between consumers.
I've visited many audio dealers and NONE have ever made such statement when trying to sell me an electrostatic speaker.
Did these many audio dealers all have you listen in mono directly in front of a room centered electrostatic speaker, 3' off the front wall?
What was your impression of the sound?
That is like saying a formula one racing car is an example of a failed design of an automobile because it only seats one person. One can't judge success or failure of a design without considering the purpose for which it is designed.
But I would bet it says in the Formula One's operator's manual that only person can sit in it, and it's probably clear when you see it. The drivers manual probably warns you not to drive the car except on a professional racing track under optimal driving conditions; it's probably clearly understood by the owner that it's not a practical car you can drive in rain, snow or a wide range of real world driving conditions: otherwise the car will operate at 30% of its performance, and be unsafe and dangerous for the driver and people around the car.
There is no such warning in this electrostatic speaker's manual that says: "WARNING: This $3800 loudspeaker is only intended for 1 listener in one specific spot in one specific room of specific size and acoustic treatment. IF your listening room, speaker position and single listening spot DOES NOT meet these exact conditions DO NOT PURCHASE IT: OTHERWISE its sound quality may fall into the lower 30 percentile of modern day loudspeakers ( and will sound significantly worst than a $500 loudspeaker that has no such specific requirements)
How about a comparison using the speakers as they were intended to be used?
I will send you the plans of our room, if you can provide me the exact locations where the speaker and listener should be located so that the test meets the "purpose for which the speaker was designed". Do that and I would be happy to rerun the test. Also, let me know the specific music tracks for which it was designed.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Did these many audio dealers all have you listen in mono directly in front of a room centered electrostatic speaker, 3' off the front wall?
What was your impression of the sound?
Believe me, some of the setups I've used at audio dealers are far worse than any claimed issues with Dr. Olive's test. Perhaps that's why I've never been impressed with electrostatics when I listen at dealers. Beyond that, my comments were not directed at the validity of Dr. Olive's testing methodology. I was refuting analog scotts' claims that a speaker with terrible off axis performance is a sensible design and that the formula one car analogy is a valid one.
As an owner of electrostatic speakers I would certainly say yes, they accomodate 1 listener and/or one listening seat in the room. How is that a gross admission of failure of the design oof the loudspeaker? That is like saying a formula one racing car is an example of a failed design of an automobile because it only seats one person. One can't judge success or failure of a design without considering the purpose for which it is designed.
When you purchased it, did the manufacturer of the speaker or dealer explicitly state this was how the speaker was designed to be used? Did they talk to you about the sweet spot, ideal speaker positioning, ideal seating position, and state all of this was very particular and important to this particular speaker? Did they tell you this behavior or design diverged quite a bit from the typical dynamic speaker, which most people are used to?
I've visited many audio dealers and NONE have ever made such statement when trying to sell me an electrostatic speaker. I don't recall seeing any of this in the marketing literature either.
If they are assuming that the buyer will know or setup their room to use the speaker that way, that is a faulty assumption. Many people do not have the luxury of changing the whole room for a pair of speakers. Beyond that, even people will dedicated listening rooms may have multiple couches, chairs, seating surfaces, etc. In my home I sometimes listen to music and read while lying down or sitting in a recliner that is not in the sweet spot. I would certainly be irritated to learn that the manufacturer of an allegedly superior speaker designed them with the assumption that I would not perform the common activity of listening outside the sweet spot.
Other than consumers like Richard Branson or the CEO of Oracle, few people purchase formula one cars. Many "average joes" with no knowledge outside of what their dealers and marketing literature has told them will buy electrostatics.
As for the formula car, it's pretty obvious from looking at the car that it only seats one person. And everybody knows the cars are designed for a specific purpose and not one professional associated with Formula one would say otherwise. The Formula one car is not advertised as a car for daily use. Is it obvious from looking at an electrostatic that it's off axis performance is different from a dynamic speaker? I think it's safe to assume speakers sold to consumers will be used as "daily use" items, meaning they have to accommodate to the different ways consumers use them and the differences between consumers.
Odigg,
I was writing my response as you were writing this response. We clearly think along the same lines. I like you
Cheers
Sean
Audio Musings (http://seanolive.blogspot.com)
I will send you the plans of our room, if you can provide me the exact locations where the speaker and listener should be located so that the test meets the "purpose for which the speaker was designed". Do that and I would be happy to rerun the test. Also, let me know the specific music tracks for which it was designed.
Here ya go Sean
Try some classical or big band jazz...something with real acoustic instruments to compare against your stored memory. See if you get more spatial realism when switching from the boxes to the stats (hopefully the "trained" listeners will recognize this instantly ).
Btw, hopefully the room isn't the smaller version diagrammed above, because the 362 monopoles placed 2' off the sidewalls, even toed in, could get a bit nasty with the early reflections (no side nulls like the dipoles)...and require some of those "acoustic treatment" bandaids that kill spaciousness and realism.
Let us know the results .
cheers,
AJ
Believe me, some of the setups I've used at audio dealers are far worse than any claimed issues with Dr. Olive's test.
Can you explain that to me, I'm having trouble making sense of it - The dealer setups were "worse than any claimed issues with Dr. Olive's test"? TIA.
Were any of them in mono with the speaker centered on the front wall?
I was refuting analog scotts' claims that a speaker with terrible off axis performance is a sensible design
Where did he claim that?
What speaker has "terrible off axis performance"?
cheers,
AJ
Did these many audio dealers all have you listen in mono directly in front of a room centered electrostatic speaker, 3' off the front wall?
What was your impression of the sound?
Believe me, some of the setups I've used at audio dealers are far worse than any claimed issues with Dr. Olive's test. Perhaps that's why I've never been impressed with electrostatics when I listen at dealers. Beyond that, my comments were not directed at the validity of Dr. Olive's testing methodology. I was refuting analog scotts' claims that a speaker with terrible off axis performance is a sensible design and that the formula one car analogy is a valid one.
According to the manual of this speaker it's supposed to be 3 ft from the front wall. On page 10 it states:
"By now your speakers should be placed approximately two to three feet from the front wall, the wall in front of the listening position, and about two feet from the side walls"
Is that a misprint? Or does this confirm I got the distance from the front wall correct in our testl
A few paragraphs later it contradicts the recommendation for placement 2 ft from the side walls:
"A good rule of thumb is to have the side walls as far away from the speaker sides as possible... An ideal side wall, however, is no side wall at all. "
This indicates that I also got the distance from the side walls perfect according to manufacturer's instructions! 2 out of 2
So according to the electrostatic owners' manual I set the speakers up perfectly as instructed in terms of distances to the front and side walls.
They recommend a 72 hour break-in period, which to me is always a Red Flag for electrodynamic loudspeakers at least. It's more of a cognitive effect they hope for where listener adaptation to the speaker over several days will alleviate buyer remorse due to issues about its poor sound quality.
" Now that you have positioned your speaker system, spend time listening. Wait to make any major changes in your initial setup for the next few days as the speaker system itself will change subtly in its sound. Over the first 72 hours of play the actual tonal quality will change slightly with deeper bass and more spacious highs resulting. After a few days of listening you can begin to make refinements and hear the differences."
I'm not saying that's the case here but it would be interesting to see some physical and perceptual evidence that this break-in has measurable effects on its sound. I've actually done this with our own loudspeakers (motivated by a former marketing person who insisted on putting "break-in labels on a model of our speakers). Both measurements and listening tests proved the break-in was a crock. The break-in stickers were removed, and the marketing person was fired shortly thereafter.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
As an owner of electrostatic speakers I would certainly say yes, they accomodate 1 listener and/or one listening seat in the room. How is that a gross admission of failure of the design oof the loudspeaker? That is like saying a formula one racing car is an example of a failed design of an automobile because it only seats one person. One can't judge success or failure of a design without considering the purpose for which it is designed.
When you purchased it, did the manufacturer of the speaker or dealer explicitly state this was how the speaker was designed to be used? Did they talk to you about the sweet spot, ideal speaker positioning, ideal seating position, and state all of this was very particular and important to this particular speaker? Did they tell you this behavior or design diverged quite a bit from the typical dynamic speaker, which most people are used to?
I bought my current speakers, the Soundlab A3s second hand. But as a long time owner of ML CLSs I already knew what I was getting. Way back when I bought the CLSs the from Optimal Enchantment Randy Cooley went out of his way as a dealer to accomidate me as a consumer. He even proctored a SB comparison between the MLs and a pair of Apogee Duettas. Yes he made it very clear that these were one person speakers. The in store audition and the week long home audition made it clear as well. Randy more than did his job. I was well informed.
I've visited many audio dealers and NONE have ever made such statement when trying to sell me an electrostatic speaker. I don't recall seeing any of this in the marketing literature either.
If they are assuming that the buyer will know or setup their room to use the speaker that way, that is a faulty assumption.
I can not speak for all ML dealers. But Randy at Optimal Enchantment didn't make any such assumptions. They gave me excellent instructions on home setup and use and came by to help tweak my speaker and listener positioning.
Many people do not have the luxury of changing the whole room for a pair of speakers.
Of course not. And this is a significant point. Some of us do. Some of us have the luxury of having a custom dedicated listening room. For those of us with that luxury speakers that demand such things as a dedicated listening room with treatment for optimization are not a "failed design" because of those needs any more than a formula one car is a "failed design" because it has practical limitations.
Beyond that, even people will dedicated listening rooms may have multiple couches, chairs, seating surfaces, etc.
Or they may not. I didn't.
In my home I sometimes listen to music and read while lying down or sitting in a recliner that is not in the sweet spot. I would certainly be irritated to learn that the manufacturer of an allegedly superior speaker designed them with the assumption that I would not perform the common activity of listening outside the sweet spot.
I would think that if you did a propper extended home audition this would have come to your attention. It certainly was more than clear to me after my audition.
Other than consumers like Richard Branson or the CEO of Oracle, few people purchase formula one cars. Many "average joes" with no knowledge outside of what their dealers and marketing literature has told them will buy electrostatics.
Highend electrostatic speakers are IMO no more for the average Joe than a formula one car. I know ML has made a wide range of product long after the CLSs and I haven't paid much attention to those products. I know they even have a surround sound system. I don't know if that system has the same limitations of the CLSs or not. But if I were in the market for those speakers I would find out before buying. The top of the line Soundlabs are 9 feet tall! Ya think someone would buy these without knowing a thing or two about how they work in a room?
As for the formula car, it's pretty obvious from looking at the car that it only seats one person. And everybody knows the cars are designed for a specific purpose and not one professional associated with Formula one would say otherwise. The Formula one car is not advertised as a car for daily use. Is it obvious from looking at an electrostatic that it's off axis performance is different from a dynamic speaker? I think it's safe to assume speakers sold to consumers will be used as "daily use" items, meaning they have to accommodate to the different ways consumers use them and the differences between consumers.
IMO the same is quite obvious for high end electrostats an is made clear with just the simplest and most brief audition. Cerianly the two channel high end systems I have had and have listened to. If someone bought electrostats without ever auditioning them they may be in for a surprise. But I would think some folks would actually do their homework first. I don't think the potential pitfalls of buying speakers without a propper audition and some homework would be unique to electrostatic speakers. Evey design has it's unique characteristics and optimal room configuration.
That is like saying a formula one racing car is an example of a failed design of an automobile because it only seats one person. One can't judge success or failure of a design without considering the purpose for which it is designed.
But I would bet it says in the Formula One's operator's manual that only person can sit in it, and it's probably clear when you see it. The drivers manual probably warns you not to drive the car except on a professional racing track under optimal driving conditions; it's probably clearly understood by the owner that it's not a practical car you can drive in rain, snow or a wide range of real world driving conditions: otherwise the car will operate at 30% of its performance, and be unsafe and dangerous for the driver and people around the car.
There is no such warning in this electrostatic speaker's manual that says: "WARNING: This $3800 loudspeaker is only intended for 1 listener in one specific spot in one specific room of specific size and acoustic treatment. IF your listening room, speaker position and single listening spot DOES NOT meet these exact conditions DO NOT PURCHASE IT: OTHERWISE its sound quality may fall into the lower 30 percentile of modern day loudspeakers ( and will sound significantly worst than a $500 loudspeaker that has no such specific requirements)
How about a comparison using the speakers as they were intended to be used?
I will send you the plans of our room, if you can provide me the exact locations where the speaker and listener should be located so that the test meets the "purpose for which the speaker was designed". Do that and I would be happy to rerun the test. Also, let me know the specific music tracks for which it was designed.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
I have a hunch Formula One cars don't come with any manual.
As I have stated in other posts the reality of the one listener limitation of the electrostaic speakers I encountered was plainly obvious with even the most limited auditions. Um, the Revel Ultimas that I had the pleasure of listening to had a sweat spot too. The damage was obviously less severe as one moved from the sweat spot with the Revels but for me it was still enough to call those speakers a one man speaker as well along with every other two channel speaker system I have ever heard. It's just that with electrostats the damage is more severe. I can't say that I have ever heard any two channel system that did not suffer from moving out of the sweat spot. I'm pretty sure the physics dictates that the imaging will always be affected by moving out of the sweat spot. Call me crazy but I didn't go out of my way to get the best sounding system I could afford with the idea of screwing up that sound by sitting somewhere other than the sweat spot. If I owned Revel speakers I would still have a dedicated listening room and listen from the sweat spot.
I'd be more than happy to look at the plans of your facility and make suggestions on how to better situate electrostatic speakers in that room and what sort of treatment to use to further optimise things. But it would just be a starting point. I would also be happy to provide you with a play list of material that works for me as an audiophile when doing such auditions. I am glad to hear you would consider these variables in follow up tests.
Did these many audio dealers all have you listen in mono directly in front of a room centered electrostatic speaker, 3' off the front wall?
What was your impression of the sound?
Believe me, some of the setups I've used at audio dealers are far worse than any claimed issues with Dr. Olive's test. Perhaps that's why I've never been impressed with electrostatics when I listen at dealers. Beyond that, my comments were not directed at the validity of Dr. Olive's testing methodology. I was refuting analog scotts' claims that a speaker with terrible off axis performance is a sensible design and that the formula one car analogy is a valid one.
According to the manual of this speaker it's supposed to be 3 ft from the front wall. On page 10 it states:
"By now your speakers should be placed approximately two to three feet from the front wall, the wall in front of the listening position, and about two feet from the side walls"
Is that a misprint? Or does this confirm I got the distance from the front wall correct in our testl
A few paragraphs later it contradicts the recommendation for placement 2 ft from the side walls:
"A good rule of thumb is to have the side walls as far away from the speaker sides as possible... An ideal side wall, however, is no side wall at all. "
This indicates that I also got the distance from the side walls perfect according to manufacturer's instructions! 2 out of 2
So according to the electrostatic owners' manual I set the speakers up perfectly as instructed in terms of distances to the front and side walls.
They recommend a 72 hour break-in period, which to me is always a Red Flag for electrodynamic loudspeakers at least. It's more of a cognitive effect they hope for where listener adaptation to the speaker over several days will alleviate buyer remorse due to issues about its poor sound quality.
" Now that you have positioned your speaker system, spend time listening. Wait to make any major changes in your initial setup for the next few days as the speaker system itself will change subtly in its sound. Over the first 72 hours of play the actual tonal quality will change slightly with deeper bass and more spacious highs resulting. After a few days of listening you can begin to make refinements and hear the differences."
I'm not saying that's the case here but it would be interesting to see some physical and perceptual evidence that this break-in has measurable effects on its sound. I've actually done this with our own loudspeakers (motivated by a former marketing person who insisted on putting "break-in labels on a model of our speakers). Both measurements and listening tests proved the break-in was a crock. The break-in stickers were removed, and the marketing person was fired shortly thereafter.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
The break in on Martin Logans is real. I have done side by side comparisons between brand new and broken in MLs. It is a substantial difference. You are dealing with a very different driver here. Dealer demo speakers do tend to have more than 72 hours on them so it should not be an issue with any auditions using dealer demo speakers.
Believe me, some of the setups I've used at audio dealers are far worse than any claimed issues with Dr. Olive's test.
Can you explain that to me, I'm having trouble making sense of it - The dealer setups were "worse than any claimed issues with Dr. Olive's test"? TIA.
Were any of them in mono with the speaker centered on the front wall?
I was refuting analog scotts' claims that a speaker with terrible off axis performance is a sensible design
Where did he claim that?
What speaker has "terrible off axis performance"?
cheers,
AJ
What is your problem with mono? Do you not believe the research results I've point you towards indicating that a) mono test results track stereo results b) listeners are more discriminating in mono c) that mono signals exist in stereo music recordings and multichannel music/films? Your rationale for requiring stereo signals for loudspeaker tests escapes me given the body of research that suggests otherwise.
Check the off-axis measurements for speaker C in slide 28 (http://docs.google.com/fileview?id=0B97zTRsdcJTfZmM1ZjhlN2MtMTQwZC00MzVmLWE1OTQtMWQzN2IzNGNiMjc0&hl=en) compared to the Infinity. The mid/treble is significantly down in level relative to the bass: they will sound very dull as you move off-axis. The reflected sounds arriving at the listener will also sound dull and colored compared to the direct sound.
There is a strong argument that this speaker will sound better the further it's away from side walls (as we tested it and as the owner's manual recommends) since there will be less colored reflected sound arriving at the listener.
Cheers
Sean Olive
Audio Musings (http://docs.google.com/fileview?id=0B97zTRsdcJTfZmM1ZjhlN2MtMTQwZC00MzVmLWE1OTQtMWQzN2IzNGNiMjc0&hl=en)
Scott, just FYI, it's "Sweet" spot.
Sweat spots are what happens when you try to read the manual or I go to church.
What is your problem with mono?
Nothing. I use it myself, along with pink noise to do something like what you do...hear specific non-linearity issues in my builds.
Plus I like really old recordings . My "problem" is the extrapolation to stereophonic soundfield perception predictions. I urge reconsideration.
Do you not believe the research results I've point you towards indicating that a) mono test results track stereo results b) listeners are more discriminating in mono c) that mono signals exist in stereo music recordings and multichannel music/films? Your rationale for requiring stereo signals for loudspeaker tests escapes me given the body of research that suggests otherwise.
Why did you use a stereo test for the mp3 vs CD? Why not have them listen to a central located speaker in mono?
Check the off-axis measurements for speaker C in slide 28 (http://docs.google.com/fileview?id=0B97zTRsdcJTfZmM1ZjhlN2MtMTQwZC00MzVmLWE1OTQtMWQzN2IzNGNiMjc0&hl=en) compared to the Infinity. The mid/treble is significantly down in level relative to the bass: they will sound very dull as you move off-axis.
Or placed 2' from the sidewalls (stereo listening, as both are intended for), they won't sound as overly bright and smeared as the 362's, due to that same dipole null. Depends on how you view it Sean.
The reflected sounds arriving at the listener will also sound dull and colored compared to the direct sound.
Was that the perception in stereo? Can you link the results of the 362 vs ML's when listened to this way, in stereo, positioned around corners of a typical living room? TIA
Gotta run...
cheers,
AJ
The 360's measured very well by Stereophile and as the 362's predecessor, seems very similar, so I'm hoping your answer is yes.
Yes, the only differences should be cosmetic ones.
The 360s have one woofer per box, while the 362s have two.
I have a pair of 360s and they are very smooth.
Arnie: That's not correct. Both the Primus 360 and 362 (the update) have the same driver compliment: 2 woofers, midrange + tweeter. The drivers are identical except I think the color of the cone or surround changed.
Cheers
Sean Olive
Audio Musings (http://http:/seanolive.blogspot.com)
Why did you use a stereo test for the mp3 vs CD? Why not have them listen to a central located speaker in mono?
We were testing stereo codecs - not loudspeakers -- and the codec might produce spatial artifacts. Loudspeakers don't create spatial artifacts (the artifacts are in the recordings), but they produce differences in apparent image width/depth, which are related to their directivity and frequency response. These difference are apparent in both mono and stereo. No need to test them in both modes.
they won't sound as overly bright and smeared as the 362's, due to that same dipole null. Depends on how you view it Sean.
Smeared?? What does that term mean? Do you greater ASW (apparent source width) or spaciousness? Most listeners consider that a good feature unless they are tight-assed imaging purist puritans [it's getting late so my tongue is getting looser]
Was that the perception in stereo? Can you link the results of the 362 vs ML's when listened to this way, in stereo, positioned around corners of a typical living room? TIA
Gotta run...
The speaker sounds dull off-axis whether in mono or stereo. Do a walk by of that speaker versus the Primus and they are night and day: the timbre of the Primus is much more constant as you walk to the left and right of the reference axis. The electrostatic sounds like someone is putting socks in the vocalist's mouth as you walk to the left or right. There is nothing magical about the psychoacoustics of stereo that changes the character and perception of the off-axis sound (except you may be more forgiving of certain problems).
Stereo does produce a more colored, unstable phantom center image compared to a center channel due to the interaural cross-talk between the ears and the left and right speakers. For this reason, the sweet spot restriction and the spatially deprived experience of 2 channel sound, I personally can't stand listening to stereo loudspeaker recordings anymore except when I'm in another room preferably drinking a glass of Pinot Noir or another Rhone varietal. Multichannel or up-mixed stereo is my preferred choice, although the perfect up-mixer hasn't been developed yet.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
As I have stated in other posts the reality of the one listener limitation of the electrostaic speakers I encountered was plainly obvious with even the most limited auditions. Um, the Revel Ultimas that I had the pleasure of listening to had a sweat spot too. The damage was obviously less severe as one moved from the sweat spot with the Revels but for me it was still enough to call those speakers a one man speaker as well along with every other two channel speaker system I have ever heard. It's just that with electrostats the damage is more severe. I can't say that I have ever heard any two channel system that did not suffer from moving out of the sweat spot. I'm pretty sure the physics dictates that the imaging will always be affected by moving out of the sweat spot. Call me crazy but I didn't go out of my way to get the best sounding system I could afford with the idea of screwing up that sound by sitting somewhere other than the sweat spot. If I owned Revel speakers I would still have a dedicated listening room and listen from the sweat spot.
It's called a "sweet spot" not a "sweat spot" Perhaps this was a Freudian slip you made stemming from the fact the "sweet spot" of your electrostatics is so incredibly tight and small that it causes you to profusely sweat trying to stay within it, thus leaving a permanent sweat mark on the chair
The dispersion of the Revel Salon/Studio II mid/tweeter with its very shallow waveguide is extremely wide and smooth, and not even in the same ballpark compared to that of most electrostatics I've tested. I'm not sure what you are talking about but it makes we wonder if you might be talking about stereo image stability while moving around within the speaker spot versus spatial variance in timbre as you move around the seat.
I'd be more than happy to look at the plans of your facility and make suggestions on how to better situate electrostatic speakers in that room and what sort of treatment to use to further optimise things. But it would just be a starting point. I would also be happy to provide you with a play list of material that works for me as an audiophile when doing such auditions. I am glad to hear you would consider these variables in follow up tests.
I am always willing to experimentally test a question or challenge. This forum stemmed from a challenge I made to the Jonathan Berger Mp3 study, which I felt didn't jive with my experience and others when listening to MP3 at 128 kpbs. It's good to have constructive differences in opinion and challenges, although I always seem to be the one who ends up doing all the work
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Stereo does produce a more colored, unstable phantom center image compared to a center channel due to the interaural cross-talk between the ears and the left and right speakers. For this reason, the sweet spot restriction and the spatially deprived experience of 2 channel sound, I personally can't stand listening to stereo loudspeaker recordings anymore except when I'm in another room preferably drinking a glass of Pinot Noir or another Rhone varietal. Multichannel or up-mixed stereo is my preferred choice, although the perfect up-mixer hasn't been developed yet.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
So what do you listen to when you want to listen to music? IMO most of the good stuff is in two channel stereo.
As I have stated in other posts the reality of the one listener limitation of the electrostaic speakers I encountered was plainly obvious with even the most limited auditions. Um, the Revel Ultimas that I had the pleasure of listening to had a sweat spot too. The damage was obviously less severe as one moved from the sweat spot with the Revels but for me it was still enough to call those speakers a one man speaker as well along with every other two channel speaker system I have ever heard. It's just that with electrostats the damage is more severe. I can't say that I have ever heard any two channel system that did not suffer from moving out of the sweat spot. I'm pretty sure the physics dictates that the imaging will always be affected by moving out of the sweat spot. Call me crazy but I didn't go out of my way to get the best sounding system I could afford with the idea of screwing up that sound by sitting somewhere other than the sweat spot. If I owned Revel speakers I would still have a dedicated listening room and listen from the sweat spot.
It's called a "sweet spot" not a "sweat spot" Perhaps this was a Freudian slip you made stemming from the fact the "sweet spot" of your electrostatics is so incredibly tight and small that it causes you to profusely sweat trying to stay within it, thus leaving a permanent sweat mark on the chair
The dispersion of the Revel Salon/Studio II mid/tweeter with its very shallow waveguide is extremely wide and smooth, and not even in the same ballpark compared to that of most electrostatics I've tested. I'm not sure what you are talking about but it makes we wonder if you might be talking about stereo image stability while moving around within the speaker spot versus spatial variance in timbre as you move around the seat.
I'd be more than happy to look at the plans of your facility and make suggestions on how to better situate electrostatic speakers in that room and what sort of treatment to use to further optimise things. But it would just be a starting point. I would also be happy to provide you with a play list of material that works for me as an audiophile when doing such auditions. I am glad to hear you would consider these variables in follow up tests.
I am always willing to experimentally test a question or challenge. This forum stemmed from a challenge I made to the Jonathan Berger Mp3 study, which I felt didn't jive with my experience and others when listening to MP3 at 128 kpbs. It's good to have constructive differences in opinion and challenges, although I always seem to be the one who ends up doing all the work
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Not so much a Freudian slip as a consequence of growing up with dyslexia. My spelling simply sucks. I don't sweat it much.
I'm simply pointing out that the sweet spot may be tighter for electrostatic speakers and more destructive when you move out of it but that all speakers have the sweet spot with two channel stereo and the sonic performance of any speaker system suffers as one moves away from the ideal listening position. I don't get up and walk around when listening to music. That certainly destroys any illusion of an aural soundspace. Yes it is worse with electorstatic speakers. I notice it when i get up to get a drink or go to the bathroom. It doesn't bother me. So I don't see how the more severe breakdown outside the sweet spot becomes a failure in the design of the speakers. Certainly it is an inconvenience if you want to wander the room or if, for whatever reason ,you need to move your head around a lot. But I think it is obvious that electrostatic speakers are not for headbangers anyway. That was the point of the Formula One analogy. High end electrostatic speakers are built for performance not convenience. The old Martin Logan CLSs could create an uncanny illusion of live music with excellent source material, the right room, the right set up and the right equipment. At 2,400 dollars they were pretty unique in that ability at that time. My Soundlabs do much the same thing only with a much wider variety of music. At 3,000 dollars used they are pretty tough to beat for an illusion of live music from a wide range of source material.
When I am back in L.A. I'd be happy to come by and help you with any follow up tests of electrostatic speakers.
I still don't get the argument. I understand, that an "exclusive" sweet spot can be conceived as a feature of high priced gear. But what does this discussion have to do with the fact that frequency response within the sweet spot is also worse than that of the 362?
I still don't get the argument. I understand, that an "exclusive" sweet spot can be conceived as a feature of high priced gear. But what does this discussion have to do with the fact that frequency response within the sweet spot is also worse than that of the 362?
1. It is not a "feature" it is an inconvenience. the question is whether or not that inconvenience is a deal breaker. For me it is not. Apparently for Sean it is. He called it a failed design because of that. The conversation moved from the specific speakers to the nature of electrostatic speakers in general. There are no assumptions about the on axis frequency response of all high end electrostatic speakers in the current discussion of the sweet spot they share as a family of speakers and whether or not that inconvenience constitutes a failed design.
Have you conducted the same experiment by gender? I would find those results very interesting.
Don't women have a higher upper limit anyway? That one woman's 100% CD preference seems to support that.
It's possible that it's been suggested by older data taken for example when men were more likely to have spent time working in heavy industry with inadequate hearing protection, but I've never heard anyone claim women
naturally have a higher upper limit to hearing. If anyone knows otherwise please send me a reference since I'd be interested to read.
Dude take a chill pill. One doesn't have to argue for a personal aesthetic choice so such a post would be kind of silly.
I thought there was a discussion about controlled and repeatable measurements and disciplined judgements of quality going on, not an exchange of aesthetic choices (about which, also, reasoned discussion can be undertaken, but not on HA).
But I am interested. I have a smallish room which I could set up as primarily a listening room, where I would listen by myself. I wouldn't normally think of spending thousands of bucks on speakers, but if I were persuaded that it would really make a difference, I could spring. My musical tastes are various, but I'm especially concerned with Bach solo instrumental and chamber music, Haydn and Shostakovich. What sort of benefit might I expect to get from electrostatics, compared with more modest speakers that measure flat, like the Infinity speakers mentioned here?
The 360's measured very well by Stereophile and as the 362's predecessor, seems very similar, so I'm hoping your answer is yes.
Yes, the only differences should be cosmetic ones.
The 360s have one woofer per box, while the 362s have two.
I have a pair of 360s and they are very smooth.
Arnie: That's not correct. Both the Primus 360 and 362 (the update) have the same driver compliment: 2 woofers, midrange + tweeter. The drivers are identical except I think the color of the cone or surround changed.
Audio Musings (http://http:/seanolive.blogspot.com)
I just double checked my speakers and I misrembered the model numbers. You're right, and I'm wong. Sorry.
But, they still sound great! ;-) I think that speakers this good in this price range are quite a technical accomplishment.
Dude take a chill pill. One doesn't have to argue for a personal aesthetic choice so such a post would be kind of silly.
I thought there was a discussion about controlled and repeatable measurements and disciplined judgements of quality going on, not an exchange of aesthetic choices (about which, also, reasoned discussion can be undertaken, but not on HA).
But I am interested. I have a smallish room which I could set up as primarily a listening room, where I would listen by myself. I wouldn't normally think of spending thousands of bucks on speakers, but if I were persuaded that it would really make a difference, I could spring. My musical tastes are various, but I'm especially concerned with Bach solo instrumental and chamber music, Haydn and Shostakovich. What sort of benefit might I expect to get from electrostatics, compared with more modest speakers that measure flat, like the Infinity speakers mentioned here?
It was in response to Sean's comment "I personally can't stand listening to stereo loudspeaker recordings anymore except when I'm in another room preferably drinking a glass of Pinot." Looks like an aesthetic choice or at least an easthetic reaction to me. I simply thought it was an aesthetic value that carries a tremendous handicap for any music lover given the body of work recorded in stereo for stereo speaker playback. so I asked Sean what he personally listens to given his personal aesthetic dislike for two channel playback and the obvious handicap it presents any music lover.
IME with that sort of music in particular in a smallish dedicated listening room one simply gets a more convincing illusion of live music in the original space with electrostatic speakers. It's almost the ideal listener needs for one to consider some of the less expensive models from Soundlab. I am a big fan of going used. You can get some real bargains. But you should hear them first. Here are some fine speakers (IMO) for reasonable prices that fit the bill.
http://www.audiogon.com/cgi-bin/cls.pl?spk...mp;/Quad-ESL-63 (http://www.audiogon.com/cgi-bin/cls.pl?spkrplan&1282059174&/Quad-ESL-63)
http://www.audiogon.com/cgi-bin/cls.pl?spk...mp;/Quad-ESL-63 (http://www.audiogon.com/cgi-bin/cls.pl?spkrplan&1281315022&/Quad-ESL-63)
http://www.audiogon.com/cgi-bin/cls.pl?spk...p;/Quad-ESL-988 (http://www.audiogon.com/cgi-bin/cls.pl?spkrplan&1281037726&/Quad-ESL-988)
http://www.audiogon.com/cgi-bin/cls.pl?spk...p;/Quad-esl-988 (http://www.audiogon.com/cgi-bin/cls.pl?spkrplan&1280888662&/Quad-esl-988)
http://www.audiogon.com/cgi-bin/cls.pl?spk...p;/Quad-esl-989 (http://www.audiogon.com/cgi-bin/cls.pl?spkrplan&1280888534&/Quad-esl-989)
I didn't see any ML CLS IIzs or any lower end Soundlabs for sale at Audiogon hence all the Quads.
The break in on Martin Logans is real. I have done side by side comparisons between brand new and broken in MLs. It is a substantial difference. You are dealing with a very different driver here. Dealer demo speakers do tend to have more than 72 hours on them so it should not be an issue with any auditions using dealer demo speakers.
An obvious TOS 8 infracftion. :-(
[quote name='analog scott' date='Jun 23 2010, 06:49' post='711099']
" I personally can't stand listening to stereo loudspeaker recordings anymore except when I'm in another room preferably drinking a glass of Pinot."
Note that the text above does not seem to come Sean's OP in the matter. It appears that Scott rooted about in the collected writings of Sean Olive until he could find something that he could take out of context and present as incriminating evidence.
It looks to be like this whole thread has veered solidly off-topic. Instead of being about the preferences of Gen Y listeners, it has turned into a rehash of the battle of 2-channel stereo versus surround sound.
The break in on Martin Logans is real. I have done side by side comparisons between brand new and broken in MLs. It is a substantial difference. You are dealing with a very different driver here. Dealer demo speakers do tend to have more than 72 hours on them so it should not be an issue with any auditions using dealer demo speakers.
An obvious TOS 8 infracftion. :-(
Perhaps but maybe a little context would be helpful. the reason I compared them was because I though my upgraded pannels and/or traansformer was defective. I brought the speakers back to the dealer to demonstrate the gross problems. In a side by side comaprison the problems were quite apparent. The dealer insisted this was a "break in issue." I didn't believe it. The differences were to gross IMO. I suppose....I could have imagined the differences between the new pannels due to bias effects.....
I reluctantly took them back home vowing to return them if the "break in" didn't fix the problem. After one hundred hours of continuous play at loud levels I took them straight back without even a re-evaluation due to my skepticism about that kind of "break in." Guess who went home with a little egg on his face after the second side by side comparison. They were now pretty close to indistinguishable. Reverse subliminal bias expectation effect? Possibly.....Sure surprised me.
IME with that sort of music in particular in a smallish dedicated listening room one simply gets a more convincing illusion of live music in the original space with electrostatic speakers.
I guess I'm especially interested in what it is about electrostatics specifically. The claim for illusion of presence of live music is one that seems to be made for all excellent gear, and IME, using only a couple of cheapy little KEFs, that depends a very great deal on the recording. MA Recordings, for instance, have a couple of disks of Bach on lute which definitely convey an ambiance, but which are also so close-miked that the sound of fingers on strings is actually distracting. But other recordings, not so much, with the same speakers or headphones. So, electrostatics aren't as linear as other speakers (this seems to be non-controversial); what do they win in exchange? Is it to do with having a larger driving area? My only experience is hearing some Quads long long ago, when I was hardly a critical listener.
" I personally can't stand listening to stereo loudspeaker recordings anymore except when I'm in another room preferably drinking a glass of Pinot."
Note that the text above does not seem to come Sean's OP in the matter. It appears that Scott rooted about in the collected writings of Sean Olive until he could find something that he could take out of context and present as incriminating evidence.
It looks to be like this whole thread has veered solidly off-topic. Instead of being about the preferences of Gen Y listeners, it has turned into a rehash of the battle of 2-channel stereo versus surround sound.
Hopefully the mods won't be too hard on Sean for steering so far off topic with his comment on stereo speaker playback. Not sure what "battle" you are refering to. I have no problem with Sean's preference for mulitchannel. I just was wondering how he deals with the fact that so much music is in stereo and designed to be heard on stereo speakers. I don't see how that is any attempt on my part to "incriminate" anyone or anything.
IME with that sort of music in particular in a smallish dedicated listening room one simply gets a more convincing illusion of live music in the original space with electrostatic speakers.
I guess I'm especially interested in what it is about electrostatics specifically. The claim for illusion of presence of live music is one that seems to be made for all excellent gear, and IME, using only a couple of cheapy little KEFs, that depends a very great deal on the recording. MA Recordings, for instance, have a couple of disks of Bach on lute which definitely convey an ambiance, but which are also so close-miked that the sound of fingers on strings is actually distracting. But other recordings, not so much, with the same speakers or headphones. So, electrostatics aren't as linear as other speakers (this seems to be non-controversial); what do they win in exchange? Is it to do with having a larger driving area? My only experience is hearing some Quads long long ago, when I was hardly a critical listener.
I would suggest listening again as a critical listener. If your experience runs contrary to mine then you don't need an explination as to why. But if you get the same impression then I suppose the question is worth persuing. But we can only ask why the experience is what it is after the experience. I can offer some *speculation* in private if you like as to what the advantages are. But not here. It is speculation.
Sean just introduced his opinion of two channel stereo to the thread.
He introduced an argumentation, not an opinion. So a good way of following up would picking up the elements that you consider flawed. A bad way is calling the argument as a whole an opinion and posting yours beside.
" I personally can't stand listening to stereo loudspeaker recordings anymore except when I'm in another room preferably drinking a glass of Pinot."
Looks a whole lot like a comment that reflects a personal opinion to me. I don't see any "argumentation" present in that comment. Do you? Really? And, again, the *question* I asked in response is a fair one given the body of recorded music is mostly two channel recordings designed to be listened to with a stereo pair of loudspeakers. And.... it_was_a_question. There was no "argument" form me nor any assertion of an "alternative opinion."
I still listen to 2-channel stereo music, but it's seldom listened to through 2 speakers: it's listened through 5 to 7 channels via an up-mixer like Logic 7.Sorry if I didn't make that clear. The up-mixing gives me a center channel (missing in stereo,)a much wider sweet spot (also missing), and a sense of envelopment and spaciousness that is entirely devoid in 2 channel reproduction. I'd rather be in the concert hall listening to Bach that looking through a window into the concert hall.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
" I personally can't stand listening to stereo loudspeaker recordings anymore except when I'm in another room preferably drinking a glass of Pinot."
Note that the text above does not seem to come Sean's OP in the matter. It appears that Scott rooted about in the collected writings of Sean Olive until he could find something that he could take out of context and present as incriminating evidence.
It looks to be like this whole thread has veered solidly off-topic. Instead of being about the preferences of Gen Y listeners, it has turned into a rehash of the battle of 2-channel stereo versus surround sound.
I agree that the discussion on 2 channel versus multichannel is off topic and I am happy to abandon it.
I only brought it up because we seem to have people in this forum willing to sit in a tiny sweet spot while listening to overpriced, directional speakers with terrible off-axis response that we are told are both highly room-dependent and loudspeaker/listener position dependent -- all for the purpose of what? To listen to stereo, which Bell Lab scientists said back in the early 1930s was completely inadequate to convey the realism of a live music performance to an audience.
For me, that is a terribly misguided use of effort and money because there too many inherent compromises in sound quality, given what is possible today with music recording and reproduction science and technology. We can do so much better, and we should.
As an industry we have failed to learn and acknowledge the Bell Lab science that is almost 80 years old! And we are now repeating ourselves by ignoring the loudspeaker science that has been known since the mid-1980s from Floyd Tooles' work at the National Research Council.
That is the last word I will say about loudspeakers and stereo versus multichannel in this particular forum. Let's talk about my study and high school studentl preferences for lossless versus low quality lossy music formats.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
To return, I hope, to the thread topic, it makes sense to me to reduce variables as much as possible in such a study. Stereo imaging is affected by things other than the frequency response of a speaker so it makes sense to eliminate it as a variable since a good speaker with a +-3db response over the audio range but +-2db matching errors between pairs at the listening point may still have worse stereo imaging than a basically worse speaker with a +- 5db curve but still maintaining +-0.5db matching between pairs, assuming everything else is equal.
But, for the purposes of the study, perhaps the variations in sound quality shouldn't be done by varying the speakers, but by varying the source itself. Say, switching an equalizer in and out and then seeing if listeners then preferred the flatter response.
I realize you have sophisticated speaker moving capabilities but perhaps in this case you are tempted by the technology you have available to use that when it might be better not to. Just a thought.
At about this point, I suspect Dr Olive is getting a bit exasperated. As the blog entry shows, this was an opportunistic test, to have a rather more disciplined look at the claims that The Young are so corrupted that they prefer MP3 to the real thing. Every once in a while, The Young delight one by having better sense than they're normally credited with, and what is, after all, a quick and dirty test supports the view that the original claim was highly dodgy.
And while demonstrating the technology, keep it real by seeing what sort of speaker they prefer. Surprise, and this is perhaps more of a surprise, they prefer nice accurate speakers. But maybe the sort of high school kids who go on a tour of an acoustic research lab aren't the target for doof-doof. You wouldn't use this finding as the basis for your designs for the youf market, but it is nice to know about it.
As for it not being fair to test electrostatics as mono: as I thought, Quad electrostatics were on the market in 1957, according to Wikipedia, which is a few years before stereo became at all widely available in the UK.
To return, I hope, to the thread topic, it makes sense to me to reduce variables as much as possible in such a study. Stereo imaging is affected by things other than the frequency response of a speaker so it makes sense to eliminate it as a variable since a good speaker with a +-3db response over the audio range but +-2db matching errors between pairs at the listening point may still have worse stereo imaging than a basically worse speaker with a +- 5db curve but still maintaining +-0.5db matching between pairs, assuming everything else is equal.
But, for the purposes of the study, perhaps the variations in sound quality shouldn't be done by varying the speakers, but by varying the source itself. Say, switching an equalizer in and out and then seeing if listeners then preferred the flatter response.
I realize you have sophisticated speaker moving capabilities but perhaps in this case you are tempted by the technology you have available to use that when it might be better not to. Just a thought.
This isn't really on topic either but your questions are too tempting for me to not answer.
I've been doing some cross-cultural research on preferred spectral balance of music reproduction where we are only manipulating the spectral response of a linear loudspeaker (not speaker C) via equalization. The advantages here is that you can keep factors like distortion, directivity, speaker position and its interaction with room acoustics constant w/o the need for a speaker mover. I've been testing the sound quality preferences of Japanese, Chinese and other cultures to see if their tastes in sound quality are different or is good sound universal? I will probably be talking about this at a proposed workshop on the subject at the upcoming AES Convention in San Francisco in November. We hope to run listening experiments over the first few days at the convention and present the results on the last day.
If you capture the loudspeaker/room as a set of binaural room impulses reproduced over head-tracked headphones (http://seanolive.blogspot.com/2009/03/binaural-room-scanning-powerful-tool.html), you can repeat the experiment anywhere in the world using a portable BRS playback system. No listening room, speaker mover or loudspeakers required.
We didn't have time to run the High School students through such an experiment. Besides it is also very useful to know preference of different market segments using the actual products versus only a controlled subset of the variables.
There is some serious market research going on within Harman these days, and I will become more involved in listening tests and sound quality research driven by important market research questions.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
We were testing stereo codecs - not loudspeakers -- and the codec might produce spatial artifacts.
You used stereo because that is exactly how you would focus on spatial perception, just like you used mono to focus on very specific non-linearities, such as resonances...in a loudspeaker. You used stereo for the codec because
that is exactly how the end user is going to perceive it, yet you introduce a completely artificial environment and method for perceiving one channel of what is clearly a stereo speaker system, even though your own research (and basic common sense) indicates that our perception of the multi source (stereo) soundfield is
not the same as a single source (mono) - which is an utterly abstract way of perceiving the ML's performance. No end user is going to listen to the ML that way, any more than the codec user.
Loudspeakers don't create spatial artifacts (the artifacts are in the recordings), but they produce differences in apparent image width/depth, which are related to their directivity and frequency response.
That is completely contradictory. Quite simply, loudspeaker
do affect the spatial presentation of the source material (what is "accurate" is a whole other can of worms, hence my preference to use the term "realism" from a stored acoustic memory standpoint), including their interaction with the room...as you well know. And yes, the directivity and FR are snapshots of the entire polar field generated by the loudspeaker, which then interacts with the reverberant room...and each other.
No need to test them in both modes.
That is absurd. There is only one way to test our perception of the generated-stereophonic-soundfield-room-interaction-with listener somewhere between...and it isn't in mono with a front wall centered source with listener(s) directly in front. The dipole, large acoustic membrane ML will generate a
significantly different polar field than the piston source monopole, one which an anechoic snapshot of the frontal response will be wholly insufficient to describe. It will give you significant info, just not all of it. And you know as well as I do that those narrow band notches will be far more difficult to hear than see, the downward slope of the treble may indeed sound "dull" compared to the Infinity. Again,
some conclusions may be inferred...but not all, from a single snapshot (if they were both piston source monopoles, I'd agree with you). If you have an FR snapshot (or spatial avg)
from the listener position, with the ML placed in the corner of the room, as it would be listened to...like your codec...it would be informative to see it. Please present it if you have it, thanks.
Smeared?? What does that term mean? Do you greater ASW (apparent source width) or spaciousness? Most listeners consider that a good feature unless they are tight-assed imaging purist puritans [it's getting late so my tongue is getting looser]
"Smear" as in the spatial presentation of a wide dispersion piston source monopole (362) placed 2' from a reflective side wall (really real world living room) causing early reflections. "Smear" that won't be there in the identically positioned from sidewall dipole ML, due to the strong side null. I'm starting to wonder if you've ever listened to these very different polar field producing sources in a real room. Have you?
although I always seem to be the one who ends up doing all the work
There are both 360's and ML's present here in Tampa. I'll see about getting them together to compare perceptions of their respective stereo soundfields in a real living room, by listeners behind a (hopefully) acoustically transparent screen. You assistance and advice would be greatly appreciated. Sound reasonable ?
cheers,
AJ
Odigg,
I was writing my response as you were writing this response. We clearly think along the same lines. I like you
Thanks for the compliment! I'm sure *most* everybody likes you too as the work you do is very valuable to us. A number of us are garage audio scientists and we don't have the resources (or even degrees in the right field) to perform the type of research you are conducting. Your findings are very helpful of cutting through the BS and helping us understand what is really going on in the world of audio and listening.
analog scott - I began to respond in a comprehensive manner to your post, but realized I've already said most of what I want to say. I also would prefer to leave this thread on topic. I will add that I think you are making certain assumptions about the buyers of electrostatic speakers (ML in particular), and the dealers of such speakers, that are not accurate. I say this based on my experience at multiple dealers and having known people who own electrostatic speakers. I don't know if ML has made the same assumptions you have. Some speaker manufacturers do make these assumptions and they very plainly say so when they designate a speaker a "near field" monitor.
Dr. Olive (or somebody at Harmon) probably has better data on the buyers of such speakers. If such data exists and Dr. Olive is able to share it, perhaps another thread can be started to discuss this.
Odigg,
I was writing my response as you were writing this response. We clearly think along the same lines. I like you
Thanks for the compliment! I'm sure *most* everybody likes you too as the work you do is very valuable to us. A number of us are garage audio scientists and we don't have the resources (or even degrees in the right field) to perform the type of research you are conducting. Your findings are very helpful of cutting through the BS and helping us understand what is really going on in the world of audio and listening.
analog scott - I began to respond in a comprehensive manner to your post, but realized I've already said most of what I want to say. I also would prefer to leave this thread on topic. I will add that I think you are making certain assumptions about the buyers of electrostatic speakers (ML in particular), and the dealers of such speakers, that are not accurate. I say this based on my experience at multiple dealers and having known people who own electrostatic speakers. I don't know if ML has made the same assumptions you have. Some speaker manufacturers do make these assumptions and they very plainly say so when they designate a speaker a "near field" monitor.
Dr. Olive (or somebody at Harmon) probably has better data on the buyers of such speakers. If such data exists and Dr. Olive is able to share it, perhaps another thread can be started to discuss this.
How is it that my experience as an owner of electrostatic speakers are "assumptions" and your experience as someone who has not had that experience...is "experience?" I really don't see how your "experience' which is certainly less informed on certain levels is "experience" where as my twenty years experience as an owner who has delt with two different manufacturers of electrostatic speakers as an owner, who has dealt with three different dealers as a customer and, believe it or not, has actually met and/or corisponded with "a few" fellow owners of electrostatic speakers along the way? who is really making the assumptions here?
How is it that my experience as an owner of electrostatic speakers are "assumptions" and your experience as someone who has not had that experience...
You can have 100 years of experience and still count as
noise in a scientific sense, if you refuse to go any further than holistic, singular experience comparisons. Objectivity and repeatability are the driving forces behind scientific success. That requires certain amounts of abstraction. You don't seem to be willing to follow that path. Your arguments are of the same type over and over again and all lead into the same direction: one cannot say anything until you have basically my setup (or that of another ML expert). Those then also cannot present any criteria, but infer from ears and experience. You could go over to the Stereophile forum, where people hug each other for experiences like this, but HA might be the wrong place. Nailing down significants without resentments against abstraction rather seems to be the culture here.
I bet the silent majority of readers with no particular axe to grind are thinking that testing stereo speakers in mono is a strange thing to do.
I've read the justifications. I don't buy them. I can see how on-axis mono listening can give more repeatable results where differences are more easily audible - but maybe that's another way of saying that normal stereo listening depends far more on the interaction between the speakers and the room.
Like it or not, most of us are stuck with interaction between our speakers and our listening rooms. Obviously it would complicate things, but I think many people are interested in which speakers sound best in the locations where the speakers look+fit best in their living room. Other people are interested in which speakers sound best given the opportunity to place the speakers at the optimum location for that pair of speakers. Other people are interested in creating the best listening room, placing the speakers optimally, and getting the best sound.
These are harder questions to answer, and I don't want to make the best the enemy of the good, but the concerns are real.
I'm sure Sean can envisage a Mark II speaker switcher room with a speaker switcher in each corner to deliver stereo tests. How much space have you got Sean?
Cheers,
David.
I bet the silent majority of readers with no particular axe to grind are thinking that testing stereo speakers in mono is a strange thing to do.
No doubt.
I've read the justifications. I don't buy them.
I presume that is related to what you know and don't know about speakers.
While I pesonally have only built a few dozen speakers that were one-offofjects, for the past 30 or more years I've hung pretty closely with a number of people whose accomplisments with respect to speakers are virtually unknown in the world of high end audio, but who people like Sean know eactly about. I'm talking about people who have presented many important AES conference papers and even a few JAES articles. People like Clark and Geddes. I've also hung with people who each have designed complete audio systems that have sold in the 100,000s. People who have contributed indirectly to Sean's employer's bottom line.
I'm not going to claim that I've a speaker expert because these people are my friends, but these people don't work in a vacuum. They each have something like "Kitchen Cabinets" where they present their ideas and look for feedback from people whose opinions they value. I've been part of those discussions and experiments for decades.
I see the current discussion in that context. I so strongly disagree with much that has been said lately about evaluating speakers that I'm tempted to start a thread called "Is there such a thing as a center channel speaker". The point being that While we talk about speakers as being center channel speakers and main speakers and surround and ambience speakers, those are more terms of art and merchandising and conveniece than technical definitions or requirements.
For example, one of the most satisfying and revealing systems for listening to stereo recordings that I know of is nominally a surround system with 9 identical nearly full-range speakers and 2 subwoofers with gargantuan bass extension and dynamic range. Its owner who is merely an AES Fellow, obviously doesn't believe that a no-compromise audio system needs to have speakers whose design is different for the various roles, other than subwoofers. He's hardly unique. Indeed, he's following the same game plan as Sean.
I can see how on-axis mono listening can give more repeatable results where differences are more easily audible - but maybe that's another way of saying that normal stereo listening depends far more on the interaction between the speakers and the room.
I would start from this point. The ideal speaker would be a straight wire that converts electrical signals to sound. In fact we know that is naive. In broad terms, speakers have 2 outputs - their on-axis response and their off-axis response. There is very little evidence that the requirements for on-axis and off-axis response vary with the number of channels given that the final system response is appropriate. There's no doubt that it is less confusing to compare speakers the fewer channels that are active down to no more than two and perhaps one.
Like it or not, most of us are stuck with interaction between our speakers and our listening rooms.
Only in a world where speakers are locked into pre-ordained posistions and orientations and there is no such thing as equalization and adjustment of individual channel levels.
Obviously it would complicate things, but I think many people are interested in which speakers sound best in the locations where the speakers look+fit best in their living room.
Rasing the question of how does one make sound quality as independent as possible from bad choices of location in the room.
Other people are interested in which speakers sound best given the opportunity to place the speakers at the optimum location for that pair of speakers. Other people are interested in creating the best listening room, placing the speakers optimally, and getting the best sound.
The last approach can really work. I would go so far as to say that if you aren't willing to make the room the servant of the speaker systems, you aren't really trying. Dilentantes need not apply.
How is it that my experience as an owner of electrostatic speakers are "assumptions" and your experience as someone who has not had that experience...
You can have 100 years of experience and still count as noise in a scientific sense, if you refuse to go any further than holistic, singular experience comparisons. Objectivity and repeatability are the driving forces behind scientific success. That requires certain amounts of abstraction. You don't seem to be willing to follow that path. Your arguments are of the same type over and over again and all lead into the same direction: one cannot say anything until you have basically my setup (or that of another ML expert). Those then also cannot present any criteria, but infer from ears and experience. You could go over to the Stereophile forum, where people hug each other for experiences like this, but HA might be the wrong place. Nailing down significants without resentments against abstraction rather seems to be the culture here.
You cite your experience which clearly has no advantage over my experience and is pretty clearly less inofrmed than my experience and then you call my experiences assumptions while using your experience as an objective reference. That is simply a double standard and a coupling of two logical fallacies. You appeal to the authority of your own experience and then use a classic logical fallacy of movement of the goal posts against my experience by invoking "science" when attacking the validity of my experience. Are you seriously waving the science flag here? We were talking about whether or not customers were duly informed about the nature of the product! How can you possibly wave the science flag? Your "experience" with how well informed the buying decisions of electrostatic speaker owners is "scientific?"I have spoken of my personal experience in this regard. *I* was duly informed as a customer. I have yet to run across *any* owner of electrostatic speakers that have expressed any problems about having been duly informed and/or educated on the nature of the sweet spot before they bought their speakers. I also clearly noted that my experience was in no way a universal indicator on how every dealer deals with every customer. I also have asserted that anyone (unless they are hearing impared) can and should recognize this particular aspect of any electrostatic speaker that exhibits it with the simplest of auditions. It is my position that for one to be unaware of the sweet spot issues with a given pair of electrostatic speakers one would have had to not do any kind of meaningful audition.
You have some "science" on the matter to offer up that refutes my experiences or my assertions?
I've read the justifications. I don't buy them. I can see how on-axis mono listening can give more repeatable results where differences are more easily audible - but maybe that's another way of saying that normal stereo listening depends far more on the interaction between the speakers and the room.
Well, I haven't actually read the papers Olive mentions, but he says that:
a) mono test results track stereo results b) listeners are more discriminating in mono
do you not believe these? or if you do, do you have a reason to ignore them? because, if the mono results do track the stereo results, just with less "noise", it would seem reasonable to use mono tests, or not?
I bet the silent majority of readers with no particular axe to grind are thinking that testing stereo speakers in mono is a strange thing to do.
I've read the justifications. I don't buy them.
Sounded strange to me too but I'm buying. I don't think we'll get much further by discussing it. What needs to happen next is someone needs to try to reproduce Sean's findings WRT mono vs. stereo listening tests. Has this happened?
Well, I haven't actually read the papers Olive mentions, but he says that:
a) mono test results track stereo results b) listeners are more discriminating in mono
do you not believe these? or if you do, do you have a reason to ignore them? because, if the mono results do track the stereo results, just with less "noise", it would seem reasonable to use mono tests, or not?
I haven't really thought about it before reading this thread, so in terms of what works when developing speakers, I'm happy to defer to those who actually do it.
However, "mono test results track stereo results" - well, yes, kind of. I'm sure I can build a speaker where this doesn't work! I only have to paint the red terminal black, and the black terminal red, on one of a pair of speakers and we all know what will happen. one speaker listening = great. two speaker listening = no bass! There are also speaker designs with side firing drivers - is anyone seriously suggesting they sound or measure the same in Sean's speaker changer as they do toed-in near-ish to the corner of a room?
I think it's unwise to quantify X by measuring Y, just because "X tracks Y" pretty well in most situations. More importantly, if you're going to publish these results, it opens them up to easy and obvious criticism.
As for "listeners are more discriminating in mono", well, if we're trying to find the best value for money, doesn't this nullify the test? Listen in mono, and the results may show an audible improvement (achieved by spending more, maybe) which is inaudible in stereo, and therefore is a waste of money.
Don't get me wrong - I think these experiments are wonderful - I bet that listening blind (rather than sighted) improves the quality of the results 100x more than listening in mono (rather than in stereo) diminishes the quality of the results.
In my own limited experience however, it is the speaker/room interaction that causes me the most problems. To my ears, different rooms (+ speaker placements in those rooms) sound
more different than different speakers...
Rasing the question of how does one make sound quality as independent as possible from bad choices of location in the room.
Yes, exactly. I don't think Sean's current experiments are intended to tell you anything about this at all. Which is fine. But it may mean that the "best" speaker in his test isn't the "best" speaker in my room.
Other people are interested in which speakers sound best given the opportunity to place the speakers at the optimum location for that pair of speakers. Other people are interested in creating the best listening room, placing the speakers optimally, and getting the best sound.
The last approach can really work. I would go so far as to say that if you aren't willing to make the room the servant of the speaker systems, you aren't really trying. Dilentantes need not apply.
Well, indeed , but some of us are married, and even if we are lucky enough to have "spare" rooms in our houses, they're not always the right size or shape or construction to make a good listening room. I know many of us laugh at the amount audiophiles spend on their systems, but if you're going to factor in the cost of a 3x4x5metre room in the UK, that's about £20,000 - £40,000.
Hence the value of a speaker nearly defying the laws of physics, and working really well in a corner of the living room.
Cheers,
David.
Well, I haven't actually read the papers Olive mentions, but he says that:
a) mono test results track stereo results b) listeners are more discriminating in mono
do you not believe these? or if you do, do you have a reason to ignore them? because, if the mono results do track the stereo results, just with less "noise", it would seem reasonable to use mono tests, or not?
I haven't really thought about it before reading this thread, so in terms of what works when developing speakers, I'm happy to defer to those who actually do it.
However, "mono test results track stereo results" - well, yes, kind of. I'm sure I can build a speaker where this doesn't work! I only have to paint the red terminal black, and the black terminal red, on one of a pair of speakers and we all know what will happen. one speaker listening = great. two speaker listening = no bass! There are also speaker designs with side firing drivers - is anyone seriously suggesting they sound or measure the same in Sean's speaker changer as they do toed-in near-ish to the corner of a room?
That is a good point: it cannot be generally true. So it comes down to what type of speakers was used, and in what environment, during those tests. A good point.
I think it's unwise to quantify X by measuring Y, just because "X tracks Y" pretty well in most situations. More importantly, if you're going to publish these results, it opens them up to easy and obvious criticism.
As for "listeners are more discriminating in mono", well, if we're trying to find the best value for money, doesn't this nullify the test? Listen in mono, and the results may show an audible improvement (achieved by spending more, maybe) which is inaudible in stereo, and therefore is a waste of money.
I am not sure what you are saying here. Surely, if you were to accept that results of mono tests "track" those of stereo tests, only with less statistical variation between different observers, they would be useful (yes, I realise you don't accept it, but if). Decreasing the variance is one thing, decreasing the mean another.
Don't get me wrong - I think these experiments are wonderful - I bet that listening blind (rather than sighted) improves the quality of the results 100x more than listening in mono (rather than in stereo) diminishes the quality of the results.
In my own limited experience however, it is the speaker/room interaction that causes me the most problems. To my ears, different rooms (+ speaker placements in those rooms) sound more different than different speakers...
This is true in my experience, too. And using a single speaker will drastically change this. So a good point.
You cite your experience which clearly has no advantage over my experience
Scott, that your experience is more or less advantageous than someone elses experiences is not a given.
You started this subtread with the following:
How is it that my experience as an owner of electrostatic speakers are "assumptions" and your experience as someone who has not had that experience...is "experience?"
Let me address that question by recalling that you recently posted the contents of your system here, and if memory serves, your system included an incredible number of items that are well-known to be audio jewelry - audio components that cannot possibly enchance sound quality. Please correct me if I misremembered that.
From that information Scott, we must conclude that very many of your so-called experiences with audio components are highly likely to be based on false assumptions and not based on actual reliably-heard expereinces.
Well, I haven't actually read the papers Olive mentions, but he says that:
a) mono test results track stereo results b) listeners are more discriminating in mono
do you not believe these? or if you do, do you have a reason to ignore them? because, if the mono results do track the stereo results, just with less "noise", it would seem reasonable to use mono tests, or not?
I haven't really thought about it before reading this thread, so in terms of what works when developing speakers, I'm happy to defer to those who actually do it.
However, "mono test results track stereo results" - well, yes, kind of. I'm sure I can build a speaker where this doesn't work! I only have to paint the red terminal black, and the black terminal red, on one of a pair of speakers and we all know what will happen. one speaker listening = great. two speaker listening = no bass! There are also speaker designs with side firing drivers - is anyone seriously suggesting they sound or measure the same in Sean's speaker changer as they do toed-in near-ish to the corner of a room?
The above are either straw man or excluded-middle arguments depending on how you take them.
Don't get me wrong - I think these experiments are wonderful - I bet that listening blind (rather than sighted) improves the quality of the results 100x more than listening in mono (rather than in stereo) diminishes the quality of the results.
While I don't want to diminish the value of bias controls, it is a fact that bias controls often become less important the more audible the differences are. Bias controls have produced their most compelling results in those cases where there were no actual audible differences, those where the perception of audible differences was completely based on listener bias.
You have to ask yourself what occurs when you put the second speaker into the room. The speaker by itself rarely makes much of a difference until you apply a signal to it. One you apply a signal to the second speaker the consequences diverge depending on whether the signals applied to the speakers correlate with each other. So immediately, you've added a relevant, highly audible variable that does not depend on the speakers at all.
To summarize, if you listen to just one speaker then its sound is dependent on the speaker itself and how it interacts with the room. If you add a second speaker, the sound of the pair of speakers is dependent on each speaker and how it interacts with the room times two, plus the nature and degree of the correlation between the signals applied to the speakers and how it interacts with the room.
In short, adding the second speaker increases the dependencies on things other than the speaker from 1 to 4 or 5 depending on how you count.
I thing that the thing about mono comparisons that people react to the most is our general lack of familiarity with listening in mono. IME one quickly adjusts to this and starts appreciating the ease of making comparisons with fewer dependencies that don't directly relate to the speaker.
In my own limited experience however, it is the speaker/room interaction that causes me the most problems. To my ears, different rooms (+ speaker placements in those rooms) sound more different than different speakers...
I completely agree with that. However, I don't see this fact as working against mono comparisons. Instead I count the variables this creates and find that listening in mono simplifies the comparison. Try it, you may like it! I do.
However, "mono test results track stereo results" - well, yes, kind of. I'm sure I can build a speaker where this doesn't work! I only have to paint the red terminal black, and the black terminal red, on one of a pair of speakers and we all know what will happen. one speaker listening = great. two speaker listening = no bass! There are also speaker designs with side firing drivers - is anyone seriously suggesting they sound or measure the same in Sean's speaker changer as they do toed-in near-ish to the corner of a room?
I think it's unwise to quantify X by measuring Y, just because "X tracks Y" pretty well in most situations. More importantly, if you're going to publish these results, it opens them up to easy and obvious criticism.
As for "listeners are more discriminating in mono", well, if we're trying to find the best value for money, doesn't this nullify the test? Listen in mono, and the results may show an audible improvement (achieved by spending more, maybe) which is inaudible in stereo, and therefore is a waste of money.
If you are trying to make a weather forecast for Europe and find that measuring the pressure in India historically has been a better predictor than using previous measurements of the weather in Europe, what method would you choose?
I have just "accepted" that mono listening tests gives less measurement noise. I did not do any tests of my own and did not try to understand those citations that led to this. If listening in mono really means that the same expected conclusions can be found after 3 hours instead of 6, I am all for it. If listening to dry dull recordings of tamburine instead of reall, engaging music mean faster, more robust conclusions, I am all for it.
-k
As for "listeners are more discriminating in mono", well, if we're trying to find the best value for money, doesn't this nullify the test? Listen in mono, and the results may show an audible improvement (achieved by spending more, maybe) which is inaudible in stereo, and therefore is a waste of money.
I am not sure what you are saying here. Surely, if you were to accept that results of mono tests "track" those of stereo tests, only with less statistical variation between different observers, they would be useful (yes, I realise you don't accept it, but if). Decreasing the variance is one thing, decreasing the mean another.
OK I later realised that you took "more discriminating" to mean, essentially, "harder to please" (ie, they marked speakers down when listened to in mono vs listening in stereo). I took it to mean "there was less variance between different listener's ratings in mono than there was in stereo".
Well, I haven't actually read the papers Olive mentions, but he says that:
a) mono test results track stereo results b) listeners are more discriminating in mono
do you not believe these? or if you do, do you have a reason to ignore them? because, if the mono results do track the stereo results, just with less "noise", it would seem reasonable to use mono tests, or not?
I haven't really thought about it before reading this thread, so in terms of what works when developing speakers, I'm happy to defer to those who actually do it.
However, "mono test results track stereo results" - well, yes, kind of. I'm sure I can build a speaker where this doesn't work! I only have to paint the red terminal black, and the black terminal red, on one of a pair of speakers and we all know what will happen. one speaker listening = great. two speaker listening = no bass! There are also speaker designs with side firing drivers - is anyone seriously suggesting they sound or measure the same in Sean's speaker changer as they do toed-in near-ish to the corner of a room?
The above are either straw man or excluded-middle arguments depending on how you take them.
Don't get me wrong - I think these experiments are wonderful - I bet that listening blind (rather than sighted) improves the quality of the results 100x more than listening in mono (rather than in stereo) diminishes the quality of the results.
While I don't want to diminish the value of bias controls, it is a fact that bias controls often become less important the more audible the differences are. Bias controls have produced their most compelling results in those cases where there were no actual audible differences, those where the perception of audible differences was completely based on listener bias.
You have to ask yourself what occurs when you put the second speaker into the room. The speaker by itself rarely makes much of a difference until you apply a signal to it. One you apply a signal to the second speaker the consequences diverge depending on whether the signals applied to the speakers correlate with each other. So immediately, you've added a relevant, highly audible variable that does not depend on the speakers at all.
To summarize, if you listen to just one speaker then its sound is dependent on the speaker itself and how it interacts with the room. If you add a second speaker, the sound of the pair of speakers is dependent on each speaker and how it interacts with the room times two, plus the nature and degree of the correlation between the signals applied to the speakers and how it interacts with the room.
In short, adding the second speaker increases the dependencies on things other than the speaker from 1 to 4 or 5 depending on how you count.
I thing that the thing about mono comparisons that people react to the most is our general lack of familiarity with listening in mono. IME one quickly adjusts to this and starts appreciating the ease of making comparisons with fewer dependencies that don't directly relate to the speaker.
In my own limited experience however, it is the speaker/room interaction that causes me the most problems. To my ears, different rooms (+ speaker placements in those rooms) sound more different than different speakers...
I completely agree with that. However, I don't see this fact as working against mono comparisons. Instead I count the variables this creates and find that listening in mono simplifies the comparison. Try it, you may like it! I do.
While I appreciate you attributing all those wise words to me, I didn't actually write them
I agree with you, but I think that his position amounts to: speakers and room are inseparable in reality, so isolating one of the two won't necessarily give a meaningful result.
(PS: I left all the quotes in place on purpose, otherwise my comment does not make sense)
Speakers and room are inseparable in reality, so isolating one of the two won't necessarily give a meaningful result.
This statement contradicts what Sean has found in controlled listening tests. If you have not done any of your own tests, as per TOS#8, I don't think you're in a position to make such statements.
Speakers and room are inseparable in reality, so isolating one of the two won't necessarily give a meaningful result.
This statement contradicts what Sean has found in controlled listening tests. If you have not done any of your own tests, as per TOS#8, I don't think you're in a position to make such statements.
But what I actually wrote was: "I think that his position amounts to: speakers and room are inseparable in reality, so isolating one of the two won't necessarily give a meaningful result."
You cite your experience which clearly has no advantage over my experience
Scott, that your experience is more or less advantageous than someone elses experiences is not a given.
You started this subtread with the following:
How is it that my experience as an owner of electrostatic speakers are "assumptions" and your experience as someone who has not had that experience...is "experience?"
Let me address that question by recalling that you recently posted the contents of your system here, and if memory serves, your system included an incredible number of items that are well-known to be audio jewelry - audio components that cannot possibly enchance sound quality. Please correct me if I misremembered that.
From that information Scott, we must conclude that very many of your so-called experiences with audio components are highly likely to be based on false assumptions and not based on actual reliably-heard expereinces.
The issue was whether or not customers of ML or other electrostatic speakers were duly informed on the nature of the sweet spot. What does the rest of my system have to do with anything in regards to my experience in being duly informed as an actual person who bought electrostatic speakers from a dealer and has dealt with two other dealers in regards to electrostatic speakers since then? What does it have to do with my discussions over the years with many other owners of electrostatic speakers? What audio "jewelry" are you refering to? The diamond stylus of my cartridge is an actual functional part, not jewelry. Can't think of anything else you might be refering to. Do I also need to cite the multiple logical fallacies contained in your so called "must" conclusion?
What audio "jewelry" are you refering to? The diamond stylus of my cartridge is an actual functional part, not jewelry. Can't think of anything else you might be refering to.
Not sure what particular piece of jewelry in the collection Arnie might be referring to, but since this thread involves the soundfield generated by loudspeakers, could you tell us specifically how the "Bybee Pro Filter" and "VPI Brick" (http://cgi.audioasylum.com/systems/3229.html) physically effect the soundwaves emanating from the Soundlabs?
TIA
What audio "jewelry" are you refering to? The diamond stylus of my cartridge is an actual functional part, not jewelry. Can't think of anything else you might be refering to.
Not sure what particular piece of jewelry in the collection Arnie might be referring to, but since this thread involves the soundfield generated by loudspeakers, could you tell us specifically how the "Bybee Pro Filter" and "VPI Brick" (http://cgi.audioasylum.com/systems/3229.html) physically effect the soundwaves emanating from the Soundlabs?
TIA
The Bybee filter was something I tried on a whim since i was able to get one for very little money compared to it's retail price. Can't say that it made any difference whatsoever. I sold it for a little more than my purchase price. The VPI bricks (two of them) were given to me. I'm not sure if they make that much of a difference or not but they sit on my Vandersteen subwooofer. Without them my two Emmy's that also sit on the subwoofer tend to ring when the bass gets loud. With the VPI bricks the on top of the sub the sub's top seems to be damped enough to stop the ringing. I don't like hearing the Emmys ring so that does make an audible improvement. I suppose the Emmys could be seen as audio jewelry since they sit on the sub, and are gold plated and very shiny??? I never thought of them as an actual part of the system though.
I so strongly disagree with much that has been said lately about evaluating speakers that I'm tempted to start a thread called "Is there such a thing as a center channel speaker". The point being that While we talk about speakers as being center channel speakers and main speakers and surround and ambience speakers, those are more terms of art and merchandising and conveniece than technical definitions or requirements.
For example, one of the most satisfying and revealing systems for listening to stereo recordings that I know of is nominally a surround system with 9 identical nearly full-range speakers and 2 subwoofers with gargantuan bass extension and dynamic range. Its owner who is merely an AES Fellow, obviously doesn't believe that a no-compromise audio system needs to have speakers whose design is different for the various roles, other than subwoofers. He's hardly unique. Indeed, he's following the same game plan as Sean.
Thank you Arnold. This clarifies things for me a great deal.
The above are either straw man or excluded-middle arguments depending on how you take them.
OP: "mono test results track stereo results"
Me: "not always, here's two trivial counter examples"
You: "straw man"
HA search says you've claimed "straw man" 11 times now. You should build a haystack.
In my own limited experience however, it is the speaker/room interaction that causes me the most problems. To my ears, different rooms (+ speaker placements in those rooms) sound more different than different speakers...
I completely agree with that. However, I don't see this fact as working against mono comparisons. Instead I count the variables this creates and find that listening in mono simplifies the comparison. Try it, you may like it! I do.
Well, kind of, but it doesn't tell you how much the speaker's sound will fall apart as its pushed towards the corners of my living room.
Cheers,
David.
yet you introduce a completely artificial environment and method for perceiving one channel of what is clearly a stereo speaker system, even though your own research (and basic common sense) indicates that our perception of the multi source (stereo) soundfield is not the same as a single source (mono) - which is an utterly abstract way of perceiving the ML's performance. No end user is going to listen to the ML that way, any more than the codec user.
Look, if you don't accept the rationale and published scientific evidence for doing mono loudspeaker evaluations that is your prerogative. I've already proven this to myself and a group of scientists who peer-reviewed my work, as has Floyd Toole in his landmark papers on loudspeaker evaluation back in the 1980s. Besides Infinity, JBL, and Revel, many others companies like Axiom, Paradigm, PSB, B&O,etc also do mono loudspeaker comparisons because they understand and accept the science, and have probably confirmed the science works in practice.
The published science tells us that mono comparisons are more revealing of problems with the loudspeakers, and therefore produce a more sensitive test, which is one of the cornerstones of a good subjective measurement. If a loudspeaker wins the test in mono with the highest overall ratings, it also will win the test in stereo: The spectral and spatial ratings also track across mono and stereo tests. The ecological validity of using mono signals is also legitimate since many stereo and surround recordings contain strong monophonic signal components.
It's all summarized in Floyd Toole's book "Sound Reproduction", section 8.2.1 "Testing the Effects of Loudspeaker Directivity on Imaging and Space". He shows an example where two direct radiator loudspeakers (a 3-way and a 2-way) and a full-range dipole electrostatic loudspeaker (Quad) are evaluated in both mono and stereo using double-blind controlled listening tests. This is very relevant to our discussion since the Martin Logan I tested and the Quad Floyd tested are both electrostatic dipoles (although the ML is a hybrid with a monopole subwoofer, and the Quad is a full range dipole).
In this test, the Quad happened to have the better off-axis performance of the three speakers but was it was the most directional of the three loudspeakers, with much reduced dispersion (look at the anechoic measurements). If you look at Figure 8.12, you are shown the mean sound quality (timbre + distortion) and spatial quality scores of each loudspeaker in both mono and stereo tests. It shows us that the Quad (speaker BB) had the lowest sound quality and spatial quality ratings for both mono and stereo tests. The results track across mono and stereo. What happens in stereo, however, is that the recording strongly comes into play (note: recordings are more of a nuisance variable in stereo evaluations in general because the stereophonic recording techniques strongly dictate the spatial imagery we hear).
Toole argues that the decorrelated stereo signals in the recordings produced spaciousness (lower IACC values at the ears) that compensated for the Quad's lack of perceived spaciousness in the mono tests (note: the wider dispersion direct radiator loudspeakers produced stronger lateral reflections that are known to increase the apparent source width and spaciousness of the spatial imagery). As a result, the Quad loudspeaker gets higher scores in stereo - but still not as high as the other two loudspeakers. The rank order of the loudspeakers perfectly tracked across mono and stereo playback conditions -- even for the electrostatic dipole.
Toole shows that the stereo ratings for the Quad strongly depend on the stereo information in the recordings (not a good attribute for a loudspeaker unless you only purchase recording made with spaced stereo microphones). For recordings that had less decorrelated signals (e.g. +90% of all pop, jazz recordings that typically use pan potted stereo signals, and some classical recordings that use coincident stereo microphone techniques), the Quad quality ratings dropped in the stereo evaluations, and the less directional speaker's ratings increased.
The scientific proof is there. If you choose not to accept it - that is your choice. If you want to challenge it, then i suggest you do your own double-blind controlled listening experiments on dipole speakers in both mono and stereo, and report back the results.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Thanks for your patience Sean.
This bit...
Toole shows that the stereo ratings for the Quad strongly depend on the stereo information in the recordings (not a good attribute for a loudspeaker unless you only purchase recording made with spaced stereo microphones). For recordings that had less decorrelated signals (e.g. +90% of all pop, jazz recordings that typically use pan potted stereo signals, and some classical recordings that use coincident stereo microphone techniques), the Quad quality ratings dropped in the stereo evaluations, and the less directional speaker's ratings increased.
...is especially interesting.
Cheers,
David.
Odigg,
I was writing my response as you were writing this response. We clearly think along the same lines. I like you
Thanks for the compliment! I'm sure *most* everybody likes you too as the work you do is very valuable to us. A number of us are garage audio scientists and we don't have the resources (or even degrees in the right field) to perform the type of research you are conducting. Your findings are very helpful of cutting through the BS and helping us understand what is really going on in the world of audio and listening.
analog scott - I began to respond in a comprehensive manner to your post, but realized I've already said most of what I want to say. I also would prefer to leave this thread on topic. I will add that I think you are making certain assumptions about the buyers of electrostatic speakers (ML in particular), and the dealers of such speakers, that are not accurate. I say this based on my experience at multiple dealers and having known people who own electrostatic speakers. I don't know if ML has made the same assumptions you have. Some speaker manufacturers do make these assumptions and they very plainly say so when they designate a speaker a "near field" monitor.
Dr. Olive (or somebody at Harmon) probably has better data on the buyers of such speakers. If such data exists and Dr. Olive is able to share it, perhaps another thread can be started to discuss this.
I'm sure we have some demographic data on electrostatic speaker buyers.
I recently purchased a pair of ML speakers at Magnolia for purposes of testing and evaluation. I asked the salesman if I could purchase just one 1 speaker since I like to listen in mono. The sales guy thought I was nuts -- just like many of the people in this forum (I can't remember if he forced me to buy 1 or 2 speakers). But then I also purchased several other speakers from different companies, and he instantly forgot how nuts I was.
He told me not to worry if the loudspeaker sounded "unusual" at first since it takes 72 hours or even longer to break in (and this isn't even one of their electrostatic models). I told him about human adaptation to sound over time, and how in isolation we grow used to things sounding normal that aren't -- like MP3 @ 128kbps. That was exactly the explanation behind Berger's MP3 results, that listeners had grown accustomed to that type of sound. I smiled and told him not to worry, that I wouldn't be returning the ML speakers after 72 hours. I could sense the relief in his voice after I told him this.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Look, if you don't accept the rationale and published scientific evidence for doing mono loudspeaker evaluations that is your prerogative. I've already proven this to myself and a group of scientists who peer-reviewed my work, as has Floyd Toole in his landmark papers on loudspeaker evaluation back in the 1980s. Besides Infinity, JBL, and Revel, many others companies like Axiom, Paradigm, PSB, B&O,etc also do mono loudspeaker comparisons because they understand and accept the science, and have probably confirmed the science works in practice.
Are you sure about that? Perhaps you can scratch Axiom off that list Audioholics Axiom Blind Testing (http://www.audioholics.com/news/editorials/axiom-blind-listening-test)
Many companies conduct their listening tests in mono, claiming if a speaker wins in mono it will almost always win in stereo. While this may be a good approach for listening tests that host multiple listeners NOT sitting near the sweet spot, I don’t believe you get the whole picture on how a speaker system interacts as a pair listening in mono. Ian Colquhoun and I are on the same page here, which is why all of Axiom’s testing is done in stereo.
You may want to check with the others as well (obviously not JBL, Infinity or Revel ).
The published science tells us that mono comparisons are more revealing of problems with the loudspeakers
Yes,
specific problems, as I've repeatedly agreed to. Just like you listened to the codecs in stereo, not mono, to listen for
specific reproduction problems, as would be faced by the (stereo+ listening) end user, not some abstraction.
...and therefore produce a more sensitive test, which is one of the cornerstones of a good subjective measurement. If a loudspeaker wins the test in mono with the highest overall ratings, it also will win the test in stereo: The spectral and spatial ratings also track across mono and stereo tests.
So where are the stereo ratings test results of the ML vs to 362 so we can confirm this asserted correlation? This may very well be the case, but I don't see the data.
The scientific proof is there.
Yes, that we are
less sensitive to issues that we may be more sensitive to in mono, when we listen in stereo...or more channels. As you have confirmed.
If you choose not to accept it - that is your choice. If you want to challenge it, then i suggest you do your own double-blind controlled listening experiments on dipole speakers in both mono and stereo, and report back the results.
I'll do my best, hopefully with your help . (Not sure how "double" the blind part will be if I administer, even hidden from the test subjects).
Question #1. Where do you source your screen cloth to hide the speakers from view? What is the measured FR loss of this material? TIA
cheers,
AJ
Are you sure about that? Perhaps you can scratch Axiom off that list Audioholics Axiom Blind Testing (http://www.audioholics.com/news/editorials/axiom-blind-listening-test)
OK. That is news to me, since Ian did mono tests at NRC when he used those facilities. I'm actually quite happy to discover that competitors are now testing their speakers in stereo rather than mono because it means that their test results are not as sensitive as ours.
Both Floyd Toole and I recently had many conversations with Gene D. at Audioholics about why he should do mono loudspeaker comparisons. Unfortunately, he chose to ignore our recommendations in his last loudspeaker comparison shoot-out for the same arguments that you make. So what was the outcome of his stereo loudspeaker test results? I statistically analyzed the data and there were no statistically significance differences in listeners' preference ratings among any of the loudspeakers tested: it was just noise. Of course, several other important listening test nuisance variables were not controlled (e.g. seat and loudspeaker position, and a grille cloth blind screen with 1-2dB broadband effects) but it would have been interesting to see if they had less noise in the data if the tests were conducted in mono to confirm what we've know from 25 years of running these tests.
[note: I don;t mean to dismiss Gene's well-intentioned efforts at conducting proper controlled listening tests, which I support 100%. He should be applauded for attempting to do what many other review sites don't have the political courage, time/money or skill to do].So where are the stereo ratings test results of the ML vs to 362 so we can confirm this asserted correlation? This may very well be the case, but I don't see the data.
I don't need to test every product in mono and stereo to confirm what has already been well established. In a similar fashion, we don't routinely test other known listening test nuisance variables (e.g. blind vs sighted, level matched versus non level matched) just to satisfy the doubts of misguided audiophiles who believe they are immune to sighted and loudness-related biases in listening evaluations. With that said, you may see the mono versus stereo results in due course. Right now, it is not my top priority.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
it would have been interesting to see if they had less noise in the data if the tests were conducted in mono to confirm what we know from 25 years of running these tests.
Perhaps. But Gene seems primarily focused on "end user" data. That would be stereo, real room.
Personally, I would expect minor differences, but then again, there is a generic "sameness" to the total soundfield generated by these dome over cone monopole boxes...to my ears...rather than these dramatic difference, epic odes often written by users and the Audiophile press.
I don't need to test every product in mono and stereo to confirm what has already been well established. In a similar fashion, we don't routinely test other known listening test nuisance variable (e.g. blind vs sighted, level matched versus non level matched) just to satisfy the doubts of misguided audiophiles who might think they are immune to sighted and loudness-related biases in listening evaluations. With that said, you may see the mono versus stereo results in due course. Right now, it is not my the top priority.
Fair enough and I thank you for your discussion of the subject. Obviously, you are well positioned to discuss it.
Any thoughts on the screen cloth? As I asked Gene on AH, do you measure the total sound power loss due to the screen material?
Suggestions as to where I might acquire suitable material? Is DSP correction a consideration?
cheers,
AJ
I posted a new blog posting (http://seanolive.blogspot.com/2010/06/some-new-evidence-that-generation-y.html) that summarizes some recent experimental evidence where I tested a group of teenagers' preferences in loudspeakers and MP3 versus CD music formats. This is just the beginning of a more thorough study, so the results are very preliminary. Still I thought it would be interesting to get some feedback.
I could find no evidence that these high school students preferred the "sizzling sounds of MP3" over higher quality lossless formats, as reported by Jonathan Berger. I also found they preferred the most accurate, neutral loudspeakers when given the opportunity to hear and compare them with something less accurate and neutral.
These results are not too surprising to me, but the media seems to have been reporting a different story over the past year.
Cheers
Sean
Audio Musings (http://seanolive.blogspot.com)
My article was cited in a blog at Gramophone Magazine. (http://www.gramophone.co.uk/forum/general-discussion/some-new-evidence-that-generation-y-may-prefer-accurate-sound-reproduction). Interesting to see how classical music lovers respond to this or not. Of course most classical music lovers are past their prime listening years ;
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
What audio "jewelry" are you refering to? The diamond stylus of my cartridge is an actual functional part, not jewelry. Can't think of anything else you might be refering to. Do I also need to cite the multiple logical fallacies contained in your so called "must" conclusion?
From your equipment list on AA:
Speaker Cables/Interconnects: Audioquest Clear speaker cable, MIT Shotgun interconnects
Other (Power Conditioner, Racks etc.): Equitech 1.5Q line conditioner and Bybee pro filter
Tweaks: Aurios Pros under Subwoofer, Aurios 1.2 under 3/4 slab of acrylic under preamp, Aurios 1.0 under 2 one inch slabs of acrulic sandwiching seven home made discs of silicone elastomer with a shore hardness of 9, Tip Toes under power amp , VPI bricks over transformers on Martin Logans
I seriously doubt that any of this would test positive for audible differences in a DBT.
I don't want to hazard a guess in public as to how many $1,000s are involved.
To me this is the clear black-and-white part of the story of your collection of audio jewelry.
The following is a little more about shades of gray:
I actually see no components whatsover in your system that is anything like the minimum that would be reasonably required for the best possible sound.
IME much of it if not all of it would very likely found in a series of DBTs to be vastly infererior to far less expesnive equipment in terms of sound quality as judged by the criteria of sonic accuracy.
Of course, the single-minded devotion to an obsolete, inherently sonically flawed medium speaks for itself. You could be the most tin-eared person I've ever heard of. Even Fremer listens to an occasional digital recording, apparently on his primary system. You can't.
What audio "jewelry" are you refering to? The diamond stylus of my cartridge is an actual functional part, not jewelry. Can't think of anything else you might be refering to. Do I also need to cite the multiple logical fallacies contained in your so called "must" conclusion?
From your equipment list on AA:
Speaker Cables/Interconnects: Audioquest Clear speaker cable, MIT Shotgun interconnects
Other (Power Conditioner, Racks etc.): Equitech 1.5Q line conditioner and Bybee pro filter
Tweaks: Aurios Pros under Subwoofer, Aurios 1.2 under 3/4 slab of acrylic under preamp, Aurios 1.0 under 2 one inch slabs of acrulic sandwiching seven home made discs of silicone elastomer with a shore hardness of 9, Tip Toes under power amp , VPI bricks over transformers on Martin Logans
I seriously doubt that any of this would test positive for audible differences in a DBT.
I don't want to hazard a guess in public as to how many $1,000s are involved.
To me this is the clear black-and-white part of the story of your collection of audio jewelry.
The following is a little more about shades of gray:
I actually see no components whatsover in your system that is anything like the minimum that would be reasonably required for the best possible sound.
IME much of it if not all of it would very likely found in a series of DBTs to be vastly infererior to far less expesnive equipment in terms of sound quality as judged by the criteria of sonic accuracy.
Of course, the single-minded devotion to an obsolete, inherently sonically flawed medium speaks for itself. You could be the most tin-eared person I've ever heard of. Even Fremer listens to an occasional digital recording, apparently on his primary system. You can't.
Way to stay on topic Arny. Maybe you have a few more tips on comedy for me as well? Arny you need to get over this personal vendetta. You are just embarrassing yourself. But hey, if you want to fantasize about how bad my system sounds in order to feel better about yours that's fine. I can't say that I enjoy my system any less as a result.
Way to stay on topic Arny. Maybe you have a few more tips on comedy for me as well? Arny you need to get over this personal vendetta. You are just embarrassing yourself. But hey, if you want to fantasize about how bad my system sounds in order to feel better about yours that's fine. I can't say that I enjoy my system any less as a result.
Name calling won't change the fact that he's caught you.
This is the Hydrogen Audio site. The folks here are nowhere near as gullible as you seem to think they are.
Way to stay on topic Arny.
You asked the question, and I answered it. The question was OT. How can the answer to it be anything but OT?
Maybe you have a few more tips on comedy for me as well?
You lecture me about OT posts and then you post about comedy?
Arny you need to get over this personal vendetta.
There's nothing personal about it. My comments were about a list of equipment that you posted. If you think that I'm the only person on HA that perceives that equipment list like I do, then you really need to look around and listen to the other people who post here. Most people who post here realize that HA is part of the cure for equipment lists like that.
You are just embarrassing yourself.
That equipment list would embarass a great many people if it was what their prime system was. Think I'm wrong? Look at the equipment lists that most HA members have posted.
But hey, if you want to fantasize about how bad my system sounds in order to feel better about yours that's fine.
Scott you don't understand me at all! If I tried to post an equipment list, I wouldn't know where to start. I have so many systems that I control 100% ranging from a Microtrack and a pair of HD 280 headphones to an auditorium that seats up to 600 people and serves up to 45 performers concurrently.
I can't say that I enjoy my system any less as a result.
That's the most important thing Scott, that you enjoy it. I mean that quite seriously. Have a good day!
The Bybee filter was something I tried on a whim since i was able to get one for very little money compared to it's retail price. Can't say that it made any difference whatsoever. I sold it for a little more than my purchase price. The VPI bricks (two of them) were given to me. I'm not sure if they make that much of a difference or not but they sit on my Vandersteen subwooofer. Without them my two Emmy's that also sit on the subwoofer tend to ring when the bass gets loud. With the VPI bricks the on top of the sub the sub's top seems to be damped enough to stop the ringing. I don't like hearing the Emmys ring so that does make an audible improvement. I suppose the Emmys could be seen as audio jewelry since they sit on the sub, and are gold plated and very shiny??? I never thought of them as an actual part of the system though.
Interesting. You may have indeed affected the soundfield with the bricks in that fashion. Clearly with the Bybee, your system either lacked the resolution/resolvingness and/or you lacked the "hearing" ability (or self training) to perceive them. Which raises alarming Audiophile street cred issues, both system and personal in nature.
But I agree with you, we digress. I'll try to get back on topic.
My article was cited in a blog at Gramophone Magazine. (http://www.gramophone.co.uk/forum/general-discussion/some-new-evidence-that-generation-y-may-prefer-accurate-sound-reproduction). Interesting to see how classical music lovers respond to this or not. Of course most classical music lovers are past their prime listening years ;
Hmmm. Maybe. As most likely are (fans of) Big Band Jazz and other acoustic, non-amplified ensembles. But what better alternative is there for testing loudspeaker performance fidelity?
Any thoughts on the screen material for blinding?
cheers,
AJ
In my own limited experience however, it is the speaker/room interaction that causes me the most problems. To my ears, different rooms (+ speaker placements in those rooms) sound more different than different speakers...
I completely agree with that. However, I don't see this fact as working against mono comparisons. Instead I count the variables this creates and find that listening in mono simplifies the comparison. Try it, you may like it! I do.
Well, kind of, but it doesn't tell you how much the speaker's sound will fall apart as its pushed towards the corners of my living room.
Pushing a single speaker towards the corner of your room, any room will tell you volumes about how the sound of that speaker is affected.
Doing it with two speakers just adds more variables. Are the corners the same? Are the speakers positioned identically with respect to the corner?
The answer to the question "how does a corner affect the sound of this speaker" is first and foremost an abstraction.
The answer to the question "how do these corners in my living room affect these speakers" can be thought of being composed of the answers to a goodly number of questions, only one of which is "how does a corner affect the sound of this speaker".
If you're going to chose speakers based on how they sound when pushed into corners, then the most important question is how the various speakers sound when just one of them is pushed into just one corner. Most of the other questions have the same answers no matter which speakers you test.
Note that many of the sound quality changes to speakers when they are pushed into a corner are actually the same no matter which speaker you use. That's because they happen at low frequencies where virtually all home speakers are omnidirectional. All 5 1/2 inch direct radiators in typical boxes have basically the same radiation patterns at 50-100 Hz and probably running higher. Their bandpasses will vary, but this is all pretty simple and predictable.
[note: I don;t mean to dismiss Gene's well-intentioned efforts at conducting proper controlled listening tests, which I support 100%. He should be applauded for attempting to do what many other review sites don't have the political courage, time/money or skill to do].
Audioholics' credo is "Let our rigorous testing and reviews be your guidelines to A/V equipment – not marketing slogans ". Unfortunately their commitment to rigorous testing falls somewhat short of complete. (Pointing this out to them one time too many might get you 'shadowbanned', as I was for awhile)
I rank them somewhat above Stereophile in their science-mindedness, but like that magazine, Audioholics practices an aggravating combination of detailed objective measurement and poorly-controlled subjective evaluation.
Way to stay on topic Arny. Maybe you have a few more tips on comedy for me as well? Arny you need to get over this personal vendetta. You are just embarrassing yourself. But hey, if you want to fantasize about how bad my system sounds in order to feel better about yours that's fine. I can't say that I enjoy my system any less as a result.
Name calling won't change the fact that he's caught you.
This is the Hydrogen Audio site. The folks here are nowhere near as gullible as you seem to think they are.
Caught me what? How gullible did your tarot cards tell you I believe the folks here to be? Or what form of mind reading did you use to determine that? What was the title of this thread? "Let's try to catch Scott at something or another??"
Arny caught me using isolation for my turntable and protecting my system with a power supply/purifier that would literally save my irreplaceble airbearings on my tutrntable, flywheel and pickup arm in the case of a power outage and that proves that my experience in regards to whether or not I was duly made aware of the sweet spot before buying electrostatic speaker should be dismissed as unreliable experience? Really? Do I really need to explain just how dumb and desperate that is?
Arny really needs to get over this perosnal demon he has with me and move on. it really is just a source of embarrassment for him. You speak of my beliefs about the gullibility of the members of HA. I don't think the folks on HA are gullible enough to think that this isn't a personal issue for Arny or that Arny really has a point about the validity of my experiences with the sweet spot of electrostaic speakers as a long time owner of electrostatic speakers. Are you that gullible as to believe this isn't something personal? Do you really think this stuff belongs in this thread? Didn't the modrators already remove some of Arny's other comically inept off topic attempts at ad hominem against me from this thread, really wierd stuff like lecturing me on the structure of comedy and blasting me for "libeling" the unnamed writers of manuals of formula one racing cars? Hey, for me it's just something to poke fun at but there does come a point where it just gets in the way of the thread. If Arny wants to keep lobbing them up there I keep knocking them out of the park. But the moderators will have to keep cleaning up the mess. I am sure that is getting old for them. Arny really needs to let it go.
Note that many of the sound quality changes to speakers when they are pushed into a corner are actually the same no matter which speaker you use.
Absolutely false, given the context that the (speaker) discussion with Sean is a dipole (ML) vs a monopole (362). About the only thing similar will be their LF corner loading.
That's because they happen at low frequencies where virtually all home speakers are omnidirectional. All 5 1/2 inch direct radiators in typical boxes have basically the same radiation patterns at 50-100 Hz and probably running higher. Their bandpasses will vary, but this is all pretty simple and predictable.
That unfortunate fact is true, they are both monopoles (omnipolar) in that range..and the near corner placement will have significant impact, including gain at LF. So the "sounds thin" by looking at an anechoic snapshot ML, may sound anything but in the real room listener area, while the flat as a pancake anechoic 362 may sound bass heavy. Or not. Then we add a second channel, the soundfield becomes more complex...and real life stereo+.
cheers,
AJ
Arny caught me using isolation for my turntable and protecting my system with a power supply/purifier that would literally save my irreplaceble airbearings on my tutrntable, flywheel and pickup arm in the case of a power outage and that proves that my experience in regards to whether or not I was duly made aware of the sweet spot before buying electrostatic speaker should be dismissed as unreliable experience? Really? Do I really need to explain just how dumb and desperate that is?
You're minimizing what I caught you at, Scott. What you don't realize is that you just dug yourself deeper because you reiterated more stuff that makes no reliably detectable audible difference, like the flywheel.
Arny really needs to get over this perosnal demon he has with me and move on.
Asked and answered, but just to repeat it - its not personal, Scott. *Anybody* who proudly posts such an incredible list of cosmetic and dysfunctional audio gear would get about the same response, were they also pontificating like you do.
What's embarassing is someone who plays the class card as often as you do while pretending to be interested in just discussing audio. That may play at The Hollywood Spa or the Midtowne, but its getting pretty old here.
Note that many of the sound quality changes to speakers when they are pushed into a corner are actually the same no matter which speaker you use.
Absolutely false, given the context that the (speaker) discussion with Sean is a dipole (ML) vs a monopole (362). About the only thing similar will be their LF corner loading.
Note that I wasn't discussing with Sean, but rather I was discussing with 2Bdecided, and the context was whether doing this experiment with 2 speakers in 2 corners was the better idea than doing it with 1 speaker in 1 corner.
Dipoles are such a tiny, tiny fraction of the market that I tend to totally forget that they exist. We could all agree to forget them and not much would change in the world of audio. They just make things more complex, and for what?
I suspect that if you push a true dipole into a corner the results are a lot like those from a monopole with similar directivity until the front lobe starts engaging the walls. The strong effect of dipoles at low frequencies is that the lobes become so large that they overlap and cancel, and that has its strongest effect on the overall bass bandpass.
Since I used the word "many" I was allowing that there were exceptions. This makes your judgement of "Absolutely False" absolutely false. ;-)
Note that I wasn't discussing with Sean, but rather I was discussing with 2Bdecided, and the context was whether doing this experiment with 2 speakers in 2 corners was the better idea than doing it with 1 speaker in 1 corner.
Well apparently you need to read the article (http://seanolive.blogspot.com/2010/06/some-new-evidence-that-generation-y.html) that the thread is about. The one speaker in the experiment isn't anywhere near a corner. It's in the middle of the front wall. I wasn't suggesting that they listen to one, in one corner either.
Dipoles are such a tiny, tiny fraction of the market that I tend to totally forget that they exist. We could all agree to forget them and not much would change in the world of audio. They just make things more complex, and for what?
For your requirements of fidelity, this is true. But irrelevant, since Sean chose to use one (and single it out) in his experiment. So you could argue with him as to why they are in his testing, given your requirements.
I suspect that if you push a true dipole into a corner the results are a lot like those from a monopole with similar directivity until the front lobe starts engaging the walls. The strong effect of dipoles at low frequencies is that the lobes become so large that they overlap and cancel, and that has its strongest effect on the overall bass bandpass.
I suspect you need to pick up some basic acoustics texts or go the Linkwitz's site. It's vast..and free.
A true dipole will have completely different directivity than a monopole. At any frequency it will be true dipole up to around 1/3 wavelength of the source size, then exhibit "dipolar" behavior, where the nulls are not fully formed, then eventually more directive as wavelength decreases. But their will still be radiated sound in both directions.
Bottom line...the interaction of the ML and 362, placed the same distance near corners, as they would be in real life, will be completely different. More so in stereo.
Gotta run.
cheers,
AJ
Note that I wasn't discussing with Sean, but rather I was discussing with 2Bdecided, and the context was whether doing this experiment with 2 speakers in 2 corners was the better idea than doing it with 1 speaker in 1 corner.
Well apparently you need to read the article (http://seanolive.blogspot.com/2010/06/some-new-evidence-that-generation-y.html) that the thread is about.
Childish rhetoric like that gets you noplace with me.
Note that many of the sound quality changes to speakers when they are pushed into a corner are actually the same no matter which speaker you use.
That unfortunate fact is true, they are both monopoles (omnipolar) in that range..and the near corner placement will have significant impact, including gain at LF. So the "sounds thin" by looking at an anechoic snapshot ML, may sound anything but in the real room listener area, while the flat as a pancake anechoic 362 may sound bass heavy. Or not. Then we add a second channel, the soundfield becomes more complex...and real life stereo+.
But the problem AJ, is that ML doesn't recommend you put their speaker in the corner. They talk about putting it 3-4 feet from the front wall and ideally they say there should be no side walls near the speaker. It's right on the owners manual on page 10.
Side walls
"A good rule of thumb is to have the side walls as far away from the speaker sides as possible ". .. "An ideal side wall, however, is no side wall at all."
Do you work at ML or have inside information that indicates the owners' manual is wrong? If so, you should contact them and let them know.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
I bet the silent majority of readers with no particular axe to grind are thinking that testing stereo speakers in mono is a strange thing to do.
I've read the justifications. I don't buy them. I can see how on-axis mono listening can give more repeatable results where differences are more easily audible - but maybe that's another way of saying that normal stereo listening depends far more on the interaction between the speakers and the room.
Like it or not, most of us are stuck with interaction between our speakers and our listening rooms. Obviously it would complicate things, but I think many people are interested in which speakers sound best in the locations where the speakers look+fit best in their living room. Other people are interested in which speakers sound best given the opportunity to place the speakers at the optimum location for that pair of speakers. Other people are interested in creating the best listening room, placing the speakers optimally, and getting the best sound.
These are harder questions to answer, and I don't want to make the best the enemy of the good, but the concerns are real.
I'm sure Sean can envisage a Mark II speaker switcher room with a speaker switcher in each corner to deliver stereo tests. How much space have you got Sean?
Cheers,
David.
Our MLL speaker mover already accommodates stereo comparisons of up to 4 sets of loudspeakers, and we have conducted both stereo and mono tests for some 12 years now.
No one I know of -- including Martin Logan (read their owner's manual) recommends putting loudspeakers in the corner unless they are subwoofers or the speakers were designed to be put there (e,g Klipschhorns). So, it would be a waste of money to build a speaker mover specifically to test corner-deigned speakers since not many exist.
We are in the process of building some side walls that will come in closer to the speaker mover and create stronger, earlier lateral reflections that will contribute more weight to the listeners' overall impression of the loudspeaker.
This should even make our listening tests even more sensitive to loudspeakers with problems in their off-axis response. The ratings should increase forwell designed speakers that have well-maintined smooth off-axis responses should increase, and the ratings of loudspeakers with poor off-axis responses should decrease.
Given what we know about the on and off-axis performance of the Primus 362 and the ML, how do you think these two speakers will fare in this newly designed room?
See slide 28 and compare Primus versus Loudspeaker C measurements (http://docs.google.com/fileview?id=0B97zTRsdcJTfZmM1ZjhlN2MtMTQwZC00MzVmLWE1OTQtMWQzN2IzNGNiMjc0&hl=en).
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Arny really needs to get over this perosnal demon he has with me and move on. it really is just a source of embarrassment for him.
I think we all (other than you of course) know who should be embarrassed here, and that's not Arnie.
Given what we know about the on and off-axis performance of the Primus 362 and the ML, how do you think these two speakers will fare in this newly designed room?
See slide 28 and compare Primus versus Loudspeaker C measurements (http://docs.google.com/fileview?id=0B97zTRsdcJTfZmM1ZjhlN2MtMTQwZC00MzVmLWE1OTQtMWQzN2IzNGNiMjc0&hl=en).
I enjoyed the presentation, but I thought the generalization to an entire generation a bit strong, and too strongly worded. I think it is valid and interesting evidence, but without confirmation the apparent suggestion that the evidence is conclusive seems too strong based on one study and the fairly small sample size. Also of course, given that the students came from a single school it might be a biased sample and wasn't really randomized as far as I can see.
But still very interesting and thanks for presenting it to us.
]
A true dipole will have completely different directivity than a monopole. At any frequency it will be true dipole up to around 1/3 wavelength of the source size, then exhibit "dipolar" behavior, where the nulls are not fully formed, then eventually more directive as wavelength decreases. But their will still be radiated sound in both directions.
Bottom line...the interaction of the ML and 362, placed the same distance near corners, as they would be in real life, will be completely different. More so in stereo.
Gotta run.
The ML we tested is not a true dipole but a hybrid with subwoofer that is omnidirectional up to ~ 150 Hz. So below 150 Hz, it was no more disadvantaged as the other loudspeakers in the test, which were rated much higher.
The Orion speaker that you own (based on the drawings you sent me) is also a dipole but it is NOT an electrostatic. It rather uses conventional dynamic drivers that have a better behaved on and off-axis frequency response than the typical electrostatic panels we've tested over the past 25 years. Since frequency response is the dominant factor in listener's preference of a loudspeaker, the Orion would likely be rated higher than the ML, in a controlled listening test. It would be an interesting test to conduct to see if that were true.
The only common link between your speaker and the ML is that one is a dipole and the other is a hybrid dipole. Your speaker is a full range dipole (I think) and its bass would be influenced differently by its distance to a wall/corner than the ML, which is an omnipole below 150 Hz.
There have been a number of experiments recently done by David Clark on the $8k Orion where it was compared against a $300 powered 2-way semi-professional well designed direct radiator type speaker. From what I've heard from some of the participants, there was nothing about the dipole that distinguished itself in terms of the manufacturer's claims that the speaker can present an extraordinary stereo sound stage in any room/position compared to a well-designed direct radiator loudspeaker. Under scientific scrutiny, the manufacturer's claims could not be substantiated. I look forward to you substantiating your claims with some data.
I had a conversation with Floyd Toole yesterday about the Quad Electrostatic dipole listening tests done at the NRC. One interesting fact, that is in his book, is that the Quad produced some of the widest variance in ratings that was related to hearing loss. Listeners with normal hearing didn't like it much. But listeners with hearing loss tended to like it more. My theory is that its increased directivity probably helped increase intelligibility of speech and vocals, which could be a benefit for people with hearing loss.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
There have been a number of experiments recently done by David Clark on the $8k Orion where it was compared against a $300 powered 2-way semi-professional well designed direct radiator type speaker. From what I've heard from some of the participants, there was nothing about the dipole that distinguished itself in terms of the manufacturer's claims that the speaker can present an extraordinary stereo sound stage in any room/position compared to a well-designed direct radiator loudspeaker. Under scientific scrutiny, the manufacturer's claims could not be substantiated. I look forward to you substantiating your claims with some data.
Audio Musings (http://seanolive.blogspot.com)
If anybody wants more information on these tests here is a link http://home.provide.net/~djcarlst/smwtms.htm (http://home.provide.net/~djcarlst/smwtms.htm)
In particular, here is a link (copied from the page above) to .PDF that talks about the test. http://home.provide.net/~djcarlst/SLReport10.05.pdf (http://home.provide.net/~djcarlst/SLReport10.05.pdf)
The "designed direct radiator type speaker" Dr. Olive is talking about seems to be the Behringer Truth B2031A
Arnold, I'm not usually one to jump in and defend analog scott here, but come on. Discussion of his system and components that you yourself admit likely have no discernible effect on audio is the very definition of off-topic.
Sure, he may like his snake oil. But he's not pushing it, nor is he breaking any Terms of Service.
Please just let it be? Though by-and-large the users here tend to have pretty convergent views on audio equipment, let's not ostracize those who adhere to our philosophies yet hold variant positions on what hardware they prefer.
There have been a number of experiments recently done by David Clark on the $8k Orion where it was compared against a $300 powered 2-way semi-professional well designed direct radiator type speaker. From what I've heard from some of the participants, there was nothing about the dipole that distinguished itself in terms of the manufacturer's claims that the speaker can present an extraordinary stereo sound stage in any room/position compared to a well-designed direct radiator loudspeaker. Under scientific scrutiny, the manufacturer's claims could not be substantiated.
I particpated in these listening tests and can confirm the above statement.
I look forward to you substantiating your claims with some data.
And of course with the same or better bias controls.
We were flankly quite surprised by the similarity of the sound quality of the two systems given the vast differences in their design and performance.
There have been a number of experiments recently done by David Clark on the $8k Orion where it was compared against a $300 powered 2-way semi-professional well designed direct radiator type speaker. From what I've heard from some of the participants, there was nothing about the dipole that distinguished itself in terms of the manufacturer's claims that the speaker can present an extraordinary stereo sound stage in any room/position compared to a well-designed direct radiator loudspeaker. Under scientific scrutiny, the manufacturer's claims could not be substantiated. I look forward to you substantiating your claims with some data.
Audio Musings (http://seanolive.blogspot.com)
If anybody wants more information on these tests here is a link http://home.provide.net/~djcarlst/smwtms.htm (http://home.provide.net/~djcarlst/smwtms.htm)
In particular, here is a link (copied from the page above) to .PDF that talks about the test. http://home.provide.net/~djcarlst/SLReport10.05.pdf (http://home.provide.net/~djcarlst/SLReport10.05.pdf)
The "designed direct radiator type speaker" Dr. Olive is talking about seems to be the Behringer Truth B2031A
What is the 'reference' referred to in these tests?
]
A true dipole will have completely different directivity than a monopole. At any frequency it will be true dipole up to around 1/3 wavelength of the source size, then exhibit "dipolar" behavior, where the nulls are not fully formed, then eventually more directive as wavelength decreases. But their will still be radiated sound in both directions.
Bottom line...the interaction of the ML and 362, placed the same distance near corners, as they would be in real life, will be completely different. More so in stereo.
Gotta run.
The ML we tested is not a true dipole but a hybrid with subwoofer that is omnidirectional up to ~ 150 Hz. So below 150 Hz, it was no more disadvantaged as the other loudspeakers in the test, which were rated much higher.
The Orion speaker that you own (based on the drawings you sent me) is also a dipole but it is NOT an electrostatic. It rather uses conventional dynamic drivers that have a better behaved on and off-axis frequency response than the typical electrostatic panels we've tested over the past 25 years. Since frequency response is the dominant factor in listener's preference of a loudspeaker, the Orion would likely be rated higher than the ML, in a controlled listening test. It would be an interesting test to conduct to see if that were true.
The only common link between your speaker and the ML is that one is a dipole and the other is a hybrid dipole. Your speaker is a full range dipole (I think) and its bass would be influenced differently by its distance to a wall/corner than the ML, which is an omnipole below 150 Hz.
There have been a number of experiments recently done by David Clark on the $8k Orion where it was compared against a $300 powered 2-way semi-professional well designed direct radiator type speaker. From what I've heard from some of the participants, there was nothing about the dipole that distinguished itself in terms of the manufacturer's claims that the speaker can present an extraordinary stereo sound stage in any room/position compared to a well-designed direct radiator loudspeaker. Under scientific scrutiny, the manufacturer's claims could not be substantiated. I look forward to you substantiating your claims with some data.
I had a conversation with Floyd Toole yesterday about the Quad Electrostatic dipole listening tests done at the NRC. One interesting fact, that is in his book, is that the Quad produced some of the widest variance in ratings that was related to hearing loss. Listeners with normal hearing didn't like it much. But listeners with hearing loss tended to like it more. My theory is that its increased directivity probably helped increase intelligibility of speech and vocals, which could be a benefit for people with hearing loss.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
FYI - Just to clarify my point made above: The data showing the interaction between listener hearing loss and the ratings of the Quad ESL-63 are shown in Figure 17.10 (page 356) of Floyd's book "Sound Reproduction". The Quad ESL-63 is speaker X and shows a lot of variance in ratings among the hearing impaired (the high standard deviation group), some really liking it, others not liking it. The listeners with normal hearing (listeners with low standard deviations in their ratings) clearly did not like it.
What is interesting is that the highest rated loudspeaker (PSB Passive II) in the test was also technically the most accurate loudspeaker, and it received the highest ratings among both normal and hearing impaired listeners. Also significant is the high degree of consensus in the high ratings among the listeners in both normal and hearing impaired groups.
This evidence suggests that an accurate loudspeaker is the best design target if you want to please the most listeners regardless of whether their hearing is normal or impaired. That is reassuring since it would otherwise be a challenge to design and market a loudspeaker aimed specifically for the hearing impaired audiophile. No audiophile or reviewer would be willing to audition it
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
But the problem AJ, is that ML doesn't recommend you put their speaker in the corner. They talk about putting it 3-4 feet from the front wall and ideally they say there should be no side walls near the speaker. It's right on the owners manual on page 10.
Side walls
"A good rule of thumb is to have the side walls as far away from the speaker sides as possible ". .. "An ideal side wall, however, is no side wall at all."
Do you work at ML or have inside information that indicates the owners' manual is wrong? If so, you should contact them and let them know.
You can continue to spin this, cherry pick and play word games with in/near/around corners all you want Sean. The reality of the situation is that the ML's and the 362's, will both be listened to, in stereo >channels (just like your codec listening), in real rooms, many with corners, where the speakers are going to be placed in the area of the corners, the exact distances varying. For the ML, it would be at least 3' out from the front wall. Distance from sidewall would clearly depend of the room dimensions, or whether there was a sidewall.
Whether or not you accept this reality is irrelevant. Do a survey of owners if you need to.
typical boxes have basically the same radiation patterns at 50-100 Hz and probably running higher.
That unfortunate fact is true, they [**ML and 362**] are both monopoles (omnipolar) in that range....
The ML we tested is not a true dipole but a hybrid with subwoofer that is omnidirectional up to ~ 150 Hz. So below 150 Hz, it was no more disadvantaged as the other loudspeakers in the test
Who stated differently???????
Do you have a listening area FR of a (area of) corner placed 362 or ML, rather than just the anechoic FR?
The Orion speaker that you own (based on the drawings you sent me) is also a dipole but it is NOT an electrostatic.
I cloned the Orion platform circa 2004 (after hearing them....in stereo, in a living room....with corners). I changed the design substantially (waveguide loaded ring radiator, larger woofers, etc.). It would be a stretch (and a disservice to SL) to call them proper Orions. Plus, those have been shelved for years.
And maybe it's old age, but I don't recall sending you drawings. What drawings??
The only common link between your speaker and the ML is that one is a dipole and the other is a hybrid dipole. Your speaker is a full range dipole (I think) and its bass would be influenced differently by its distance to a wall/corner than the ML, which is an omnipole below 150 Hz.
As I have repeatedly stated, it not the fact that you are testing the ML (model X) per se, it's the
method used and the conclusions implied that I take issue with. As I have also repeatedly said, the mono preference speaker may indeed remain the stereo preference speaker. I just disagreed with how you reached that projection.
There have been a number of experiments recently done by David Clark on the $8k Orion where it was compared against a $300 powered 2-way semi-professional well designed direct radiator type speaker....
I have briefly skimmed it...and can't make heads or tails of what is going on there. I'll look closer when I get a chance....but much more importantly:
I have asked you more than once, so I will take no answer as an answer as well -
Do you have data for the screen cloth used in your tests (I see none for the Clark one either)?
Do you have on and off axis FR data for the losses. Do you have data for the total power loss at the listening area with the screen in place?
I would be especially curious to see the data of the 362 and ML, as measured from the listening area, both with and without the screen. And you did say the corner behind the speakers (and the screen) are typically reflective of living rooms correct? TIA
Really looking forward to your response on the screen.
cheers,
AJ
Given what we know about the on and off-axis performance of the Primus 362 and the ML, how do you think these two speakers will fare in this newly designed room?
See slide 28 and compare Primus versus Loudspeaker C measurements (http://docs.google.com/fileview?id=0B97zTRsdcJTfZmM1ZjhlN2MtMTQwZC00MzVmLWE1OTQtMWQzN2IzNGNiMjc0&hl=en).
I enjoyed the presentation, but I thought the generalization to an entire generation a bit strong, and too strongly worded. I think it is valid and interesting evidence, but without confirmation the apparent suggestion that the evidence is conclusive seems too strong based on one study and the fairly small sample size. Also of course, given that the students came from a single school it might be a biased sample and wasn't really randomized as far as I can see.
But still very interesting and thanks for presenting it to us.
Thanks Ed. Good points. Of course, you are right that the sample size (18) is too small and not randomized to generalize the results to an entire population. The students came on a field trip to Harman and agree to participate in two 30 minute tests, so I had little to do with their selection per se.
In the conclusions of the blog posting I qualified the findings to just the tested sample:
"While this study is still in its early phase, these preliminary results suggest that the teenagers could reliably discriminate among different degradations in sound quality in music reproduction. When given the opportunity to hear and compare different qualities of sound reproduction, the high school students preferred the higher quality, more accurate reproduction over the lower quality choices."
We will continue to test more young people using a proper recruited samples of young people to see if the results hold up, and the extent to which they can be more generalized to a population. I don't know how many students Berger tested or if his sample was properly randomized,etc., but we want to be able to better understand the important factors behind what makes a consumer from different market segments prefer more or less accurate sound reproduction.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Do you have data for the screen cloth used in your tests (I see none for the Clark one either)?
Do you have on and off axis FR data for the losses. Do you have data for the total power loss at the listening area with the screen in place?
I would be especially curious to see the data of the 362 and ML, as measured from the listening area, both with and without the screen. And you did say the corner behind the speakers (and the screen) are typically reflective of living rooms correct? TIA
Really looking forward to your response on the screen.
Based on many samples, we chose a very open weave Lacoste knit polyester cloth that we measured both anechoically and in the listening room with the cloth at different distances from the speaker (the angle of sound incidence and distance of the cloth from the speaker are factors on how much attentuation you see). That was quite awhile (12 years?) ago, so I don't have that information at my finger tips. I will see if I can find it.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Sure, he may like his snake oil. But he's not pushing it, nor is he breaking any Terms of Service.
You must have missed his TOS 8 infraction in post 58, this thread.
-
Do you have data for the screen cloth used in your tests (I see none for the Clark one either)?
The Orions that Clark tested were obtained from Linkwitz just a few months back, so you can obtain grille cloth data from Linkwitz, The Behringer B 2031A speakers have no grille cloth.
If you're talking about the scrim used in the test, it was the same for both speakers, so the tiny variations it caused are the same for both speakers and don't matter.
There have been a number of experiments recently done by David Clark on the $8k Orion where it was compared against a $300 powered 2-way semi-professional well designed direct radiator type speaker. From what I've heard from some of the participants, there was nothing about the dipole that distinguished itself in terms of the manufacturer's claims that the speaker can present an extraordinary stereo sound stage in any room/position compared to a well-designed direct radiator loudspeaker. Under scientific scrutiny, the manufacturer's claims could not be substantiated. I look forward to you substantiating your claims with some data.
Audio Musings (http://seanolive.blogspot.com)
If anybody wants more information on these tests here is a link http://home.provide.net/~djcarlst/smwtms.htm (http://home.provide.net/~djcarlst/smwtms.htm)
In particular, here is a link (copied from the page above) to .PDF that talks about the test. http://home.provide.net/~djcarlst/SLReport10.05.pdf (http://home.provide.net/~djcarlst/SLReport10.05.pdf)
The "designed direct radiator type speaker" Dr. Olive is talking about seems to be the Behringer Truth B2031A
What is the 'reference' referred to in these tests?
B2031A speakers were the reference. Page 14 of the PDF actually shows 2 copies of the reference type speaker. The one that was used for listening tests is the one that is suspended by a chain in close proximity and orientation to the Orion below it.
If you're talking about the scrim used in the test
Yes, I'm referring to the "Scrim" (now that I've Googled it).
it was the same for both speakers, so the tiny variations it caused are the same for both speakers and don't matter.
Fantastic. So you obviously have total sound power measurements at the listening area for both speakers, with an without scrim and found "tiny variations" that "don't matter". Can we see them? Link?
Thanks.
cheers,
AJ
Of course, you are right that the sample size (18) is too small and not randomized to generalize the results to an entire population. The students came on a field trip to Harman and agree to participate in two 30 minute tests, so I had little to do with their selection per se.
In the conclusions of the blog posting I qualified the findings to just the tested sample:
Yes you did and I should have mentioned that. I intended to, as I recall, but forgot. I was only talking about the wording of the last "slide" of the link you posted. Of course it would be perfectly normal to see that kind of phrasing if it were meant as add copy.
we measured both anechoically and in the listening room with the cloth at different distances from the speaker (the angle of sound incidence and distance of the cloth from the speaker are factors on how much attentuation you see).
Yep. Hence my interest. Especially when we are no longer comparing polar responses/reverberation of monopole vs monopole.
Based on many samples, we chose a very open weave Lacoste knit polyester cloth.
That was quite awhile (12 years?) ago, so I don't have that information at my finger tips. I will see if I can find it.
Thank you very much.
You must have missed his TOS 8 infraction in post 58, this thread.
Post 58 is by Sean Olive. You're really trying our patience, Arnold.
Post 58 is by Sean Olive. You're really trying our patience, Arnold.
I believe he is referencing the comment about electrostatic break-in in post 53 (http://www.hydrogenaudio.org/forums/index.php?showtopic=81708&view=findpost&p=711070)
That being said, I think the attacks on analog scott are off-topic and not justified. He and I were having a disagreement about the information available to and given by electrostatic speaker buyers and dealers and somehow people claimed he was wrong because he owns equipment people on HA consider to be snake-oil.
Such an attack is not justified and it doesn't address his comments at all. If analog scott is disgusted by the behavior his feeling is justified.
You must have missed his TOS 8 infraction in post 58, this thread.
Post 58 is by Sean Olive. You're really trying our patience, Arnold.
Sorry, I meant post 53.
If you're talking about the scrim used in the test
Yes, I'm referring to the "Scrim" (now that I've Googled it).
Ever seen a scrim in action? Maybe not, since their purpose in life is to not be seen by the audience even though they are being used. The point is that scrims can be unbelievably light and gauzy (and thus be almost totally acoustically transparent) and through some simple tricks with lighting do an amazing job of concealing objects behind them.
Clark used a scrim to conceal the speakers being tested. I'm not sure how many experimenters have done this. Anybody with real-world experience with the sound absorbtion of fabrics would dismiss concerns about the sound absorbtion of this scrim were they to actually see it.
it was the same for both speakers, so the tiny variations it caused are the same for both speakers and don't matter.
Fantastic. So you obviously have total sound power measurements at the listening area for both speakers, with an without scrim and found "tiny variations" that "don't matter". Can we see them? Link?
Clark did a lot of very sophisticated acoustical measurements to set up the tests. Check the cover page of the PDF that we have been referencing for his email address.
Ever seen a scrim in action?
No idea, since I first saw the word today.
Clark used a scrim to conceal the speakers being tested. I'm not sure how many experimenters have done this.
And just to be clear, the "scrim" is the hanging black thing in the picture (labeled "Opaque screen in place"), correct?
(http://i216.photobucket.com/albums/cc73/AJinFLA/screen.jpg)
Anybody with real-world experience with the sound absorbtion of fabrics would dismiss concerns about the sound absorbtion of this scrim were they to actually see it.
Well, I'm a simpleton, so I would still like to see some soundwave measurements, rather than speculate with my eyes.
Clark did a lot of very sophisticated acoustical measurements to set up the tests. Check the cover page of the PDF that we have been referencing for his email address.
Thanks, I will.
cheers,
AJ
Clark used a scrim to conceal the speakers being tested. I'm not sure how many experimenters have done this.
And just to be clear, the "scrim" is the hanging black thing in the picture (labeled "Opaque screen in place"), correct?
(http://i216.photobucket.com/albums/cc73/AJinFLA/screen.jpg)
Yes. If the light is moved to the back of the scrim, then you can easily see the items behind the screen. With the lighting arragement shown, the scrim seems to be solidly opaque. Since the goal here is maximum concealment, the scrim is black. In theatrical use, the scrim is used to reveal and conceal objects behind the scrim at will. The scrim is usually a light color or white. Images projected on the front the scrim can also be quite convincing. There's another variation of this trick using an angled sheet of glass.
Anybody with real-world experience with the sound absorbtion of fabrics would dismiss concerns about the sound absorbtion of this scrim were they to actually see it.
Well, I'm a simpleton, so I would still like to see some soundwave measurements, rather than speculate with my eyes.
You may have to speculate if you want to be totally skeptical from a distance, but if you're there it is pretty obvious that the scrim is acoustically transparent enough to not be a problem. First off, with normal lighting the actual thinness and transparency of the scrim is obvious. Secondly, if someone walks behind it while speaking, there is no audible change in their voice. Thirdly, recordings sound natural and are unchanged when you pull the scrim back. Fourthly, if you saw how the speakers were positioned, you'd see that there is no resaonable way that one speaker would have its sound affected by the scrim without the other speaker being similarly affected because of how the speakers are mounted (I've previously referenced the picture in the PDF that shows this.)
Post 58 is by Sean Olive. You're really trying our patience, Arnold.
I believe he is referencing the comment about electrostatic break-in in post 53 (http://www.hydrogenaudio.org/forums/index.php?showtopic=81708&view=findpost&p=711070)
That being said, I think the attacks on analog scott are off-topic and not justified.
There was discussion of Scott's equipment, which is not an attack on Scott personally.
The above are either straw man or excluded-middle arguments depending on how you take them.
OP: "mono test results track stereo results"
Me: "not always, here's two trivial counter examples"
You: "straw man"
HA search says you've claimed "straw man" 11 times now. You should build a haystack.
When faced with poor quality rhetoric, one may start building haystacks to pass the time while waiting for intelligent disucssion to resume.
You may have to speculate if you want to be totally skeptical from a distance, but if you're there it is pretty obvious that the scrim is acoustically transparent enough to not be a problem. First off, with normal lighting the actual thinness and transparency of the scrim is obvious.
Again Arnie, I don't have this all figured out, like an EE as yourself. I'm just a yahoo with a keyboard. What is obvious to you, raises questions for me. So I would much rather see an acoustic measurement, than hear about subjective anecdotes or see light behavior. I figure with HA being a scientific forum, one would seem more appropriate than the other. I know this is all simple stuff and things are cut and dry settled for yourself. But I'm not interested in how light passes through it. I'm interested in how sound waves, including reflected ones, at varying angles of incidence, are affected.
Secondly, if someone walks behind it while speaking, there is no audible change in their voice.
Limited bandwidth, monopolar source with not much spatial information.
Thirdly, recordings sound natural and are unchanged when you pull the scrim back.
Tone, timbre...or spatially, especially with the dipoles? What does a "natural"
recording sound like? What are you acclimated to hearing with reproduction?
Fourthly, if you saw how the speakers were positioned
I did. Which is why my first reaction was "Whats going on here?". Did they remain vertically "stacked" during listening? With the polar response of the B2031 inverted, including the non-coincident driver frequency dependent vertical lobing? Did the screen prevent them from being identified by soundstage height? Could they be identified by height cues, with or without the screen? Could you visually tell which was playing without the screen...and if not what
was the purpose of the screen? What was the EQ used for?...I could go on, but as you pointed out, David has an e-mail address.
Btw, I did see the question raised along the lines of "How much live acoustic music do you attend/have stored in memory"? I didn't see the answers by the listeners. But I digress...
....you'd see that there is no resaonable way that one speaker would have its sound affected by the scrim without the other speaker being similarly affected because of how the speakers are mounted (I've previously referenced the picture in the PDF that shows this.)
You mean because both speakers have a very similar frontal polar field impinging upon the scrim at an equal distance/angle relationship....but only one has a (near) identical rear polar field (the opposite of the other speaker), which plays a significant role in it's signature, reflect around and then impinge upon the scrim at varying angles??
Well, as I said before, it's all simple and figured out to someone like you Arnie. Me, I have to sludge slowly through the mud, one step at a time, with still no guarantee that I''ll figure out all this simple stuff. Stick with me on this.
Are you sure about that? Perhaps you can scratch Axiom off that list Audioholics Axiom Blind Testing (http://www.audioholics.com/news/editorials/axiom-blind-listening-test)
OK. That is news to me, since Ian did mono tests at NRC when he used those facilities. I'm actually quite happy to discover that competitors are now testing their speakers in stereo rather than mono because it means that their test results are not as sensitive as ours.
Both Floyd Toole and I recently had many conversations with Gene D. at Audioholics about why he should do mono loudspeaker comparisons. Unfortunately, he chose to ignore our recommendations in his last loudspeaker comparison shoot-out for the same arguments that you make. So what was the outcome of his stereo loudspeaker test results? I statistically analyzed the data and there were no statistically significance differences in listeners' preference ratings among any of the loudspeakers tested: it was just noise. Of course, several other important listening test nuisance variables were not controlled (e.g. seat and loudspeaker position, and a grille cloth blind screen with 1-2dB broadband effects) but it would have been interesting to see if they had less noise in the data if the tests were conducted in mono to confirm what we've know from 25 years of running these tests.
[note: I don;t mean to dismiss Gene's well-intentioned efforts at conducting proper controlled listening tests, which I support 100%. He should be applauded for attempting to do what many other review sites don't have the political courage, time/money or skill to do].
So where are the stereo ratings test results of the ML vs to 362 so we can confirm this asserted correlation? This may very well be the case, but I don't see the data.
I don't need to test every product in mono and stereo to confirm what has already been well established. In a similar fashion, we don't routinely test other known listening test nuisance variables (e.g. blind vs sighted, level matched versus non level matched) just to satisfy the doubts of misguided audiophiles who believe they are immune to sighted and loudness-related biases in listening evaluations. With that said, you may see the mono versus stereo results in due course. Right now, it is not my top priority.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Sean, I'm having trouble with a few of your characterizations here and perhaps the premises behind them. You claim that the results in mono are more sensitive and therefore better/more useful but it seems to me that the tests in stereo are inevitably going to be a truer test of actual in the field usage and while they may be less "sensitive" they are in fact more *accurate* and truer representation of listener preferences. You speak of "noise." What is the "noise" in the data? How are listener impressions more or less "noisy" in blind comparisons?
You may have to speculate if you want to be totally skeptical from a distance, but if you're there it is pretty obvious that the scrim is acoustically transparent enough to not be a problem. First off, with normal lighting the actual thinness and transparency of the scrim is obvious.
Again Arnie, I don't have this all figured out, like an EE as yourself. I'm just a yahoo with a keyboard. What is obvious to you, raises questions for me.
I have no problem with raising questions, but at some point the non-acceptance of reasonable answers becomes tedius and suggests a lack of respect.
So I would much rather see an acoustic measurement,
Acoustic measurements of the sound transmission of materials are a matter of the public record. Secondly, non-acceptance of reasonable answers suggests that if measurements show say a 0.001 dB loss at 1 megahertz, they will be abused as proof that the whole set of tests was fatally flawed.
than hear about subjective anecdotes or see light behavior.
Measurements are as easy to abuse as subjective observations.
I figure with HA being a scientific forum, one would seem more appropriate than the other.
The logical flaw here is that using measurements is no more or less scientific than subjective observations. It all depends. Measurements by themselves are meaningless. The only way that we can make any sense out of measruements is to correlate them with subjective observations.
I know this is all simple stuff and things are cut and dry settled for yourself.
If you really believe that, why not just accept my judgements? Why not do your own research on the acoustic properties of thin fabrics to see whether or not my statements are reasonable?
But I'm not interested in how light passes through it.
I suspect that you are in fact interested in how light passes through the scrim, because if the scrim passes light inappropriately we could still see which speaker was playing and the scrim would be useless. This discussion is really about the ratio between visual concealment and sonic transparency. Futhermore, there is a loose correlation between how light passes through common fabrics and how sound passes through them. Actually, we want the scrim to break the usual pattern by passing sound and obstructing light.
I'm interested in how sound waves, including reflected ones, at varying angles of incidence, are affected.
The fact that sound is direct or reflected generally has no effect on the acoustic transmisivity of fabrics. Thin cloth materials generally don't care about the incidence angles of sound when the angles are mostly near perpendicular which they were here.
But why should I bother answering your questions? You've treated my existing statements like they have zero credibility. Why do you bother to ask any questions at all if you are going to throw away the answers you don't like? Why beleive measurements when there is a possibility that they are fatally flawed? If you know nothing about audio, why would you believe a measurement is right when you disbelieve all subjective observations?
Secondly, if someone walks behind it while speaking, there is no audible change in their voice.
Limited bandwidth, monopolar source with not much spatial information.
Wrong. While the human voice can tolerate bandwidth reduction to a degree its bandwidth is as wide as many musical instruments. One word: sibilants. The voice comes from at least two different parts of the body - the mouth and the chest so it is not a monopole. The voice normally picks up considerable spatial content by being reflected by nearby objects.
The point is that there is an unending list of problems that can be hypothesized by someone who doesn't want to believe the outcome of a test. Being ignorant makes all things possible. Being knowlegable greatly reduces the list of things that are possible. You've worn me out!
<This post was basically destroyed by conference limts on number of blocks of quoted text.>
<This post was basically destroyed by conference limts on number of blocks of quoted text.>
Oh no, it was great just the way it is!
Sheesh Arnie, I'm really looking forward to the bitterness of old age...and the infallibility that goes with it. Like knowing what model speaker I own
I'll probably be listening exclusively to classical music by then.
Have a good one.
cheers,
AJ
Sean, I'm having trouble with a few of your characterizations here and perhaps the premises behind them. You claim that the results in mono are more sensitive and therefore better/more useful but it seems to me that the tests in stereo are inevitably going to be a truer test of actual in the field usage and while they may be less "sensitive" they are in fact more *accurate* and truer representation of listener preferences. You speak of "noise." What is the "noise" in the data? How are listener impressions more or less "noisy" in blind comparisons?
Stereo loudspeaker listening tests are "noisier" because there is more variance in the listener preference ratings from interactions between the loudspeakers and the recordings, and possibly less influence of off-axis sounds produced from the loudspeakers. The influence of these factors diminishes once you have stereo recordings with strong monophonic components. So listening tests in stereo are less revealing of problems in loudspeakers that are more easily heard in mono. If the loudspeaker wins the test in mono, it will win the test in stereo.
I will not argue that our tests are designed to be entirely representative of typical field conditions. In many ways they are not, and for good reasons. The tests are designed to be more sensitive to worst-case scenarios and toughest customer. Other ways that our tests are more sensitive than typical field conditions include:
-- use of trained listeners with normal hearing
-- the tests are double-blind to remove influence of sighted biases
-- we don't allow listeners to carry on conversations, read newspapers, talk on their cell phones or drink wine or beer before or during the tests
-- we carefully select short program segments known to reveal problems in loudspeakers
The same approach has been taken in the development and testing of audio codecs. When an audio codec is developed they don't just test them using top 40 Billboard music samples, or use listeners off the street, even though that may be more representative of typical field usage. Instead they carefully select the most sensitive music tracks ( harpsichord, horn pipe, clavichord, classical guitar) that probably 0.00001 of the population might actually listening to. This ensures that the performance of the codec meets the toughest worst-case signals it's likely to encouter.They also carefully select and train listeners so they are experts, and do controlled double-blind tests that are much more sensitive to hearing codec artifacts. We already suspect that in typical field conditions ( ipod earbuds listening to dynamically compressed thrash metal) non-discerning listeners will tolerate artifacts from low quality MP3 lossy coding. Under more sensitive listening conditions, they cannot tolerate these artifacts as shown in my recent study.
Our listening tests are not designed to measure what can we get away with under typical field usage conditions but rather satisfying the most discerning customer under the most sensitive listening conditions.
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
If the loudspeaker wins the test in mono, it will win the test in stereo.
Maybe .
Within the limited scope of your testing with those type of trained listeners. Where apparently, the spatial aspects of the stereophonic soundfield reproduction have no bearing. May also be why you can stand the stereophonic soundfield performance of your personal 2 channel loudspeaker system. If it is spatially unrealistic due to the closed loop format used to develop them, that makes perfect sense. Adding the multi channels would seem mandatory for such a 2 dimensional stereo front stage.
The same approach has been taken in the development and testing of audio codecs.
Except the mono part.
When an audio codec is developed they don't just test them using top 40 Billboard music samples, or use listeners off the street, even though that may be more representative of typical field usage. Instead they carefully select the most sensitive music tracks ( harpsichord, horn pipe, clavichord, classical guitar) that probably 0.00001 of the population might actually listening to. This ensures that the performance of the codec meets the toughest worst-case signals it's likely to encouter.
Hmmm. Why couldn't the same method (stereo) and specific music samples be used for (your?) loudspeakers???
They also carefully select and train listeners so they are experts, and do controlled double-blind tests that are much more sensitive to hearing codec artifacts.
Out of curiosity, what is the "reference" or anchor here for the trainees? Is it the real instrument(s), or an uncoded recording, as reproduced via "reference" electro-acoustic system X?
Our listening tests are not designed to measure what can we get away with under typical field usage conditions but rather satisfying the most discerning customer under the most sensitive listening conditions.
Which for the customer, is never mono (I hope ).
cheers,
AJ
p.s. also out of curiosity, can you make heads or tails out of what David was doing out there??? Do you share any of my concerns, especially about the "scrim" TF on power response at the listening area when trying to establish an "Audio Scene"?
Sean, I'm having trouble with a few of your characterizations here and perhaps the premises behind them. You claim that the results in mono are more sensitive and therefore better/more useful but it seems to me that the tests in stereo are inevitably going to be a truer test of actual in the field usage and while they may be less "sensitive" they are in fact more *accurate* and truer representation of listener preferences. You speak of "noise." What is the "noise" in the data? How are listener impressions more or less "noisy" in blind comparisons?
The fallacy here is that tests that duplicate or at least approach field usage are always best.
A major problem is that field usage is not always the same. A given speaker may be used a number of different ways even if we restrict the differences to numbers - in pairs, in quintuples, even in groups of 9 or 11. Which is the field usage that we use to test, given that testing resources are always finite? The logical answer is that the test should be done in the way that is most demanding and revealing. What if the most revealing and demanding usage condition is none of the above, but is in fact composed of just one speaker?
AFAIK, none of the critics of single speaker testing have actually done any real world testing of their hypotheses. None of them have presented the results of any technical models or simulations. None of them have used acoustical theory to support their beliefs. All I see from them are fuzzy assertions.
I presented a model of speaker testing that involved counting up the interactions between speakers under test, and categorizing which would be highly dependent on loudspeaker quality and which are simply there because of the number of speakers used. AFAIK there has been no rebuttal of it.
I think its time for the critics of single speaker testing to actually do something other than digging in their heels and repeatedly asserting that they and their anxieties and imaginations are right, and that people who have done technical tests of speakers and have developed speakers that have sold thousands, tens of thousands and even hundereds of thousands over a period decades are wrong simply because they say so.
Sean, I'm having trouble with a few of your characterizations here and perhaps the premises behind them. You claim that the results in mono are more sensitive and therefore better/more useful but it seems to me that the tests in stereo are inevitably going to be a truer test of actual in the field usage and while they may be less "sensitive" they are in fact more *accurate* and truer representation of listener preferences. You speak of "noise." What is the "noise" in the data? How are listener impressions more or less "noisy" in blind comparisons?
>> Stereo loudspeaker listening tests are "noisier" because there is more variance in the listener preference ratings from interactions between the loudspeakers and the recordings, and possibly less influence of off-axis sounds produced from the loudspeakers.>>
I don't quite see how that is "noise." Complexity and ambiguity in test results are not "noise." They are the proverbial "signal."
>> The influence of these factors diminishes once you have stereo recordings with strong monophonic components. So listening tests in stereo are less revealing of problems in loudspeakers that are more easily heard in mono. If the loudspeaker wins the test in mono, it will win the test in stereo.>>
Then where is the "varience" in the stereo tests? How do results both "vary" and "track?"
>>I will not argue that our tests are designed to be entirely representative of typical field conditions.>>
That would be a good idea because I didn't mention "typical" field conditions. I refered to actual field conditions which run a wide gamut but pretty much always includes two channel stereo recordings.
>> In many ways they are not, and for good reasons. The tests are designed to be more sensitive to worst-case scenarios and toughest customer.>>
looks like you are hinting at a false dichotomy. "Actual in the field usage" includes worst case scenereos and your toughest customers.
>> Other ways that our tests are more sensitive than typical field conditions include:
-- use of trained listeners with normal hearing
-- the tests are double-blind to remove influence of sighted biases
-- we don't allow listeners to carry on conversations, read newspapers, talk on their cell phones or drink wine or beer before or during the tests
-- we carefully select short program segments known to reveal problems in loudspeakers
The same approach has been taken in the development and testing of audio codecs. When an audio codec is developed they don't just test them using top 40 Billboard music samples, or use listeners off the street, even though that may be more representative of typical field usage.>>
Again there was no mention of anything "typical." But it does seem that there is a pattern of narrowness designed to represent a broader spectrum. I would be concerned of an accumulative effect. How far removed from actual in field usage can you veerand narrow before the "tracking" is buried in the "noise?" That which you can "noise" is representative of preferences formed under actual usage conditions.
>> Instead they carefully select the most sensitive music tracks ( harpsichord, horn pipe, clavichord, classical guitar) that probably 0.00001 of the population might actually listening to. This ensures that the performance of the codec meets the toughest worst-case signals it's likely to encouter.>>
That seems to make more sense when testing for audible "differences" but in preference tests it seems pretty potentially problematic in representing actual in field usage. I was under the impression you used something quite different for your DBTs. Is this what you use? Harpsichord, horn pipe, clavichord and classical guitar in mono for your preference tests? If so which recordings and why those particular recordings?
>>They also carefully select and train listeners so they are experts, and do controlled double-blind tests that are much more sensitive to hearing codec artifacts. >>
Again that makes sense for detection but for preferences.....Who is an expert in taste? taste is involved in preferences. Again i see a pattern of narrowing the scope of the tests in ways that may potentially fail to tell the broader story.
>> We already suspect that in typical field conditions ( ipod earbuds listening to dynamically compressed thrash metal) non-discerning listeners will tolerate artifacts from low quality MP3 lossy coding. Under more sensitive listening conditions, they cannot tolerate these artifacts as shown in my recent study>>
Again i have made no mention of anything "typica"l in reference to actual in field usage. Actual in field usage includes, by definition, the most sensitive listening conditions that would ever be found in actual in field usage. sensitivity beyond that (That which can be heard in the most sensitive of in field usage) helps in what way?
>> Our listening tests are not designed to measure what can we get away with under typical field usage conditions but rather satisfying the most discerning customer under the most sensitive listening conditions.>>
And why exactly would tests that mimic actual usage fail to satisfy the most discerning customer? It seems to me that even the most discerning customer would not demand something beyond what they can hear in their own actual usage? I'm not trying to be a pain in the ass or just a contrarian but the choices of testing in mono with Harpsichord, clavichord, guitar and horn pipe (and how many different variations of recordings do you use for testing?) and the claims of this narrow band of test conditions tracking actual in field usgae *seems* unlikely. Maybe we need to discuss what it means that the tests in mono "track" the tests in stereo despite the added "noise" found in testing in stereo. I have a few follow up questions about that. Do they track exactly just with less magnitude of preference or do some listeners vary their preferences when tested in stereo? The source material you use, how does it compare with the wide range of source material that even your toughest customers likely may use with your speakers? It seems your findings place a strong emphasis on the importance of flat frequency response. how does this transfer to real world usage where so many recordings have skewed frequency responses? How "noisy" are the actual results one gets at home with a varitable plethora of recordings spanning the many years and many genres? How well do the results track at that point? I know that would be pretty tough to test. I'm not trying to be cynical here. But it is an issue IMO.
And why exactly would tests that mimic actual usage fail to satisfy the most discerning customer?
Tests that mimic actual usage fail to satisfy customers when their actual usage turns out to be more demanding than the actual usage that was anticipated during development.
In product development it is a general rule to test products under conditions that are somewhat more demanding than any anticipated use. Cars are commonly driven at illegal speeds during development. Foods are stored longer than their expiration date and then tested. Paint destined for cars sold in Michigan is tested in Florida where the heat, humidity and sunlight is far more prolonged and intense than Michigan.
Ever hear of accelerated life testing? Customer own and use equipment for years even decades. We don't have years and decades for product development.
Far more conditions should be tested during development than any single real world user is likely to experience since customers tend to have different needs and usage patterns.
Furthermore, an actual user might use a piece of equipment for a long period of time before he notices some fault. During product deveopment it is common to use test procedures that accelerate the manifestation and detection of faults.
For example, leaks in speaker cabinets can cause annoying whistling and rushing sounds. Therefore high pressure testing might be used to test an enclosure design or producting technique, even though the enclosure would never be highly presurized in actual use.
For another example, it is common to test speakers at higher SPLs than any real world user would be likely to use. Tweeters and woofers are often tested at frequenceis that are outide of the range that is possible in actual use due to the presence of crossovers, fuses, protection circuits, etc.
I think its time for the critics of single speaker testing to actually do something
No problem Arnie. How about I swing by tomorrow and pick up your automated loudspeaker shuffler?
And while I'm there, can I borrow your "Scrim" material that has the same amount of power loss for both monopoles
and dipoles at the listening area?
We'll work on the whole what constitutes a "trained" listener later.
TIA
The fallacy here is that.....
.... and have developed speakers that have sold thousands, tens of thousands and even hundereds of thousands over a period decades are wrong simply because they say so.
I like the transition from wagging your fingers about fallacies, right into an appeal to popularity. Nice
Once I get the screen material thing figured out, I'll do my best to get something organized.
cheers,
AJ
Tests that mimic actual usage fail to satisfy customers when their actual usage turns out to be more demanding than the actual usage that was anticipated during development.
If their "actual usage" turns out to be more demanding than the tests then the tests didn't really mimic the "actual usage" of the customers. It should be no secret what range of demands need to be represented in any such tests. The thresholds of human hearing are pretty well documented. Heck Sean already mentioned clavichord, harpsichord, classical guitar and something else as the ultimate references that represent worst case scenereos for the most demanding and discriminating listeners.
In product development it is a general rule to test products under conditions that are somewhat more demanding than any anticipated use. Cars are commonly driven at illegal speeds during development. Foods are stored longer than their expiration date and then tested. Paint destined for cars sold in Michigan is tested in Florida where the heat, humidity and sunlight is far more prolonged and intense than Michigan.
Were not talking about cars here Arny we are talking about products that deal only with aesthetic values. Let me see you use this example for say a chef who is working on a new recepie. How would that chef go about using "conditions that are somewhat more demanding than anticipated use?"
Ever hear of accelerated life testing? Customer own and use equipment for years even decades. We don't have years and decades for product development.
yeah and it's great way to test wear and tear. Not aesthetic preferences. It also has nothing to do with Sean's methodologies or my concerns with Sean's methodologies
Far more conditions should be tested during development than any single real world user is likely to experience since customers tend to have different needs and usage patterns.
My issue isn't with Sean testing "far more conditions" than "any single real world listener is likely to experience." Quite the opposite! It looks to me that the risk is that his tests encompass "far less" than "any single real world listener is likely to experience" to the point of disconnect with most of what "any single real world listener is likely to experience." I'm pretty confident that very little of what the real world listener is likely to experience' falls into the catagories of classical guitar, harpsichord and clavichord in mono.
Furthermore, an actual user might use a piece of equipment for a long period of time before he notices some fault. During product deveopment it is common to use test procedures that accelerate the manifestation and detection of faults.
It's not a question of efficiency but of accuracy. Sean has already explained how his methodolies are more efficient. I buy that. The question for me is how well his tests, which are really far removed from how the products are used in the field, translate to actual in field usage. I have asked Sean some specific questions about the tracking he asserts with his testing methodologies and actual usage. I look forward to his answers.
I think its time for the critics of single speaker testing to actually do something other than digging in their heels and repeatedly asserting that they and their anxieties and imaginations are right, and that people who have done technical tests of speakers and have developed speakers that have sold thousands, tens of thousands and even hundereds of thousands over a period decades are wrong simply because they say so.
Does that mean Bose has been right all these years?
In product development it is a general rule to test products under conditions that are somewhat more demanding than any anticipated use. Cars are commonly driven at illegal speeds during development. Foods are stored longer than their expiration date and then tested. Paint destined for cars sold in Michigan is tested in Florida where the heat, humidity and sunlight is far more prolonged and intense than Michigan.
Were not talking about cars here Arny we are talking about products that deal only with aesthetic values.
And there is the core of your problem, Scott. Audio is both an art and a science, not just an art. You have repeatedly ignored the science part of audio. Audio components are not just about pleasing people but also generally about being accurate reproducers.
Scott the following statement of yours is simply false: "we are talking about products that deal *only* with aesthetic values." It is false because of the word "only".
In the real world every product must address both aesthetic and practical values. The art of painting creates an illustration with a certain appearance, and the science of painting makes the painting last long enough for many of us to enjoy.
I can see where a makeup artist might forget many of the practical aspects of his art given that much of it lasts for a day or less. However, there is pracitcal science to even makeup art - makeup should not greatly endanger or be fatal to its wearer, for example.
You must never forget the relevant science, Scott.
Furthermore, an actual user might use a piece of equipment for a long period of time before he notices some fault. During product deveopment it is common to use test procedures that accelerate the manifestation and detection of faults.
It's not a question of efficiency but of accuracy.
Right. Can't you see how the use of efficient procedures enhances accuracy? Nobody can take forever to develop products. The quicker that inaccuracy is rooted out, the more of it can be dealt with in the time that is available.
Sean has already explained how his methodolies are more efficient. I buy that.
Then due to practical constraints, Sean has made his point. Why not give it to him?
The question for me is how well his tests, which are really far removed from how the products are used in the field, translate to actual in field usage.
Ultimately, the best evidence about the effectiveness of the development procedures lies in the products that are developed this way.
Using one of your own arguments against you Scott, when is the last time you listened to Harman Products that are developed Sean's way? We've got a specific example right here before us - the Infinity Primus 362 or a member of the same line of products. Scott, they are relatively inexpensive and widely sold. Are you open miinded enough to get some practical experience with how Sean's methodologies translate into actual products? Ironically I own a pair of some smaller products in that line, and IME they are really quite good. In contrast I've never heard a Martin Logan product that impressed me enough to make me want to stay in the room very long.
In reality mid-priced high volume products such as Primus speakers usually have far more development time put into them over all than most highly expensive speakers. The costs are divided up among a large number of units, and the cost per individual product sold can be very reasonable.
In product development it is a general rule to test products under conditions that are somewhat more demanding than any anticipated use. Cars are commonly driven at illegal speeds during development. Foods are stored longer than their expiration date and then tested. Paint destined for cars sold in Michigan is tested in Florida where the heat, humidity and sunlight is far more prolonged and intense than Michigan.
Were not talking about cars here Arny we are talking about products that deal only with aesthetic values.
And there is the core of your problem, Scott. Audio is both an art and a science, not just an art. You have repeatedly ignored the science part of audio. Audio components are not just about pleasing people but also generally about being accurate reproducers.
Scott the following statement of yours is simply false: "we are talking about products that deal *only* with aesthetic values." It is false because of the word "only".
In the real world every product must address both aesthetic and practical values. The art of painting creates an illustration with a certain appearance, and the science of painting makes the painting last long enough for many of us to enjoy.
I can see where a makeup artist might forget many of the practical aspects of his art given that much of it lasts for a day or less. However, there is pracitcal science to even makeup art - makeup should not greatly endanger or be fatal to its wearer, for example.
You must never forget the relevant science, Scott.
Sorry to break format here but I think it will help.
>> And there is the core of your problem, Scott. Audio is both an art and a science, not just an art.>>
1. so is cooking and I see you failed to show how your claim that "it is a general rule to test products under conditions that are somewhat more demanding than any anticipated use." in the development of products such as new recepies. IOW products made primarily to serve the aesthetic experience.
>> You have repeatedly ignored the science part of audio. Audio components are not just about pleasing people but also generally about being accurate reproducers. >>
You couldn't be more wrong Arny. When discussing testing methodologies for *listener preferences* it absolutely *is* JUST about pleasing people and it seems to be Sean's position that the point of accuracy in speakers is *because* his tesing demonstrates that *accuracy* on certain parameters corolates directly with "pleasing people." The questions I am asking relate to the issue of his tests tracking real world usage. That's all.
>> In the real world every product must address both aesthetic and practical values. >>
That's fine Arny but this is a discussion of testing for listener preference. It's all about aestheitc values. introducing anything opther than aesthetic values in blind listener tests for preferences is a red herring.
>>The art of painting creates an illustration with a certain appearance, and the science of painting makes the painting last long enough for many of us to enjoy.
I can see where a makeup artist might forget many of the practical aspects of his art given that much of it lasts for a day or less. However, there is pracitcal science to even makeup art - makeup should not greatly endanger or be fatal to its wearer, for example.>>
Well that speaks to your ignorance of my proffession but that is neither here nor there. The issue is not whether Sean's methodologies are sufficiently testing for potential fatal health hazards. OTOH toxicity of products used on real people in make up artistry is an issue. Unlike Sean I do have to be aware of health hazards in my work.
>>You must never forget the relevant science, Scott.>>
And you must try not forget the concept of "relevance" Arny.
The above are either straw man or excluded-middle arguments depending on how you take them.
OP: "mono test results track stereo results"
Me: "not always, here's two trivial counter examples"
You: "straw man"
HA search says you've claimed "straw man" 11 times now. You should build a haystack.
When faced with poor quality rhetoric, one may start building haystacks to pass the time while waiting for intelligent disucssion to resume.
2 days to come up with that retort. Any, you're slacking.
Believe it or not, I don't come here for rhetoric. I come here to learn and explore. Unexpectedly (well, it was to start with - I'm getting use to it now), you act as a block to that in most threads you contribute to. Arny holds the keys to all knowledge, and anyone who questions anything he says must be an audiophool subjectivist, to be treated with contempt.
While some in this thread may have a axe to grind wrt electrostatic speakers, I think most are finding it hard to believe that various departures from "real world listening" are as benign or even useful as is suggested. It doesn't make these people hard core subjectivists - it sounds more like rigorous science to me.
I
can believe that it's easier to hear certain differences in the listening set-up described by Sean than in normal stereo, but it won't always mean there's a perfect correlation between mono and stereo results (even if there "always" has been in the past, and even that misstates the truth!), and it certainly doesn't mean that the mono tests will catch problem with stereo reproduction, even if "experience" suggests that "usually" there aren't any such issues.
Cheers,
David.
The same approach has been taken in the development and testing of audio codecs. When an audio codec is developed they don't just test them using top 40 Billboard music samples, or use listeners off the street, even though that may be more representative of typical field usage. Instead they carefully select the most sensitive music tracks ( harpsichord, horn pipe, clavichord, classical guitar) that probably 0.00001 of the population might actually listening to. This ensures that the performance of the codec meets the toughest worst-case signals it's likely to encouter.
To be fair, we know
now that these are "tough" for modern codecs, but...
(a) that wasn't definitively known before the fact (e.g. the EBU SQAM CD contains almost every instrument they could find - half of them rarely reveal any coding problems!), and
(b) there are other signals, including electronic music, pure digital impulses, and various artificial test signals, which are equally, if not more problematic for codecs.
Some of the most problematic signals aren't great predictors of how the codec will perform with other signals. e.g. a string of impulses will cripple some otherwise fairly good lossy codecs, whereas ZIP handles it very efficiently (but isn't so efficient with real music!).
I agree the same principles apply - and this means that it isn't always immediately obvious which signal type will reveal the problems of a given speaker. It's less of a problem, because speakers are usually less devious in their design than an audio codec but there's still the possibility of it performing extremely well with 3 or 4 "difficult" musical signal and falling apart with some other specific signal.
Cheers,
David.
Furthermore, an actual user might use a piece of equipment for a long period of time before he notices some fault. During product deveopment it is common to use test procedures that accelerate the manifestation and detection of faults.
It's not a question of efficiency but of accuracy.
Right. Can't you see how the use of efficient procedures enhances accuracy? Nobody can take forever to develop products. The quicker that inaccuracy is rooted out, the more of it can be dealt with in the time that is available.
Sean has already explained how his methodolies are more efficient. I buy that.
Then due to practical constraints, Sean has made his point. Why not give it to him?
The question for me is how well his tests, which are really far removed from how the products are used in the field, translate to actual in field usage.
Ultimately, the best evidence about the effectiveness of the development procedures lies in the products that are developed this way.
Using one of your own arguments against you Scott, when is the last time you listened to Harman Products that are developed Sean's way? We've got a specific example right here before us - the Infinity Primus 362 or a member of the same line of products. Scott, they are relatively inexpensive and widely sold. Are you open miinded enough to get some practical experience with how Sean's methodologies translate into actual products? Ironically I own a pair of some smaller products in that line, and IME they are really quite good. In contrast I've never heard a Martin Logan product that impressed me enough to make me want to stay in the room very long.
In reality mid-priced high volume products such as Primus speakers usually have far more development time put into them over all than most highly expensive speakers. The costs are divided up among a large number of units, and the cost per individual product sold can be very reasonable.
I have acknowledged the value of efficiency in product development due to the fact that it is a business but Sean's claims *seem* to extend to a claim that his tests are in no way lacking because of the need for efficiency. His claim is that his methodlogies "track." I am just having some problems with that claim to the extent they would really track for me as an audiophile and music lover just from the magnitude of the differences between his testing conditions and in field usage for which these tests are supposed to "track." I don't think I am the only one scratching their head on this one.
I don't remember the exact date I listened to the Revel Ultima Salons but they were their flagship speaker when they came out. I thought they were very good speakers too. Do you believe the Infinity Primus 362 are better than the Revel Salons? I think we need to both be careful here of TOS #8. I did not listen and compare the Revel Salons under blind conditions to my speakers so my preference is biased. I am going to go out on a limb and guess that none of your auditions of ML speakers were done under bias controlled conditions much less any experience you may have ever had with my speakers, the Soundlab A3s.
I have acknowledged the value of efficiency in product development due to the fact that it is a business but Sean's claims *seem* to extend to a claim that his tests are in no way lacking because of the need for efficiency. His claim is that his methodlogies "track."
I agree that Sean claims that his methodolgoies track. I disagree that Sean would cliam that his methodologies in no way lacking. In the real world you don't have to be perfect, being better than any other human is more than suffcient. ;-) And you don't have to be the best all the time - being very, very good generally suffices.
I am just having some problems with that claim to the extent they would really track for me as an audiophile and music lover just from the magnitude of the differences between his testing conditions and in field usage for which these tests are supposed to "track." I don't think I am the only one scratching their head on this one.
I think the source of the head scratching is problems not fundamentally with Sean's methodologies. I think the source of the head scratching is a whole passel of misapprensions about audio.
I don't remember the exact date I listened to the Revel Ultima Salons but they were their flagship speaker when they came out. I thought they were very good speakers too.
So can you chalk that up as a point in favor of the procedures that Sean uses?
Do you believe the Infinity Primus 362 are better than the Revel Salons?
I don't see that as a relevant question. But since you asked, I would seriously hope that the Revels are the better speakers, given the cost differences.
I think we need to both be careful here of TOS #8.
I don't think that TOS 8 is relevant as long as we stick with the idea that speakers tend to sound different. I sense no controversy about that!
I did not listen and compare the Revel Salons under blind conditions to my speakers so my preference is biased.
I don't know that you actually formed a preference. I know of many speakers that sound good to me that I may not prefer. I don't even try to prefer speakers. I"m very happy enough when the speakers I am listening to sound good enough to me, right then.
I am going to go out on a limb and guess that none of your auditions of ML speakers were done under bias controlled conditions much less any experience you may have ever had with my speakers, the Soundlab A3s.
Since I was not comparing them to anything but my preferences for sound quality, there was no opportunity for a blind comparison. The sound of every speaker is profoundly affected by the room and placement in the room. I never have had any control over the rooms or placements. All I said is that I've never liked what I heard when I heard them. That's about 3 million miles from TOS 8 since they are speakers, and not speaker cables or magic blocks of wood.
I have acknowledged the value of efficiency in product development due to the fact that it is a business but Sean's claims *seem* to extend to a claim that his tests are in no way lacking because of the need for efficiency. His claim is that his methodlogies "track."
I agree that Sean claims that his methodolgoies track. I disagree that Sean would cliam that his methodologies in no way lacking. In the real world you don't have to be perfect, being better than any other human is more than suffcient. ;-) And you don't have to be the best all the time - being very, very good generally suffices.
I am just having some problems with that claim to the extent they would really track for me as an audiophile and music lover just from the magnitude of the differences between his testing conditions and in field usage for which these tests are supposed to "track." I don't think I am the only one scratching their head on this one.
I think the source of the head scratching is problems not fundamentally with Sean's methodologies. I think the source of the head scratching is a whole passel of misapprensions about audio.
I don't remember the exact date I listened to the Revel Ultima Salons but they were their flagship speaker when they came out. I thought they were very good speakers too.
So can you chalk that up as a point in favor of the procedures that Sean uses?
Do you believe the Infinity Primus 362 are better than the Revel Salons?
I don't see that as a relevant question. But since you asked, I would seriously hope that the Revels are the better speakers, given the cost differences.
I think we need to both be careful here of TOS #8.
I don't think that TOS 8 is relevant as long as we stick with the idea that speakers tend to sound different. I sense no controversy about that!
I did not listen and compare the Revel Salons under blind conditions to my speakers so my preference is biased.
I don't know that you actually formed a preference. I know of many speakers that sound good to me that I may not prefer. I don't even try to prefer speakers. I"m very happy enough when the speakers I am listening to sound good enough to me, right then.
I am going to go out on a limb and guess that none of your auditions of ML speakers were done under bias controlled conditions much less any experience you may have ever had with my speakers, the Soundlab A3s.
Since I was not comparing them to anything but my preferences for sound quality, there was no opportunity for a blind comparison. The sound of every speaker is profoundly affected by the room and placement in the room. I never have had any control over the rooms or placements. All I said is that I've never liked what I heard when I heard them. That's about 3 million miles from TOS 8 since they are speakers, and not speaker cables or magic blocks of wood.
Well perhaps I am misunderstanding the terms of service but if Sean Olive and Co. have taught the world of audio one important lesson it would that bias effects are really *more* of a concern in things like auditioning loudspeakers than in amps. cables and "magic" anything. After all if someone expressed a preference for something that makes no audible change in the system then they have done no harm to the actual sonic quality of their system. OTOH if one is forming preferences that would sound inferior to them under blind conditions then real harm is being done.
So at the very least out of respect for Sean's work I am not going to just say my speakers are better than his just because that is my biased preference. If you feel comfortable making claims about your impressions of various speakers that were formed with your biases in play on a thread started by Sean Olive and it does not violate the TOS then that is a choice you are free to make.
I would love to drag my speakers over to HK and do some blind comparisons with the Revels. I am actually interested in unbiased preferences. But I would only be interested if it were done in stereo and that I be allowed to set the position of the Soundlabs and use whatever room treatment I see fit and use my choice of source material. At least then I know it would be a fair and relevant comparison for me as a consumer. I don't see that happening. But it would be fun and informative.
I would love to drag my speakers over to HK and do some blind comparisons with the Revels. I am actually interested in unbiased preferences. But I would only be interested if it were done in stereo and that I be allowed to set the position of the Soundlabs and use whatever room treatment I see fit and use my choice of source material. At least then I know it would be a fair and relevant comparison for me as a consumer. I don't see that happening. But it would be fun and informative.
It won't be happening because Sean has limited motivation to do this work. His documented research and experimental experience indicates that stereo testing makes speakers sound more alike. Further stereo comparisons are therefore not of great interest to him.
You clearly believe otherwise. So how about expending some of your own resources and do your own testing. If you come up with interesting results, you'll get credit for the discovery, there will be something new to talk about here and Sean and others may become more motivated to dig deeper with you.
Well perhaps I am misunderstanding the terms of service but if Sean Olive and Co. have taught the world of audio one important lesson it would that bias effects are really *more* of a concern in things like auditioning loudspeakers than in amps. cables and "magic" anything.
You've over-simplified things.
Bias potentially affects evaluation more or less equally. So right away your claim that "...bias effects are really *more* of a concern in things like auditioning loudspeakers than in amps" is likely an overstatement of the true situation.
Bias is a concern no matter what you are evaluating.
However, Bias is less of an issue when the actual facts of the matter are easier to perceive. Bias is one influence, and audible differences are another influence. The strength of the influence of bias is usually less signficant when the actual facts of the matter create a very strong influence.
The application of this knowlege is different depending on the general nature of the evaluation. Some things really do sound vastly different, so when someone reports that two things sound different and they have a preference for one of them, their claim has a lot more crediblity than someone else reporting that they heard a difference that probably of certainly does not exist. If there is no audible difference, then any preference that are stated are obviously 100% due to bias. If there is a strong difference, there is a good probability that any stated preference is due that that strong difference.
When there is no difference to hear at all, then any perceptions of a different are entirely due to bias. When there is a readily-perceived differerence, then it can even overcome strongly-held biases. We can assume that there is no difference, and then use blind tests to determine the strength of the difference. If we find that certain influences are generally very strong, then we may even disregard the effects of bias.
Let's condsider an extreme case. We compare two systems that have a strong audible difference. One is working normally and the other is turned off. I say that they sound different. What sort of confirmation do I need to confirm that claim? I say that I prefer the one that is making music. What sort of confirmation do I need to confirm that claim?
Now lets consider another extreme case at the other end of the normal range of possible differences. We compare two systems that have technical performance that is essentially identical. Every aspect of each systems performance matches the other system within a small tolerance. The differences are down there in the noise created by the repeatibility of the measurements. I again say that I hear a difference and prefer one of the two systems. What sort of confirmation do I need to confirm that claim? This is where TOS 8 comes in.
After all if someone expressed a preference for something that makes no audible change in the system then they have done no harm to the actual sonic quality of their system.
They still may have caused potential harm of other kinds. They've increased the potential for system failure. They may have mislead themselves that there was in fact an improvement. When they go around posting on the internet that the alleged difference is very important, they create a lot of confusion. This happens all the time.
OTOH if one is forming preferences that would sound inferior to them under blind conditions then real harm is being done.
OK.
So at the very least out of respect for Sean's work I am not going to just say my speakers are better than his just because that is my biased preference.
You've over-personalizing the situation. Unless you have bias-controlled evaluations to back your statements up, your preference is very questionable. It doesn't matter what you think of Sean or his work.
If you claim that the two speaker systems in two different rooms sound different, that is believable, but also trivial. TOS 8 does not affect that claim. If you say that you prefer the sound of your speakers in your room, that is also believable and since it is a personal opinon, it does not require confirmation. If you say that your speakers are more accurate, or generaly make musical instruments sound live, then we have to worry about TOS 8.
If you feel comfortable making claims about your impressions of various speakers that were formed with your biases in play on a thread started by Sean Olive and it does not violate the TOS then that is a choice you are free to make.
Personal opinons stated as personal opinions don't require verification. I can say that I prefer tea over coffee without a battery of DBTs to back me up.
I would love to drag my speakers over to HK and do some blind comparisons with the Revels. I am actually interested in unbiased preferences.
But I would only be interested if it were done in stereo and that I be allowed to set the position of the Soundlabs and use whatever room treatment I see fit and use my choice of source material.
Sean may or may not want to do that for fun, but so far I don't see a lot that would be likely to be of much technical interest to him. For example you might bring in speakers that only sound right if you fill the room half way up with empty beer cans. That fits your "room treatment you see fit" criteria, right?
At least then I know it would be a fair and relevant comparison for me as a consumer. I don't see that happening. But it would be fun and informative.
Mostly fun for you, I suspect. For most of us, our involvement would just be work, and for what?
I'd be interested in actual measurements of your speakers, and a blind preference test comparing your speakers to say the Infinity Primus, with the room set up like an IEC standard listening room. Of course, no beer cans! ;-)
I would love to drag my speakers over to HK and do some blind comparisons with the Revels. I am actually interested in unbiased preferences. But I would only be interested if it were done in stereo and that I be allowed to set the position of the Soundlabs and use whatever room treatment I see fit and use my choice of source material. At least then I know it would be a fair and relevant comparison for me as a consumer. I don't see that happening. But it would be fun and informative.
You are interested in unbiased preferences.
You want your own preference of mono/stereo, speaker positioning and room treatments, and source material.
You think that would be fair and relevant.
Unbiased? Uh huh.
His documented research and experimental experience indicates that stereo testing makes speakers sound more alike.
That's not what I gathered. He seems to be saying that it is easier to hear non-linearities in mono, with some consistency.
I agree with that.
Specific non-linearities. In stereo (or as you continue to add channels/sources), the soundfield becomes more complex and our perceptions of the very same non-linearities in the individual source, becomes more difficult to discern with consistency.
What I (and obviously others) disagree with, is that the best way to determine the stereophonic performance of a pair(+) of loudspeakers, is to listen to an individual channel in mono in the middle of a front wall and then make projections.
Further stereo comparisons are therefore not of great interest to him.
That is certainly his prerogative, given his view that (2 channel) stereo is essentially dead (and i.e., using only the type of monopolar stereo speakers he develops). This despite the fact that probably 99% of music recordings over the last 40 yrs are.....stereo. He advocates extra channels (such as Logic 7) for realism (again something I can concur with) as mandatory, but feels that the 2 front channels are incapable of (or possibly should not be used for) 3 dimensionality/frontal "auditory scene" realism.
cheers,
AJ
I would love to drag my speakers over to HK and do some blind comparisons with the Revels. I am actually interested in unbiased preferences. But I would only be interested if it were done in stereo and that I be allowed to set the position of the Soundlabs and use whatever room treatment I see fit and use my choice of source material. At least then I know it would be a fair and relevant comparison for me as a consumer. I don't see that happening. But it would be fun and informative.
It won't be happening because Sean has limited motivation to do this work.
really? He strikes me as a fairly motivated worker.
His documented research and experimental experience indicates that stereo testing makes speakers sound more alike. Further stereo comparisons are therefore not of great interest to him.
That's fine but i primarily listen in stereo and so they are of prime interest to me.
You clearly believe otherwise. So how about expending some of your own resources and do your own testing.
Sorry I don't have spare millions to spend on the facilities Sean uses.
I would love to drag my speakers over to HK and do some blind comparisons with the Revels. I am actually interested in unbiased preferences. But I would only be interested if it were done in stereo and that I be allowed to set the position of the Soundlabs and use whatever room treatment I see fit and use my choice of source material. At least then I know it would be a fair and relevant comparison for me as a consumer. I don't see that happening. But it would be fun and informative.
You are interested in unbiased preferences.
You want your own preference of mono/stereo, speaker positioning and room treatments, and source material.
You think that would be fair and relevant.
Unbiased? Uh huh.
yes it would be fair to compare the two speaker systems in their optimum envirement and set up. As a long time user of the Soundlab A3s I think I could do that but it would mean positioning the Soundlabs in a way I know from experience works and doing what it takes to get the room optimised for them. Are you suggesting it is less fair to have the Soundlabs perform at their best? I would think that would be more fair.
My choice of source material would reflect my listening habbits. That would make the test relevant to me. Do you think using material I would never listen to would make the tests more relevant to me? I would think it would make them less relevant to me.
So yeah, blind and unbiased.
My choice of source material would reflect my listening habbits. That would make the test relevant to me. Do you think using material I would never listen to would make the tests more relevant to me? I would think it would make them less relevant to me.
So yeah, blind and unbiased.
The choice of source material definately biases the outcome of any listening evaluation. One common hi fi salesman trick (been there, done that for about 7 years) is to find music that makes every speaker sound good, or that makes the speakers you want to sell sound good. Sometimes the most useful recording is the one that makes the high cost, high margin speakers sound great, but makes the low magin speakers sound bad. Often this recording sounds bad on the speakers you don't want to sell because they are more accurate and reveal the general failings of the recording.
There is definately a class of music that is called variously "audiophile recordings" or "sonic spectaculars". There is another class of recording that is exceptionaly diagnostic for differences in audio gear. Sometimes the twain shall meet and sometimes they shall be kept as far apart as possible, depending on the sales strategy.
Often good diagnostic recordings sound generally ugly, JJ's *favorite* Susanne Vega recording being an example of this.
It is not uncommon for audiophiles to focus their listening on recordings that make their audio systems sound good to them, often at a huge cost in musical values. But whether the musical values are bad or good, recordings that make the current system sound good may be a big stumbling block for upgrades because they could easily sound worse on the upgraded system. This gets us into the *true* meaning of audiophile component break in, which is really all about changing the listener.
One classic comment from a Stereophile reviewer: "They want the test recordings to sound good". If an audio component has a relevant failing, as a developer I would very much prefer that it sound bad and proper choice of recording is one way to do this.
My choice of source material would reflect my listening habbits. That would make the test relevant to me. Do you think using material I would never listen to would make the tests more relevant to me? I would think it would make them less relevant to me.
So yeah, blind and unbiased.
The choice of source material definately biases the outcome of any listening evaluation. One common hi fi salesman trick (been there, done that for about 7 years) is to find music that makes every speaker sound good, or that makes the speakers you want to sell sound good. Sometimes the most useful recording is the one that makes the high cost, high margin speakers sound great, but makes the low magin speakers sound bad. Often this recording sounds bad on the speakers you don't want to sell because they are more accurate and reveal the general failings of the recording.
There is definately a class of music that is called variously "audiophile recordings" or "sonic spectaculars". There is another class of recording that is exceptionaly diagnostic for differences in audio gear. Sometimes the twain shall meet and sometimes they shall be kept as far apart as possible, depending on the sales strategy.
Often good diagnostic recordings sound generally ugly, JJ's *favorite* Susanne Vega recording being an example of this.
It is not uncommon for audiophiles to focus their listening on recordings that make their audio systems sound good to them, often at a huge cost in musical values. But whether the musical values are bad or good, recordings that make the current system sound good may be a big stumbling block for upgrades because they could easily sound worse on the upgraded system. This gets us into the *true* meaning of audiophile component break in, which is really all about changing the listener.
One classic comment from a Stereophile reviewer: "They want the test recordings to sound good". If an audio component has a relevant failing, as a developer I would very much prefer that it sound bad and proper choice of recording is one way to do this.
I think you are kind of putting the horse before the cart here for many audiophile/music lovers. Yes I have actually witnessed the audiophile who listens to a small number of audiophile recordings that show off the system. There is a fellow over on AA who has such a collection of LPs. I believe his name is John Ellison. But hey, if that is what he digs so be it. But for myself and many others the gear does not dictate our musical taste. Quite the opposite. I would not be so concerned about two channel stereo as opposed to multichannel if it were not for the fact that most of my favorite music was recorded in two channel stereo with the idea of being played through stereo speakers. So if I am guilty of anything it is tailering the gear to the music. I'm not a fan of Hendrix or the Doors because they made audiophile spectaculars. I am however a fan of guys like Steve Hoffman who took the recordings of artists like the The Doors, and Elvis, and CCR and Jim Croce just to name a few and produced what is for me the very best sounding versions of those artists' work. I am a fan of my Koetsu cartridge which consistantly delivers to my biased ears much better sound on lesser quality recordings.
So for my auditions to be relevant to me I will always insist on choosing the source material. I'm not interested in some abstract performance goals. I want to hear the music I love sound as good to my ears as possible.
When I said I would love to drag my speakers over to HK for some comparisons I was just musing. I don't for one second expect Sean to do this for me nor do I expect anyone on these boards to jump on a plane and come over to do some hard labor.
I don't think you can fake good sound with certain recordings. If a system sounds good with a given recording then it sounds good. OTOH I do think sales people can dazzle inexperienced customers with boom and sizzle. But that is a different story. That isn't good sound. Well at least not to me. But taste does come into play here. Who are any of us to judge the guy who drives down the street listening to rap with bass so loud that you heard it five minutes before his car went by? I have a feeling Sean's tests don't track this person's taste. and who are any of us to tell this guy he needs to change his taste to suit our perosnal ideals in audio? Ideals that ultimately vary from one individual to another anyway.
And for what it is worth the recordings I use for extensive auditions run the gamut of sounding really good to not so good. I have selected titles that I feel both represent the extremes and represent an accurate and thourough sampling of what I listen to. It's hardly perfect but it works for me. i am enjoying the hobby and the music. I ronically I include solo piano (not quite a harpsichord) solo classical guitar and music that does include some pretty demanding precussion.
It seems to me an obvious conclusion. Hearing deteriorates over time. However, I wonder what will happen if you can get all the subjects to return when they've reached the age group they're being tested against... I suspect our world's become a lot more noisy and many people will have worse hearing at the same age as last generation. In part because people are stupid with earphones, but also because more people live in cities, where traffic has only gotten worse.
Oh, I guess I'm misunderstanding the experiment. Apparently there's a claim that the younger generation might prefer artifacts and scratches in their recordings... Who thinks this? Vinylphiles that complain about kids these days? Even in genres of heavy, purposeful distortion (rock music played with pedals), it is really distracting and annoying to hear it interrupted by a sound that clearly doesn't belong. I bet even if you did this test on fans of John Cage they'd still be able to say that the distortion is interrupting their enjoyment of his rhythmless compositions.
I just moved this post by Chef into this discussion from the one in Validated news that was double-posted (and now deleted, please see TOS #6 if you have any questions, Sean). Perhaps it might get this discussion back on topic?
can I believe that it's easier to hear certain differences in the listening set-up described by Sean than in normal stereo, but it won't always mean there's a perfect correlation between mono and stereo results (even if there "always" has been in the past, and even that misstates the truth!), and it certainly doesn't mean that the mono tests will catch problem with stereo reproduction, even if "experience" suggests that "usually" there aren't any such issues.
That is a fair statement that I would agree with. When I can find time, I will repeat the Infinity Primus/Polk/Klipsch/Martin Logan tests in mono and stereo, and post the results for you to decide if there is "sufficient" correlation between results. I need to check that we have well-matched working stereo pairs for all models.
I don't see this thread progressing much further in the last few days, and I have moved my attention over to a new audio forum called "WhatsBestForum". They've given me my own front page forum called "Science in the Service of Art" (http://www.whatsbestforum.com/forumdisplay.php?158-Science-In-The-Service-Of-Art-Dr-Sean-Olive), and I seem to have my hands full clearing up some of misinformation being spread in the general forums about scientific methods and their application to sound reproduction.
Cheers
Sean
Audio Musings (http://seanolive.blogspot.com)
Well, the title of this thread it "Some New Evidence that Generation Y May Prefer Accurate Sound" and Sean posted details of a double blind listening test that shows precisely that: "Some Evidence". He didn't make any claim of proof and the thread title mirrors the correct conclusion from that test so far as I can see.
The arguments about whether he should or should not have used stereo speakers are utterly beside the point here. Whether or not the test would have been better or worse if stereo speakers had been used, the actual test as performed does indeed pretty much justify the claim of being "Some New Evidence that Generation Y May Prefer Accurate Sound".
Assuming for the sake of argument that stereo would have been better, so what? The test as done still shows "some evidence" perfectly clearly. The most you can claim is that it should have been "Some New Evidence that Generation Y May Prefer Accurate Sound when listening to mono speakers", but then so what? What evidence can you give that the conclusion or the evidence would have been more than slightly different if stereo speakers had been used? Certainly none is given here, only empty claims that this might have been so.
But none of this impeaches the conclusions drawn from the original study. There is evidence that at least some "gen Y" listeners prefer a more accurate sound under blind conditions. All the rest is just hot air and distraction, in my (not very humble) opinion.
My choice of source material would reflect my listening habbits. That would make the test relevant to me. Do you think using material I would never listen to would make the tests more relevant to me? I would think it would make them less relevant to me.
So yeah, blind and unbiased.
The choice of source material definately biases the outcome of any listening evaluation. One common hi fi salesman trick (been there, done that for about 7 years) is to find music that makes every speaker sound good, or that makes the speakers you want to sell sound good. Sometimes the most useful recording is the one that makes the high cost, high margin speakers sound great, but makes the low magin speakers sound bad. Often this recording sounds bad on the speakers you don't want to sell because they are more accurate and reveal the general failings of the recording.
lol. This reminds me of commercials and advertisements for HDTVs that show bright, beautiful images of marine life or something. How am I supposed to perceive what is being displayed on the actual TV there, when here I'm looking at the commercial on my on SDTV, or looking at the picture on my own computer monitor?
It's the same "trick" with a different sense, basically.
I think its time for the critics of single speaker testing to actually do something other than digging in their heels and repeatedly asserting that they and their anxieties and imaginations are right, and that people who have done technical tests of speakers and have developed speakers that have sold thousands, tens of thousands and even hundereds of thousands over a period decades are wrong simply because they say so.
Does that mean Bose has been right all these years?
Gary Eickmeier (whose Bose-based loudspeakers did very well in Clark's blind listening tests) has been saying as much for awhile now.
It seems to me an obvious conclusion. Hearing deteriorates over time. However, I wonder what will happen if you can get all the subjects to return when they've reached the age group they're being tested against... I suspect our world's become a lot more noisy and many people will have worse hearing at the same age as last generation. In part because people are stupid with earphones, but also because more people live in cities, where traffic has only gotten worse.
Oh, I guess I'm misunderstanding the experiment. Apparently there's a claim that the younger generation might prefer artifacts and scratches in their recordings... Who thinks this?
One Dr. Berger, a university music teacher whose results, derived from dubious methodology, were hyped in the audio/mainstream press awhile back. Please see Sean Olive's page, linked from the first post in this thread; it references the articles about Dr. Berger.
Well, the title of this thread it "Some New Evidence that Generation Y May Prefer Accurate Sound" and Sean posted details of a double blind listening test that shows precisely that: "Some Evidence". He didn't make any claim of proof and the thread title mirrors the correct conclusion from that test so far as I can see.
The arguments about whether he should or should not have used stereo speakers are utterly beside the point here. Whether or not the test would have been better or worse if stereo speakers had been used, the actual test as performed does indeed pretty much justify the claim of being "Some New Evidence that Generation Y May Prefer Accurate Sound".
Assuming for the sake of argument that stereo would have been better, so what? The test as done still shows "some evidence" perfectly clearly. The most you can claim is that it should have been "Some New Evidence that Generation Y May Prefer Accurate Sound when listening to mono speakers", but then so what? What evidence can you give that the conclusion or the evidence would have been more than slightly different if stereo speakers had been used? Certainly none is given here, only empty claims that this might have been so.
But none of this impeaches the conclusions drawn from the original study. There is evidence that at least some "gen Y" listeners prefer a more accurate sound under blind conditions. All the rest is just hot air and distraction, in my (not very humble) opinion.
Thanks Ed for clarifying that. I put some thought into the title carefully choosing my words so not to overstep what conclusions can be drawn from the limited data. The complete story has not yet been told.
It's funny and slightly ironic that my test, its methodology and conclusions are getting attacked by some people in this forum more so than Berger's who never publicly disclosed any of the information in his tests, and it was widely distributed by the press w/o much question. I guess it's human nature to shoot at something you can see and point to versus something that leaves us completely in the dark
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
yes it would be fair to compare the two speaker systems in their optimum envirement and set up. As a long time user of the Soundlab A3s I think I could do that but it would mean positioning the Soundlabs in a way I know from experience works and doing what it takes to get the room optimised for them. Are you suggesting it is less fair to have the Soundlabs perform at their best? I would think that would be more fair.
That's not fair - not a "level playing field". Given the purpose of the test and the constraints he had to work under, Sean's mono room setup was the "fairest" for all the speakers he used, including the MLs. The only thing I would have done differently would be to make sure the speakers used remained anonymous. All of this "unpleasantness" could have been avoided and we could have remained on topic.
For speaker reviews, it is common to optimise the positioning of the speakers in the room. It is not common to alter the room acoustics. Once learned, they become part of the reviewer's reference. It is common to use a wide range of musical styles known to be revealing of subtle differences, not just the reviewer's particular private listening favourites.
For auditioning for your own use, it is fair enough to alter the room acoustics and use "your own" music. You are then looking for the speakers that can be made to perform the best in your particular environment. Doing the "test" you suggested is just auditioning, and not a fair test when comparing against other speakers. There may well be speakers with better overall performance than the Soundlabs, but that would not perform as well in your particular environment.
So to be fair, you either set up every speaker under test in the environment that suits it best, or you use a single common environment set up to minimise environmental effects. Sean didn't have the time to optimise the performance of every speaker he used, so he took the other approach. For what it's worth, I believe his test setup was fairer to the MLs than the treatment they would normally receive in testing.
My choice of source material would reflect my listening habbits. That would make the test relevant to me. Do you think using material I would never listen to would make the tests more relevant to me? I would think it would make them less relevant to me.
It might be relevant to you, but it has no relevance to how well the Soundlabs actually perform compared to other brands.
So yeah, blind and unbiased.
You're half right.
Regards,
Don H.
The complete story has not yet been told.
Agreed
It's funny and slightly ironic that my test, its methodology and conclusions are getting attacked by some people in this forum
That's an odd generalized characterization. Can't speak for everyone it may have been aimed at, but I agree with probably 98% of what you do. I question the other 2%.
"Attack" seems a bit overly sensitive here. "Question" perhaps??
more so than Berger's
I hold you to a much higher standard that a "Berger". I could care less what he has to say. Your research is of much greater interest and relevance to me. Hence the focus....and questions .
(and hopefully we don't get one of Arnie's Born again, on the pulpit, "How dare ye asketh questions of the almighty..." type responses )
When I can find time, I will repeat the Infinity Primus/Polk/Klipsch/Martin Logan tests in mono and stereo, and post the results for you to decide if there is "sufficient" correlation between results. I need to check that we have well-matched working stereo pairs for all models.
I would be very grateful if you did. I would also be shocked if the Polk or Klipsch did any better than in mono. It's an apples to apples comparison with the 362, so I expect the correlation (to mono) to hold. My main interest would be the 362 vs ML. Remember the young folks rating discrepancy? I'm not as sure as you as to the reason. Will a similar HS group be involved, untrained (to have a strong preference for a very linear monopolar box speaker), or will it be only the "trained" listeners? Lastly, if you could, take some RTA FR's at the listening area (an avg would be fine) of both systems, with and without the opaque screen. (Maybe even see how they scored unsighted vs sighted? As that would add biases, but also remove the screen)
TIA,
cheers,
AJ
Btw, Earl Geddes has done some polar maps (http://www.gedlee.com/publish.htm) of the Orions used by David & Co. (as well as the B2031 and others).
Not sure if it's Murphy's law or otherwise, but there is a fairly large discrepancy in the FR response of that particular assembled system and what others have measured (http://theaudiocritic.com/plog/index.php?op=ViewArticle&articleId=33&blogId=1).
I'm really not sure of what to make of all that went on out there.
I'll withhold judgment for now and see what happens after the dust settles. Definitely have some....questions .
cheers,
AJ
yes it would be fair to compare the two speaker systems in their optimum envirement and set up. As a long time user of the Soundlab A3s I think I could do that but it would mean positioning the Soundlabs in a way I know from experience works and doing what it takes to get the room optimised for them. Are you suggesting it is less fair to have the Soundlabs perform at their best? I would think that would be more fair.
That's not fair - not a "level playing field". Given the purpose of the test and the constraints he had to work under, Sean's mono room setup was the "fairest" for all the speakers he used, including the MLs. The only thing I would have done differently would be to make sure the speakers used remained anonymous. All of this "unpleasantness" could have been avoided and we could have remained on topic.
For speaker reviews, it is common to optimise the positioning of the speakers in the room. It is not common to alter the room acoustics. Once learned, they become part of the reviewer's reference. It is common to use a wide range of musical styles known to be revealing of subtle differences, not just the reviewer's particular private listening favourites.
For auditioning for your own use, it is fair enough to alter the room acoustics and use "your own" music. You are then looking for the speakers that can be made to perform the best in your particular environment. Doing the "test" you suggested is just auditioning, and not a fair test when comparing against other speakers. There may well be speakers with better overall performance than the Soundlabs, but that would not perform as well in your particular environment.
So to be fair, you either set up every speaker under test in the environment that suits it best, or you use a single common environment set up to minimise environmental effects. Sean didn't have the time to optimise the performance of every speaker he used, so he took the other approach. For what it's worth, I believe his test setup was fairer to the MLs than the treatment they would normally receive in testing.
My choice of source material would reflect my listening habbits. That would make the test relevant to me. Do you think using material I would never listen to would make the tests more relevant to me? I would think it would make them less relevant to me.
It might be relevant to you, but it has no relevance to how well the Soundlabs actually perform compared to other brands.
So yeah, blind and unbiased.
You're half right.
Regards,
Don H.
So comparing the two speakers at their best, the way they would be used by me as a consumer is "not fair." Not sure where to go from there since I simply *completely* disagree. The fairness of any test is not a function of whatever constraints Sean has to work with. It really comes down to whether or not the speakers are being heard at their best. Anything less is unfair. That is especially true when the proctor has a potential conflict of interest as the representative of one of the speakers in the contest.
I'm not sure how common practice in reviews has any bearing on what makes for the best and most fair comparisons. Aren't most speaker reviews, no, all of them done under sighted conditions? Not really the standard we are looking for here is it? By the way I would have been giving up the aleged "advantage" of using my own room were I to hall my speakers over to HK. They all ready give up home field advantage because of that.
You say "So to be fair, you either set up every speaker under test in the environment that suits it best, or you use a single common environment set up to minimise environmental effects." Since different designers design and optimise for different envirements there is actually nothing fair about a common envirement. It is inherently unfair unless all speakers in question are optimised for that envirement. You either optimise the space for each speaker or you compramise that speaker's performance. That is a pretty solid either/or proposition. That Sean doesn't have time to optimise his tests in this way does not make the tests fair. Fairness is not a function of convenience.
I think the idea that my choice of source material would render any comparisons between the Soudlabs v. other brands completely irrelevant is..well...presumptuous to say the least. For all you know, among my choices are the very same recordings of classical guitar, harpsichord, clavichord referenced By Sean Olive as the most revealing sources.
It's funny and slightly ironic that my test, its methodology and conclusions are getting attacked by some people in this forum more so than Berger's who never publicly disclosed any of the information in his tests, and it was widely distributed by the press w/o much question. I guess it's human nature to shoot at something you can see and point to versus something that leaves us completely in the dark
I think part of the reason was because many people (me included) simply assumed there was some serious flaw with that study. Berger (and the details of the study) were most certainly not present in the way you are so it's hard to rigorously criticize a study when the comments are just getting sucked into the net. It's somewhat like criticizing the smelly air around a factory
Considering the amount of MP3 vs CD ABXs conducted on HA and on other forums, I think your findings have plenty of support from the evidence out there.
My apologies for not being more present in this discussion. I think some arguments have been valid (e.g. testing speakers out in the corner of the room) as far as exposing some limitations of an evaluations of speaker quality, but many others have just been the same "arguments" that are rehashed over and over on forums. I could have told you on June 22nd that somebody would eventually say they want you to test some specific speakers in some specific setup - much of which defeats the whole point of scientists trying to find generalizable and valid conclusions. I've grown very wary of responding to such arguments because a lot of time is wasted and the end result is typically the same.
I used to post reviews and results of my experiments on audio forums. If I made any strong conclusion (e.g. X is not worth it, an expensive product is not audibly different from a far cheaper one) many people would make all sorts of half-cooked arguments and basically refuse discussion otherwise. FEW were willing to conduct (or willing to publically admit their own results) their own experiments. Ultimately I'd waste a lot of time and the majority of people would consider me to be an both an idiot and deaf.
I wonder if kids today would take on a 2nd job just to purchase a $900 audio system? I already know the answer to that question.
Probably not since you can get a audio system (PMP+nice earphones) for under $100. Also, a music system is not that fashionable anymore. However, if a kid wants to buy an iphone...
I wonder if kids today would take on a 2nd job just to purchase a $900 audio system? I already know the answer to that question.
Probably not since you can get a audio system (PMP+nice earphones) for under $100. Also, a music system is not that fashionable anymore. However, if a kid wants to buy an iphone...
You're right. I just asked my 11-year son and 9-year old daughter what they would purchase if I gave them each $500.
Son: Iphone 4.0
Daughter: (1) $50 to charity with the rest spent on (2) Iphone 4.0
Note: They already have Ipods and headphones, and they associate good sound (my audio system when I play it) with "music that sucks"
Can't wait for the teenage years
Cheers
Sean Olive
Audio Musings (http://seanolive.blogspot.com)
Note: They already have Ipods and headphones, and they associate good sound (my audio system when I play it) with "music that sucks"
Have you tried playing their music on your system?
What happened?
It's funny and slightly ironic that my test, its methodology and conclusions are getting attacked by some people in this forum more so than Berger's who never publicly disclosed any of the information in his tests, and it was widely distributed by the press w/o much question. I guess it's human nature to shoot at something you can see and point to versus something that leaves us completely in the dark
Berger was trashed on HA at the time...
http://www.hydrogenaudio.org/forums/index....st&p=620628 (http://www.hydrogenaudio.org/forums/index.php?s=&showtopic=70272&view=findpost&p=620628)
...but maybe with less serious discussion because there was nothing to discuss - it was so far away from a proper test.
Whereas your work is exactly the kind of thing that many HA regulars would like to see/do, so of course there's intense interest in it, and lots of debate about incremental improvements.
Plus the poor results of certain beloved speakers have attracted the attention of a couple of people who (I suspect) do not buy into TOS 8 100%
Cheers,
David.
Since different designers design and optimise for different envirements there is actually nothing fair about a common envirement. It is inherently unfair unless all speakers in question are optimised for that envirement. You either optimise the space for each speaker or you compramise that speaker's performance. That is a pretty solid either/or proposition. That Sean doesn't have time to optimise his tests in this way does not make the tests fair. Fairness is not a function of convenience.
Do different loudspeaker companies or designers truly design and optimize their loudspeakers for different environments? Is this a true statement and how do we know what they are?
As I explained yesterday to someone over at WhatsBestForum, we design all of our loudspeakers to a design target that is based on anechoic measurements. The design target has no specific listening room in mind, because all listening rooms are different. It's not until the consumer has installed the speaker in their room that you can begin to adjust for the acoustical interaction between the loudspeaker and the room.
So it would be unwise to optimize the loudspeaker for a specific room. Instead they should design the speaker so it sounds good in as many rooms and positions as possible. This is done by optimizing the on and off-axis response so the direct and reflected sounds are neutral as you walk around the room. Then the problem comes down to mostly a low frequency one where the modal behavior of the room dominates the sound you hear.
We do validate the final design of our loudspeakers in a room but by then the design of the loudspeaker in 90% complete.
Cheers
Sean Olive
Audio Musings (http://htttp://seanolive.blogspot.com)
I mostly was interested in the reminiscing about one's first system, but I guess you could split it a lot of different ways.
Done (http://www.hydrogenaudio.org/forums/index.php?showtopic=81948).
Well, the title of this thread it "Some New Evidence that Generation Y May Prefer Accurate Sound" and Sean posted details of a double blind listening test that shows precisely that: "Some Evidence". He didn't make any claim of proof and the thread title mirrors the correct conclusion from that test so far as I can see.
The arguments about whether he should or should not have used stereo speakers are utterly beside the point here. Whether or not the test would have been better or worse if stereo speakers had been used, the actual test as performed does indeed pretty much justify the claim of being "Some New Evidence that Generation Y May Prefer Accurate Sound".
It's more than mono vs. stereo, concerns over placement and room treatment were expressed as well.
I think the pressing issue that started all this defensive posturing over electrostatic speakers lies in the choice of the word "accurate". I get the feeling that least some of the more overly-zealous fanatics don't like the insinuation that their highly prized speakers are inaccurate. I have to over the issues raised concerning scrim, however.
Do different loudspeaker companies or designers truly design and optimize their loudspeakers for different environments? Is this a true statement and how do we know what they are?
I think it's clear that some companies/designers optimize their speakers, deliberately or accidentally, for different amounts of direct vs. non-direct (I would not call it diffuse) sound.
The intent, etc, is another question.
I have to over the issues raised concerning scrim, however.
Could be be more specific?
I think the pressing issue that started all this defensive posturing over electrostatic speakers lies in the choice of the word "accurate". I get the feeling that least some of the more overly-zealous fanatics don't like the insinuation that their highly prized speakers are inaccurate.
From the pov of what is grandiosely called Discourse Analysis, it is clear that the leading terms of praise for music systems are "accurate" "precise" and so on, and that the favoured way of praising a high-end system is to say that it's like being in the, well, Living Presence ? of the music (despite the fact that this criterion is totally irrelevant for a lot of music). People who like a system, for whatever reason, are forced to claim that it is more realistic/accurate/lifelike; which is a pity, because there's no reason why you shouldn't prefer a system that colours things, just as Young People Nowadays like really accentuated bass (or are thought to do so by the advertisers). I think a decent system actually enables me to
hear the music better, in many ways, than I could at a concert, though I don't get to indulge my mild crush on the leading oboe player.
It's more than mono vs. stereo, concerns over placement and room treatment were expressed as well.
All of them irrelevant to the question of whether Sean's original post provided what it claimed to provide which, were it true (which it ain't IMO), might be at least a reasonable criticism.
I think the pressing issue that started all this defensive posturing over electrostatic speakers lies in the choice of the word "accurate". I get the feeling that least some of the more overly-zealous fanatics don't like the insinuation that their highly prized speakers are inaccurate. I have to over the issues raised concerning scrim, however.
Yes, I think that is very likely the real problem people had.
Since different designers design and optimise for different envirements there is actually nothing fair about a common envirement. It is inherently unfair unless all speakers in question are optimised for that envirement. You either optimise the space for each speaker or you compramise that speaker's performance. That is a pretty solid either/or proposition. That Sean doesn't have time to optimise his tests in this way does not make the tests fair. Fairness is not a function of convenience.
Do different loudspeaker companies or designers truly design and optimize their loudspeakers for different environments? Is this a true statement and how do we know what they are?
I think it is a true statement that other designers actually do listen to their products and evaluate them based on listening and then develope their designs further. Now, you need a room for that don't you? here is a little blurb on the ML evaluation room.
"First stop was the state of the art Martin Logan listening room. In this double-doored, tube trapped, fine-tuned room, products are evaluated for their sonic merits. The ancient ESL's you see on either sideof the reQuest ESL are there for training sessions also held in this room as well as around the facility. The room was built to reveal the slightest flaws and gains in performance of their products. They are true fanatics about the quality of their products from finish to function to final output!"
http://www.stereotimes.com/event020100.shtml (http://www.stereotimes.com/event020100.shtml)
As I explained yesterday to someone over at WhatsBestForum, we design all of our loudspeakers to a design target that is based on anechoic measurements. The design target has no specific listening room in mind, because all listening rooms are different. It's not until the consumer has installed the speaker in their room that you can begin to adjust for the acoustical interaction between the loudspeaker and the room.
Interesting. So would it be fair to say that the best room for your speakers would be an anechoic chamber? Where you do your measurements? Or the room you do the listening evaluations? You point it out yourself, every room is different so your speakers will sound different in every room no? What room would give us the very very best sound from your flagship speakers?
So it would be unwise to optimize the loudspeaker for a specific room. Instead they should design the speaker so it sounds good in as many rooms and positions as possible.
Now hold on here. I think you are making a hasty generalization. All purpose and high performance do not always go hand in hand. I don't see how in a stab at a SOTA speaker a designer may make choices that would force certain rooms or room treatments be used with them.
This is done by optimizing the on and off-axis response so the direct and reflected sounds are neutral as you walk around the room. Then the problem comes down to mostly a low frequency one where the modal behavior of the room dominates the sound you hear.
We do validate the final design of our loudspeakers in a room but by then the design of the loudspeaker in 90% complete.
Cheers
Sean Olive
Audio Musings (http://htttp://seanolive.blogspot.com)
That is fine Sean, that is your approach.
Well, the title of this thread it "Some New Evidence that Generation Y May Prefer Accurate Sound" and Sean posted details of a double blind listening test that shows precisely that: "Some Evidence". He didn't make any claim of proof and the thread title mirrors the correct conclusion from that test so far as I can see.
The arguments about whether he should or should not have used stereo speakers are utterly beside the point here. Whether or not the test would have been better or worse if stereo speakers had been used, the actual test as performed does indeed pretty much justify the claim of being "Some New Evidence that Generation Y May Prefer Accurate Sound".
It's more than mono vs. stereo, concerns over placement and room treatment were expressed as well.
I think the pressing issue that started all this defensive posturing over electrostatic speakers lies in the choice of the word "accurate". I get the feeling that least some of the more overly-zealous fanatics don't like the insinuation that their highly prized speakers are inaccurate. I have to over the issues raised concerning scrim, however.
I think I may actually be the only owner of electrostatic speakers involved in this thread. but i am not particularly concerned about whether or not other people call them "accurate." I think I have made my opinions clear before on all concerns for accuracy. Let's just say i'm not even considering the question when I form my preferences. *My* speakers were not measured by Sean. I have no idea how "accurate" they are by whatever measure one wants to use for speakers. And I don't care. No, it is really about the choice of doing the tests in mono when the speakers are used in stereo. I went on to bring up things like source material and room interactions since those are significant variables in the real world.
That is fine Sean, that is your approach.
Sure. When you get the opportunity to conduct your own series of tests you can come back and report, but don't get upset when your thread gets hijacked by people engaging in tangential discussions throwing stones.
I went on to bring up things like source material and room interactions since those are significant variables in the real world.
You've made your point numerous times and quite clearly. Now it's time to move along.
That is fine Sean, that is your approach.
Sure. When you get the opportunity to conduct your own series of tests you can come back and report, but don't get upset when your thread gets hijacked by people engaging in tangential discussions throwing stones.
I went on to bring up things like source material and room interactions since those are significant variables in the real world.
You've made your point numerous times and quite clearly. Now it's time to move along.
I'll consider conducting my own series of tests should I decide to go into the business of designing speakers.
I'm pretty sure there are some folks out there doing it differently than Sean and doing OK at it. That was the jist of my point. Sean was trying to say how *others* "should" be doing it.
If the tangent was so upsetting why let it go on all this time? Oh, maybe because it was actually interesting and some mods were actually contributing to the tangental part of the thread? But I know I know, it's my fault.
If the tangent was so upsetting why let it go on all this time? Oh, maybe because it was actually interesting and some mods were actually contributing to the tangental part of the thread? But I know I know, it's my fault.
I seriously hope you don't think we're perusing every post immediately after they happen.
But yeah, we let it go for a while. Now it's getting quite old watching the same point being argued over and over again. I don't think you're the only one feeding the cycle, but you are the one doing it at this present moment while I'm online.
Sean is going to get back with me over where he would like to see this post heading. Until then it will be closed for further discussion.