For a long time I've always thought that a good speaker could show limitations in amps and source components that a poor speaker would otherwise mask.
But I'm not sure of this line reasoning. Is it true that speakers distort sound more than a poor amp does? Would the sound not then be masked by the speakers own distortons?
Is it true that speakers distort sound more than a poor amp does?
Depends on your definition of "poor", but most modern amps produce orders of magnitude less distortion than speakers when not clipping.
Would the sound not then be masked by the speakers own distortons?
Yes.
...limitations in amps
With solid state electronics, an amplifier shouldn't have any audible defects. (
Maybe you'll hear some background noise.)
With modern electronics it's cheap & easy to build an amp with flat frequency response and very low distortion. I've NEVER heard ANY distortion from an amp that wasn't defective or being overdriven. Of course, you do have to pay more if you want a higher-power amplifier. (You can buy high-quality power amplifiers for less than one dollar per-watt.)
BEWARE of what you may read on "audiophile" websites! You may read that one amplifier sounds more "open" or "less veiled" than another, or some other NONSENSE that
seems to have a meaning, but really means nothing, or means different things to different people. They always use terminology that has no scientific or engineering meaning. They never mention the important "specs" (noise, frequency response, distortion) that are understood and measurable. In proper scientific-blind listening tests, all halfway decent amps will sound identical.
...and source components
The only "components" to worry about would be "old technology" analog turntables or cassette tape players, etc. You are more likely to have a poor quality
recording, such as a "home recording" made from the microphones built-into a laptop or iPhone, or again a cassette or vinyl recording, or a weak radio signal, etc.
...that a poor speaker would otherwise mask.
Sometime yes. For example, a cheap "full range" speaker without a tweeter will tend to filter-out high frequency noise such as tape hiss. (You can do the same thing by turning down the treble control.) A small speaker with poor bass response will tend to filter-out AC power-line hum. (You can do the same thing by turning down the bass.)
A poor quality speaker will generally
not filter-out MP3 compression artifacts. In fact, some people say that cheap headphones sometimes make MP3 artifacts easier to hear. A good quality MP3 can sound amazing with good speakers in a good room (with an amplifier powerful enough to get good volume). With a lower-bitrate MP3, you'd probably still prefer a good speaker.
Is it true that speakers distort sound more than a poor amp does?
Speakers are the weakest link and the most important component. If you go to an audio/video store and compare amplifiers, you shouldn't hear any difference (as long as they are properly level-matched* and as long as the amps are not over-driven into distortion). If you compare speakers, every speaker will sound different. Room acoustics also make a big difference, although a cheap speaker in a good room will rarely sound as good as a good speaker in a poor room.
* It's pretty-much impossible to do level-matched listening tests in a store, and the louder setup will generally "sound better", so it's easy to get fooled.
If the supposed problems with an amp or source component don't correspond with the distortion components of the speaker (or whatever), how are they going to be masked?
If the supposed problems with an amp or source component don't correspond with the distortion components of the speaker (or whatever), how are they going to be masked?
Basically what I am saying is that speakers with better resolution will reproduce the not so good sounding artifacts of a poor source where a speaker with poor resolution will not. Do you agree or disagree?
If the supposed problems with an amp or source component don't correspond with the distortion components of the speaker (or whatever), how are they going to be masked?
Basically what I am saying is that speakers with better resolution will reproduce the not so good sounding artifacts of a poor source where a speaker with poor resolution will not. Do you agree or disagree?
There is no simple answer to the question. For very poor speakers, a lot will be masked by their distortion. For a very bad amp (or one that is clipping), it will be obvious that the amp isn't working right.
You have to consider the specifics of your application. What are you trying to do?
If the supposed problems with an amp or source component don't correspond with the distortion components of the speaker (or whatever), how are they going to be masked?
Basically what I am saying is that speakers with better resolution will reproduce the not so good sounding artifacts of a poor source where a speaker with poor resolution will not. Do you agree or disagree?
There is no simple answer to the question. For very poor speakers, a lot will be masked by their distortion. For a very bad amp (or one that is clipping), it will be obvious that the amp isn't working right.
You have to consider the specifics of your application. What are you trying to do?
Trying to reproduce the musc so that the amp and source are not holding it back. I think a speaker of high resolution can potentally highlight flaws in the chain that a poor speaker with poor resolution might miss or gloss over, but then I'm asking for your input here.
Trying to reproduce the musc so that the amp and source are not holding it back.
So basically, you are considering moderate to high quality equipment that is being used correctly and within specification?
In that case, no, there will be essentially no masking of one component by another. Either an amp will be good enough or it won't be.
I think a speaker of high resolution can potentally highlight flaws in the chain that a poor speaker with poor resolution might miss or gloss over, but then I'm asking for your input here.
You have to be more specific. There is no universally valid answer that applies to all equipment. It depends on what you are testing. But for the kind of equipment people on this website would consider acceptable to use, you generally won't have to worry about this.
Trying to reproduce the musc so that the amp and source are not holding it back.
So basically, you are considering moderate to high quality equipment that is being used correctly and within specification?
In that case, no, there will be essentially no masking of one component by another. Either an amp will be good enough or it won't be.
I think a speaker of high resolution can potentally highlight flaws in the chain that a poor speaker with poor resolution might miss or gloss over, but then I'm asking for your input here.
You have to be more specific. There is no universally valid answer that applies to all equipment. It depends on what you are testing. But for the kind of equipment people on this website would consider acceptable to use, you generally won't have to worry about this.
Be more specific how?
Be more specific how?
What sort of equipment are considering, how will you be using it, and what is it you want to know about it, etc.
For a long time I've always thought that a good speaker could show limitations in amps and source components that a poor speaker would otherwise mask.
You get to be wrong! ;-)
But I'm not sure of this line reasoning. Is it true that speakers distort sound more than a poor amp does? Would the sound not then be masked by the speakers own distortons?
Right and right.
Here's the frequency response of a highly regarded loudspeaker
http://www.soundstagemagazine.com/measurem...m_studio100_v3/ (http://www.soundstagemagazine.com/measurements/paradigm_studio100_v3/)
(http://www.soundstagemagazine.com/measurements/paradigm_studio100_v3/frequency_on1530.gif)
Here are some distortion measurements at a fairly low listening level:
(http://www.soundstagemagazine.com/measurements/paradigm_studio100_v3/thd_95db.gif)
Show me an amp that measures that bad!
Show me an amp that measures only 10 times better!
For a long time I've always thought that a good speaker could show limitations in amps and source components that a poor speaker would otherwise mask.
You get to be wrong! ;-)
But I'm not sure of this line reasoning. Is it true that speakers distort sound more than a poor amp does? Would the sound not then be masked by the speakers own distortons?
Right and right.
Here's the frequency response of a highly regarded loudspeaker
http://www.soundstagemagazine.com/measurem...m_studio100_v3/ (http://www.soundstagemagazine.com/measurements/paradigm_studio100_v3/)
(http://www.soundstagemagazine.com/measurements/paradigm_studio100_v3/frequency_on1530.gif)
Here are some distortion measurements at a fairly low listening level:
(http://www.soundstagemagazine.com/measurements/paradigm_studio100_v3/thd_95db.gif)
Show me an amp that measures that bad!
Show me an amp that measures only 10 times better!
In percentage terms, how much distortion does a good speaker produce compared to a poor amplifier?
Right and right.
Here's the frequency response of a highly regarded loudspeaker
Why is that speaker considered to be highly regarded? According to whom? And I see it is like almost 10 years old.
What about a high-end speaker?
Here's the frequency response of a highly regarded loudspeaker
Why is that speaker considered to be highly regarded?
This is one of those questions that if I have to answer it, anybody who asks it probably can't grasp the answer.
I guess that there are people who never heard of Paradigm or think their stuff is all abjectly mid-fi. Fine!
I think that you are fighting a honest attempt at a solution. I tried! Good technical tests of speakers are like hen's teeth, so they don't exist for the vast majority of speakers. Opinions are like butt-holes, everybody has at least one.
According to whom? And I see it is like almost 10 years old.
The audiophile myth embedded in that statement seems to be that everything that is any good at all came out last month. The rate of real technological development in speakers is relatively slow.
The Paradigm Studio 100 is still a current model, and as a rule the differences between submodels of loudspeakers are small, often even just cosmetic.
What about a high-end speaker?
Please give me a generally agreed upon formal definition of
High End Loudspeaker and then post a link to a comparably detailed technical review.
According to whom? And I see it is like almost 10 years old.
The audiophile myth embedded in that statement seems to be that everything that is any good at all came out last month. The rate of real technological development in speakers is relatively slow.
If you read hi-fi magazines, this years model is often night-and-day amazingly better than last years model. The best just keeps getting better.
If you add together all the claimed improvements, you'd believe that the best from 20 years ago would sound like an Edison cylinder phonograph. Yet amazingly, most comparable components from 20 years ago sound pretty much like the ones available today. In some categories you can go back much further before you find any real audible inferiority.
Those hi-fi magazines couldn't be exaggerating the improvements to please their advertisers and reduce the desirability of used equipment, could they? It couldn't all just be some cynical psychological ploy designed to trick people into spending money chasing non-existent improvements, could it?
Cheers,
David.
In percentage terms, how much distortion does a good speaker produce compared to a poor amplifier?
http://www.sengpielaudio.com/calculator-thd.htm (http://www.sengpielaudio.com/calculator-thd.htm) so that you can swap between DB and % easily.
Those hi-fi magazines couldn't be exaggerating the improvements to please their advertisers and reduce the desirability of used equipment, could they? It couldn't all just be some cynical psychological ploy designed to trick people into spending money chasing non-existent improvements, could it?
Indeed, it's not like hi-fi manufacturers have to come up with reasons to make us buy their latest models each year, obviously they always sound super amazing
This is one of those questions that if I have to answer it, anybody who asks it probably can't grasp the answer.
I guess that there are people who never heard of Paradigm or think their stuff is all abjectly mid-fi. Fine!
What are you talking about???
I think that you are fighting a honest attempt at a solution. I tried! Good technical tests of speakers are like hen's teeth, so they don't exist for the vast majority of speakers.
You sent me a measurement. That was all you did. I asked you why Paradigm is so well regarded and you never gave me an answer. So????
Please give me a generally agreed upon formal definition of High End Loudspeaker and then post a link to a comparably detailed technical review.
Why are you asking me to define this? Don't high-end speakers exist in audio? Are you trying to say that all speakers the same?
Don't high-end speakers exist in audio?
What does "high-end" mean? That they are outrageously expensive? That some journalist has been paid to write that they are great?
The original question may be inherently assuming that an audio system should comprise a single massive amp driving a single multi-way speaker and passive crossover (x2 for stereo), but there are logical reasons why this is not the best solution. Those upstart companies that build active speakers with DSP filtering can give distortion measurements at the drivers that are comparable with amps, but they're measurably better in every other way, too.
By 35Hz it was thundering well at a high 99dB, while at the lower and more normal 86dB level distortion was fine, better than 1%. Up at 85Hz, 96dB sound level came in at an excellent 0.1% of second, 0.15% of third harmonic distortion. For 500Hz, 96dB we had 0.2 % second, 0.07% third. Distortion in loudspeakers can vary significantly with frequency, thus at a rather loud 96dB at 1kHz it gave 0.2% second, 0.5% of third, reducing quickly at lower powers; 96dB at 5kHz was fine, with 0.4% second, and 0.1% of third.
http://www.meridian-audio.info/public/meri...1%5B3099%5D.pdf (http://www.meridian-audio.info/public/meridian1%5B3099%5D.pdf)
This is one of those questions that if I have to answer it, anybody who asks it probably can't grasp the answer
I guess that there are people who never heard of Paradigm or think their stuff is all abjectly mid-fi. Fine!
What are you talking about???
Like I said the first time, if you don't know what I was talking about the first time, answering your question won't help you.
I think that you are fighting a honest attempt at a solution. I tried! Good technical tests of speakers are like hen's teeth, so they don't exist for the vast majority of speakers.
You sent me a measurement. That was all you did.
I've asked you to send me a measurement and you failed to do so. I guess that sending a measurement isn't as small of a thing that you seem to want to make it out to be.
I asked you why Paradigm is so well regarded and you never gave me an answer. So????
I gave you an answer, you just didn't like it. Besides that is a marketing question, not a technical question.
Please give me a generally agreed upon formal definition of High End Loudspeaker and then post a link to a comparably detailed technical review.
Why are you asking me to define this?
Because you seem to think that you know far better than I.
Don't high-end speakers exist in audio?
You mean you brought up the issue about something that you don't know even exists? That's a riddle, not a question.
Are you trying to say that all speakers the same?
I'm not so arrogant or condescending as to suggest that someone else would be so stupid or imperceptive as to believe that.
This is one of those questions that if I have to answer it, anybody who asks it probably can't grasp the answer
I guess that there are people who never heard of Paradigm or think their stuff is all abjectly mid-fi. Fine!
What are you talking about???
Like I said the first time, if you don't know what I was talking about the first time, answering your question won't help you.
So this is basically a riddle of your own making.
I asked you why Paradigm is so well regarded and you never gave me an answer. So????
I gave you an answer, you just didn't like it. Besides that is a marketing question, not a technical question.
Asking why THAT SPECIFIC speaker is so well regarded, is what I wanted you to explain and you failed to explain it. Maybe you like cat and mouse games but I don't.
Because you seem to think that you know far better than I.
Seem to? Where is all this speculation coming from? I don't think I know more than you. You are clearly taking things too personally.
Don't high-end speakers exist in audio?
You mean you brought up the issue about something that you don't know even exists? That's a riddle, not a question.
I wonder if you have problems having conversations with real-life human beings.
Are you trying to say that all speakers the same?
I'm not so arrogant or condescending as to suggest that someone else would be so stupid or imperceptive as to believe that.
So apparently there are no high-end speakers. Apparently the term high-end needs to be defined because a small group of people have been living under a rock. What speakers and system do you have? You seem like you have all the answers judging by your arrogant tone.
Why are you asking me to define this? Don't high-end speakers exist in audio? Are you trying to say that all speakers the same?
No, he's clearly asking you to post a link to measurements of what you consider to be a "high end speaker".
FWIW Stereophile publish rather good frequency measurements and waterfall plots on-line, but don't seem to publish useful distortion measurements for speakers.
Cheers,
David.
Why are you asking me to define this? Don't high-end speakers exist in audio? Are you trying to say that all speakers the same?
No, he's clearly asking you to post a link to measurements of what you consider to be a "high end speaker".
Thank you!
Since you were able to understand that, his insulting claim that I have problems having conversations with real-life human beings is in as much tatters as any of the rest of his claims.
FWIW Stereophile publish rather good frequency measurements and waterfall plots on-line, but don't seem to publish useful distortion measurements for speakers.
True and a pity, suggesting that Atkinson sees no value to any measurements of nonlinear distortion due to speakers. IOW he thinks that they all sound the same in this regard.
The measurements at the site I referenced suggest otherwise.
What about a high-end speaker?
Why are you asking me to define this?
Well, you used the term above.
I'd be curious to know your definition of the term you use as well.
cheers,
AJ
True and a pity, suggesting that Atkinson sees no value to any measurements of nonlinear distortion due to speakers. IOW he thinks that they all sound the same in this regard.
I know you have some beef with JA and I certainly don't always agree with him, but I think you've made quite a leap there. He may just think they are a bit hazily correlated to perception. I believe your friend Dr Geddes pays little mind to the speaker induced variety also, he considers the amplifier variety more insidious.
The measurements at the site I referenced suggest otherwise.
Soundstage? They (NRC) do measurements, but I'm not aware of any discussion related to audibility correlation. Link?
You seen those inexpensive Goldwood woofers JJ likes using? Rather pedestrian motor design, to be kind. No Faraday rings, etc. Certainly not "low distortion". May be a hint...
cheers,
AJ
The measurements at the site I referenced suggest otherwise.
Soundstage? They (NRC) do measurements, but I'm not aware of any discussion related to audibility correlation. Link?
The entirety of Toole/Olive's research program on measured loudspeaker performance vs preference involves NRC-type measurements, so I think that counts ;>
Them having worked at the NRC while doing the early part of that research too --
The entirety of Toole/Olive's research program on measured loudspeaker performance vs preference involves NRC-type measurements, so I think that counts ;>
Them having worked at the NRC while doing the early part of that research too --
Look at my responses to Arny. He/I referencing, specifically, "
distortion" measurements.
cheers,
AJ
True and a pity, suggesting that Atkinson sees no value to any measurements of nonlinear distortion due to speakers. IOW he thinks that they all sound the same in this regard.
I know you have some beef with JA and I certainly don't always agree with him, but I think you've made quite a leap there. He may just think they are a bit hazily correlated to perception. I believe your friend Dr Geddes pays little mind to the speaker induced variety also, he considers the amplifier variety more insidious.
That is incorrect to the point of being a complete inversion of what he writes. Geddes considers distortion of any kind in electronics to be a non problem and builds his systems accordingly. His own audio system is based on a mid-priced AVR and he advises others to do the same.
Geddes considers loudspeaker distortion to be the final frontier. The punchline to the Lee-Geddes papers about the audibility of nonlinear distortion is that since speakers and only speakers are going to be saddled with audible nonlinear distortion for the foreseeable future, shaping that distortion so that it has the least audible impact is of the essence. Hence the need for a better understanding of what is more audible and what is less audible.
With electronics, we can point at all of the equipment with all forms of distorton 80-100 dB down which means that even a flawed measure like THD suffices.
Speaker nonlinear distortion is often less than 20 dB down, with 40-50 dB under less stressful than actually used being common. I still remember one of my speaker-guru friends exclaiming that they found a driver with nonlinear distortion over a usable range that was > 60 dB down. That was and is exceptional performance.
http://www.data-bass.com/home (http://www.data-bass.com/home)
The measurements at the site I referenced suggest otherwise.
Soundstage? They (NRC) do measurements, but I'm not aware of any discussion related to audibility correlation. Link?
It is not necessarily the job of people doing measurements to explain every jot and tittle of what those measurements mean.
Absence of pontification is not evidence of absence of significance.
Please actually read Geddes before again misrepresenting him. Here's a link to full text - no excuses about having to pay for AES papers!
http://gedlee.com/distortion_perception.htm (http://gedlee.com/distortion_perception.htm)
You seen those inexpensive Goldwood woofers JJ likes using? Rather pedestrian motor design, to be kind. No Faraday rings, etc. Certainly not "low distortion". May be a hint...
That's JJ's buisness, not mine. Lots of speakers lack Faraday rings. Please show that their usage is an absolute requirement for low distortion in all speakers.
What about a high-end speaker?
Why are you asking me to define this?
Well, you used the term above.
I'd be curious to know your definition of the term you use as well.
I think we all, (including Rich B) know that there is no generally agreed-upon meaning to the phrase "High end speaker", so attacking someone's usage of the term is a cheap way to pick an argument without actually taking a position that could be discussed in a reasonable way. It seems very cowardly to me.
I picked one of the upper range of Paradigm's line because in a reasonable world (which high end audio is not) it would be considered to be a high end product. At ca. $2K a unit it costs about 10 times more than a mainstream speaker with similar function (and surprisingly similar performance). The company has a long favorable reputation originally based on the research of the NRC. Their equipment is reviewed in high end magazines. Obviously its not a $400,000 speaker or a $40,000 speaker, but it is a ca. $4,000 a pair speaker in a world where $400 a pair can get surprisingly comparable performance.
IMO high end audio at this time is partially based on keeping the wrong people from properly reviewing their overpriced products. This is easy enough to do - simply price their product so that no sane knowledgeable person would buy it, and send review copies only to people who are in on the scam.
If properly reviewed, everybody who gives any credibility to good technical tests should be asking "Why the outlandish price?". Reality is that high end speaker companies are almost all merely systems integrators. They produce no raw drivers of their own. They have no lock on some secret loudspeaker technology. They buy their drivers from the same production houses as everybody else can. A few have captive production facilities, but again they have no special magic that can't be bought on the open market. Their speakers are not technically exceptional in any way that they seem to be willing to actually demonstrate with anything from opinions from people who effectively shill for them. The outlandish pricing is based on what they can bully the market to pay through innuendo and outright false claims.
That is incorrect to the point of being a complete inversion of what he writes. Geddes considers distortion of any kind in electronics to be a non problem and builds his systems accordingly. His own audio system is based on a mid-priced AVR and he advises others to do the same.
Geddes considers loudspeaker distortion to be the final frontier.
http://www.diyaudio.com/forums/gedlee/1223...tml#post2682717 (http://www.diyaudio.com/forums/gedlee/122318-diy-waveguide-loudspeaker-kit-203.html#post2682717)
I'm not sure what you mean here? Do you mean crossover distortion in a split supply amplifier? If so then yes, this is the most insidious type of distortion. I am rather surprised that you are not familiar with our work in this area. You should read it as it will answer your questions about measureing it etc.
But amplifier distortion is beyond the scope of this thread since we are talking about loudspeakers. In loudspeakers there are no types of distortion that are significant - unless the speaker is broken. Amps can have highly audible forms of distortion, but thats a topic for an amplifier thread.
__________________
Earl Geddes
http://www.diyaudio.com/forums/multi-way/2...tml#post4034467 (http://www.diyaudio.com/forums/multi-way/260858-generic-how-big-mid-duty-driver-7.html#post4034467)
Not a trivial question at all, its right on the issue.
Let me make a few key points, some of which you have already made:
1) there is a threshold of nonlinearity below which we cannot detect it because of our ears inherent masking of such things.
2) this threshold can be vastly different for steady state signals versus transient ones. Steady state is far easier to detect.
3) (and here is the first key point) the numbers produced by THD or IMD do not correlate with the detectability of the nonlinearity in the underlying system. One form of nonlinearity, crossover distortion, is detectable at .1% THD while another, pure second order nonlinearity is not detectable even at 10%. Hence THD and IMD are useless at sorting out the "good" from the "bad". This is true for any signal steady state or transient.
4) it is the higher orders of nonlinearity that are audible, not the lower orders. Masking tells us why this is true.
5) the types of nonlinearity found in a loudspeaker are normally of the most benign type, low order. Electronics, the most insidious type, high order.
6) good loudspeakers are well below the thresholds of audibility for most signals, but certainly music, except when pushed beyond their design limits. A 15" woofer and a compression driver are never going to be pushed beyond there limits in a home situation.
There are metrics that do correlate with subjective assessment, but no one seems to want to adopt those - better to use pointless numbers than risk being shown to be at fault - being ignorant of the truth is better than being wrong.
An odd thing about the above is that it is very possible to make a loudspeaker with higher levels of THD that will actually sound better because the overall amount of higher order nonlinearities have been minimized. B&C actually does this, but they never even mention it. It is only apparent when one looks at the Klippel data. Their BL curves are deliberately NOT flat!! How's that for changing the paradigm of loudspeaker design!
I used to believe that nonlinearity was a key aspect of transducer design - my ears told me that. But the numbers don't lie and now I know that I was wrong all those years and wasted so much time chasing something that was unimportant.
__________________
Earl Geddes
It is not necessarily the job of people doing measurements to explain every jot and tittle of what those measurements mean.
Absence of pontification is not evidence of absence of significance.
A simple "ooops" from you would have sufficed, but alas....
Maybe all that dancing with Amir has taught you some moves.
That's JJ's buisness, not mine.
Indeed. I couldn't have said it better.
cheers,
AJ
http://www.diyaudio.com/forums/multi-way/2...tml#post4034467 (http://www.diyaudio.com/forums/multi-way/260858-generic-how-big-mid-duty-driver-7.html#post4034467)
Not a trivial question at all, its right on the issue.
Let me make a few key points, some of which you have already made:
1) There is a threshold of nonlinearity below which we cannot detect it because of our ears inherent masking of such things.
2) this threshold can be vastly different for steady state signals versus transient ones. Steady state is far easier to detect.
3) (and here is the first key point) the numbers produced by THD or IMD do not correlate with the detectability of the nonlinearity in the underlying system. One form of nonlinearity, crossover distortion, is detectable at .1% THD while another, pure second order nonlinearity is not detectable even at 10%. Hence THD and IMD are useless at sorting out the "good" from the "bad". This is true for any signal steady state or transient.
4) it is the higher orders of nonlinearity that are audible, not the lower orders. Masking tells us why this is true.
Earl usually talks in a structured way. The above, especially point #1 is a global qualifier for what follows. We all know that there are no good amplifiers, DACs or other modern electronics that have as much as 0.1% THD without being pushed into clipping, and there are none and have been none for decades that have measurable or audible crossover distortion. His mention of them is to make an important point which is that if nonlinear distortion is high enough to hear THD is not a diagnostic measure of nonlinear distortion.
5) the types of nonlinearity found in a loudspeaker are normally of the most benign type, low order. Electronics, the most insidious type, high order.
Earl's personal system, and his lab test setups have for years been centerpieced by mid-priced AVRs and cheap (ca. $100) amps that were not chosen based on any special sound quality properties. The last time I personally inspected such things at his house was about 6 weeks ago. This is not a guy who worries about the sound quality of good modern electronics no matter how that may be misrepresented by some.
6) good loudspeakers are well below the thresholds of audibility for most signals, but certainly music, except when pushed beyond their design limits. A 15" woofer and a compression driver are never going to be pushed beyond there limits in a home situation.
The relevant qualifier here is "good loudspeakers". Earl's ideas about what constitute good loudspeakers is easy to discern - it is what he sells which is high end both in terms of construction and pricing.
His top speakers run about $12K a pair if memory serves. There is the better part of $1K worth of drivers in each. He's not talking about all speakers used for audio or even all speakers that are used in mainstream audio by a long shot.
Earl's interest in subwoofers is fairly recent. Its wisely focused on smooth response in the upper bass, but not so much on exceptional dynamic range and low distortion below 30 Hz. That implies a history accepting either limited response below 30 Hz or a lot (audible) nonlinear distortion at serious listening levels in that range and below. My link to Data-bass should be more than enough evidence about that. There are a lot of people who disagree with the idea that performance below 30 Hz is irrelevant.
I have a foolproof method of choosing speakers, though I admit I've currently had the same pair for over twenty years. I take a range of source material with which I'm familiar, tell the shop assistant to set the volume at a relatively soft level, where I can judge the specific character of the speakers and listen, not just be hit by a wall of noise. The cabinet, cone materials and other factors will affect the character of the sound, and one man's speaker heaven is another man's hell. Technical specifications are all well and good, but your ears are the best judge of whether you can live with the specific character of a speaker.
I have a foolproof method of choosing speakers, though I admit I've currently had the same pair for over twenty years. I take a range of source material with which I'm familiar, tell the shop assistant to set the volume at a relatively soft level, where I can judge the specific character of the speakers and listen, not just be hit by a wall of noise. The cabinet, cone materials and other factors will affect the character of the sound, and one man's speaker heaven is another man's hell. Technical specifications are all well and good, but your ears are the best judge of whether you can live with the specific character of a speaker.
In store demos have to be as bad as a means for auditioning speakers as can be imagined.
For openers, speaker sound quality is profoundly affected by the acoustic environment, and no way is the dealer's demo room the area where I'm going to use the speaker.
Secondly, I generally listen to speakers for hours using dozens of recordings in a private setting before reaching any conclusions. What dealer would let me inflict that on his business?
Dealer demos have a lot of exceptionally powerful sources of bias including not only the physical environment but also the social and intellectual environment.
I buy new speakers not infrequently both for myself and others. I generally buy speakers based on technical performance and price and plan on using electrical and acoustical means to tailor their response to the actual application.
True and a pity, suggesting that Atkinson sees no value to any measurements of nonlinear distortion due to speakers. IOW he thinks that they all sound the same in this regard.
I know you have some beef with JA and I certainly don't always agree with him, but I think you've made quite a leap there.
Actually I have almost 2 decades of beefs with Atkinson. ;-)
He's staked his career on the infallibility of sighted evaluations and I've staked my public presence on the insanity of such thinking. He's staked his career on the idea that better measurements always imply better sound quality, and I've long taken the position that beyond certain often readily achievable levels, improvements in measured performance (particularly of electronics) just don't matter.
Interestingly enough, even here I have been facing arguments that seem to exactly match what I've imputed to Atkinson, which is that nonlinear distortion doesn't matter for speakers in general.
I don't know why someone would call foul on the position I've carved out for Atkinson while taking it on for themselves! ;-)
Earl usually talks in a structured way.
Spin and whirl any way you wish. I think any reasonable person can read what I wrote, you wrote and what Earl wrote, regarding loudspeaker distortion perception (as opposed to electronics) and figure out our stances on perceptual relevance with dynamic signals.
cheers,
AJ
Actually I have almost 2 decades of beefs with Atkinson. ;-)
I know.
I've called him out rather publicly myself, regarding controlled tests and "cognitive dissonance".
But I won't call him out for frivolous reasons as you did above. That isn't reasonable or objective.
Seems almost petty to me.
cheers,
AJ
Earl usually talks in a structured way.
Spin and whirl any way you wish. I think any reasonable person can read what I wrote, you wrote and what Earl wrote, regarding loudspeaker distortion perception (as opposed to electronics) and figure out our stances on perceptual relevance with dynamic signals.
Thanks for providing Earl's comments in context. I see that in that context Earls statements were commented on based on similar grounds to some of my comments.
If I excluded all speaker drivers but B&C's top products and avoided trying to reproduce bass below 30 Hz at decent SPLs I might be of a similar state of mind.
One irony is that certain unnamed persons will privately admit that they accept higher nonlinear distortion by operating B&Cs top drivers at relatively low crossover points... But of course it is never audible at the listening levels they consider to be reasonable... A certain amount of truth is dependent on making certain choices.
You can insultingly call my comments spinning and whirling and I can bite back if I wish to lower myself to use such tactics, which I won't.
Instead I will say that I think we have the same basic disagreements that we have entertained for years on at least 2 different forums, and you get to be wrong! ;-)
Re: Earl Geddes and the irrelevance of nonlinear distortion in loudspeakers, he cites Sean Olive's and Toole's views often, but I wonder if in this case he ascribes a view to them that isn't quite right.
On SO's own blog, this exchange (http://seanolive.blogspot.com/2013/04/the-relationship-between-perception-and.html?showComment=1367930646037#c6666389657949726751) occurred last year (i'm eliding some nonrelevant parts , adding emphasis and a link))
MM: [...] Kind of off-topic, you said above "From the perceptual data, a set of acoustical loudspeaker measurements has been identified from which we can model and predict listeners' loudspeaker preference ratings with about 86% accuracy". 86% is quite reasonable, but do you have any idea why your predictions aren't even more accurate? [...]
SO: Why can't we predict loudspeaker preference better than 86% accuracy? There are a number of reasons but here are two obvious ones that I pointed out in the original AES preprint (http://www.aes.org/e-lib/browse.cfm?elib=12847):
1) The prediction model only includes linear distortions in the loudspeaker measurements -- not nonlinear distortion
2) The accuracy of prediction is limited by limitations in the listening tests that created variance from context effects, elastic scale. The 86% accuracy was for the model based on 70 loudspeakers evaluated over 19-20 different listening tests. For the Consumer Reports tests, where we controlled context effects by comparing all combinations of loudspeakers the accuracy was almost 100%.
This suggests to me that SO consider the audible effects of at least some nonlinear distortions on loudspeaker preference to be both real and worthy of further investigation.
Actually I have almost 2 decades of beefs with Atkinson. ;-)
I know.
I've called him out rather publicly myself, regarding controlled tests and "cognitive dissonance".
But I won't call him out for frivolous reasons as you did above. That isn't reasonable or objective.
Seems almost petty to me.
What I see as petty is obfuscating the easy answers to a simple question about what has more nonlinear distortion, a CD player, an amplifier, or a loudspeaker.
It's a simple question with a simple answer that anybody who is truly knowledgeable about audio knows, but one that is widely obfuscated by the golden ear press.
it's even a question that someone who we both know and love may have stumbled over:
"One form of nonlinearity, crossover distortion, is detectable at .1% THD"
In fact crossover distortion is not detectable at 0.1% THD with normal musical recordings as the sound source.
I know of no proper DBTs ever being done to produce evidence about this claim, but if you download this file, you can use the FOOBAR2000 ABX tool to listen for yourself and present your results if you dare:
https://www.dropbox.com/s/9ebrlo639h99git/c...20joni.zip?dl=0 (https://www.dropbox.com/s/9ebrlo639h99git/crossover%20joni.zip?dl=0)
If the supposed problems with an amp or source component don't correspond with the distortion components of the speaker (or whatever), how are they going to be masked?
Masking does not require correspondence of distortion of distortion components, especially when the difference in levels of the various distortion components is as great as it typically is with amps and speakers.
Masking does not need to be invoked when the distortion components are below the thresholds of audibility, which is also true with most electronic equipment.
Masking is not the only source of the human inability to hear certain distortion components, such as those that are part and parcel of FM distortion (jitter).
Finally neither the amplifier nor the speaker are the only sources of masking in an audio system. Most musical program material is full of components that mask other sounds including some distortion components generated by the equipment.
The relevant qualifier here is "good loudspeakers". Earl's ideas about what constitute good loudspeakers is easy to discern - it is what he sells which is high end both in terms of construction and pricing.
Define "High End"?
If properly reviewed, everybody who gives any credibility to good technical tests should be asking "Why the outlandish price?". Reality is that high end speaker companies are almost all merely systems integrators. They produce no raw drivers of their own. They have no lock on some secret loudspeaker technology. They buy their drivers from the same production houses as everybody else can. A few have captive production facilities, but again they have no special magic that can't be bought on the open market. Their speakers are not technically exceptional in any way that they seem to be willing to actually demonstrate with anything from opinions from people who effectively shill for them. The outlandish pricing is based on what they can bully the market to pay through innuendo and outright false claims.
You seem to be demonizing companies that sell expensive speakers. What false claims are you talking about? What innuendo? Are you sure you aren't just paranoid about the high-end industry?
I get frustrated with these discussions that consist only of referencing 'Higher Authorities'.
In fact crossover distortion is not detectable at 0.1% THD with normal musical recordings as the sound source.
A fatal flaw, it seems to me is the question of what constitutes an audiophile's idea of "musical". My experience is that the audiophile taste in music is, shall we say, very 'conservative'. If all these Higher Authorities' pronouncements on audibility are based on their own idea of 'music' (such as they choose to subject their listening panels to in their listening tests, for example) then they can be taken with a pinch of salt. Why should I believe their pronouncements on audibility if my taste in music is so much more dynamic? :-)
[At this moment I am listening to some Britten. There is no way that if I were to play this in a dealers at any volume level, that it could be described as a "wall of noise", which tells me something about how other people judge a system, and what music they choose to listen to. I can only judge a system by turning the volume up to 'realistic', and I suspect that the Higher Authorities simply don't share my taste in music, and therefore couldn't hear what it is that I am listening for. They may pronounce that "low order distortion is inaudible" or "phase isn't important" but that may just be based on their taste in light jazz from the 1960s].
Define "High End"?
Rich, you were supposed to tell us what a "high end" speaker is! Help me out man, I did my best with the whole FR thing for you, which you hopefully now understand.
Would a Revel Salon qualify?
I get frustrated with these discussions that consist only of referencing 'Higher Authorities'.
Agnostic?
You mean the guys with all the AES peer reviewed work, making statements about that work, often verified over and over. Them??
Like I said early in the show, if you have some work, or links to works, overturning the established science that has misled these authorities, by all means present it! Always good to learn something new GM.
cheers,
AJ
This suggests to me that SO consider the audible effects of at least some nonlinear distortions on loudspeaker preference to be both real and worthy of further investigation.
Yes, he did say that, along with this: http://seanolive.blogspot.com/2009/01/what...ations-are.html (http://seanolive.blogspot.com/2009/01/what-loudspeaker-specifications-are.html)
1) The perception of loudspeaker sound quality is dominated by linear distortions, which can be accurately quantified and predicted using a set of comprehensive anechoic frequency response measurements.
2) Both trained and untrained listeners tend to prefer the most accurate loudspeakers when measured under controlled double-blind listening conditions.
3) The relationship between perception and measurement of nonlinear distortions is less well understood and needs further research. Popular specifications like Total Harmonic Distortion (THD) and Intermodulation Distortion (IM) do not accurately reflect the distortion’s audibility and effect on the perceived sound quality of the loudspeaker.
I don't think there was any misinterpretation by Earl, who was there...and has done his own investigations.
But hey, both SO and jj were at one time members here. Perhaps Arny could ask them directly, to avoid any risk of "inversion" (despite verbatim quotation).
cheers,
AJ
You mean the guys with all the AES peer reviewed work, making statements about that work, often verified over and over. Them??
Like I said early in the show, if you have some work, or links to works, overturning the established science that has misled these authorities, by all means present it! Always good to learn something new GM.
Indeed I do. Because once we're into declaring what distortion is audible with 'normal musical signals' then it's pure subjectivity despite the trappings of objectivity. Coming up with some tracks to listen to and listing them in our peer-reviewed paper doesn't make it science. Who chose these tracks and why?
Define "High End"?
Rich, you were supposed to tell us what a "high end" speaker is! Help me out man, I did my best with the whole FR thing for you, which you hopefully now understand.
Would a Revel Salon qualify?
I just find it funny that I'm asked to define these terms and then Arnold goes and throws the term around without defining them. No one complains when he does it! Some people have double standards.
This suggests to me that SO consider the audible effects of at least some nonlinear distortions on loudspeaker preference to be both real and worthy of further investigation.
Yes, he did say that, along with this: http://seanolive.blogspot.com/2009/01/what...ations-are.html (http://seanolive.blogspot.com/2009/01/what-loudspeaker-specifications-are.html)
1) The perception of loudspeaker sound quality is dominated by linear distortions, which can be accurately quantified and predicted using a set of comprehensive anechoic frequency response measurements.
2) Both trained and untrained listeners tend to prefer the most accurate loudspeakers when measured under controlled double-blind listening conditions.
3) The relationship between perception and measurement of nonlinear distortions is less well understood and needs further research. Popular specifications like Total Harmonic Distortion (THD) and Intermodulation Distortion (IM) do not accurately reflect the distortion’s audibility and effect on the perceived sound quality of the loudspeaker.
I don't think there was any misinterpretation by Earl, who was there...and has done his own investigations.
But hey, both SO and jj were at one time members here. Perhaps Arny could ask them directly, to avoid any risk of "inversion" (despite verbatim quotation).
cheers,
AJ
There's really no contradiction there between SO's two quotes. The SO quote I pulled didn't address the relative contributions, nor imply that nonlinear distortions were
just as important as linear. And he says here exactly what I surmised: needs more research. Olive attributes at least part of the failure of his preference model - -- which fails in 24% of cases -- to the effects of nonlinear distortion. So I'd say he considers it
potentially significant until further research shows otherwise.
And I'm not understanding the second part or what you wrote --Geddes was there? *Where*? With Olive when he did the research described in the 2004 AES preprint I linked to? I hadn't heard that. I know he's done his own testing. The point was, Geddes claims that nonlinear distortion typically has *no* importance, and he cites Olive as backup. If he only means typical THD and IM specs, that's not quite the same claim. And I don't think Sean would phrase it as strongly as Earl does..
The relevant qualifier here is "good loudspeakers". Earl's ideas about what constitute good loudspeakers is easy to discern - it is what he sells which is high end both in terms of construction and pricing.
Define "High End"?
I see two high ends in audio. One actually makes some sense - it is composed of very high performing stuff.
The other high end is just plain whatever the traffic provided by easily bullied, poorly informed people with more money than brains waste their obviously too easily obtained money on in order to get bragging rights.
Paradigm - high performing high end. Wilson Audio - for the bragging rights market.
You seem to be demonizing companies that sell expensive speakers. What false claims are you talking about? What innuendo? Are you sure you aren't just paranoid about the high-end industry?
You sure you aren't suckered by it? Would you say it's wrong to 'demonize' a >$10K loudspeaker that boasts of its performance/perfection, yet 'measures' abysmally?
If properly reviewed, everybody who gives any credibility to good technical tests should be asking "Why the outlandish price?". Reality is that high end speaker companies are almost all merely systems integrators. They produce no raw drivers of their own. They have no lock on some secret loudspeaker technology. They buy their drivers from the same production houses as everybody else can. A few have captive production facilities, but again they have no special magic that can't be bought on the open market. Their speakers are not technically exceptional in any way that they seem to be willing to actually demonstrate with anything from opinions from people who effectively shill for them. The outlandish pricing is based on what they can bully the market to pay through innuendo and outright false claims.
You seem to be demonizing companies that sell expensive speakers.
A speaker with ultra-low distortion and ultra-high dynamic range have an inherent right to sell expensive speakers. One of the problems with people like Atkinson is that their technical tests go out of their way to provide zero evidence about either of those issues.
What false claims are you talking about?
False claims to above average sound quality.
What innuendo?
The innuendo that they aren't gouging their customers.
Are you sure you aren't just paranoid about the high-end industry?
IMO there are two high end audio industries - the people who are actually trying to provide more value for more money, and the ones operating carnival side slows.
Obviously a guy who is trying to produce 130 dB SPL at < 10% THD is going to have need average pricing to have a business. But people like Sonus Faber that mark up their parts and production costs into the next universe are just playing mind games with their customer's money.
[At this moment I am listening to some Britten. There is no way that if I were to play this in a dealers at any volume level, that it could be described as a "wall of noise", which tells me something about how other people judge a system, and what music they choose to listen to. I can only judge a system by turning the volume up to 'realistic', and I suspect that the Higher Authorities simply don't share my taste in music, and therefore couldn't hear what it is that I am listening for. They may pronounce that "low order distortion is inaudible" or "phase isn't important" but that may just be based on their taste in light jazz from the 1960s].
Well the question to you then is , how do you know that what *is* important to you, is due to the level of 'low order distortion' or 'phase'?
Really, where the Higher Authorities who do *research* have the advantage is, they can describe the correlations between actual measured properties and actual listener responses.
You, and most of us, are typically just flailing about, trying to describe effects whose causes we don't know or just guess at... some of which effects may even be imaginary. At the extreme we have the 'high end' reviewers who tend towards the flowery prose of wine tasting. When challenged to explain what they mean, we get bollocks like this:
Pace, Rhythm, & Dynamics
http://www.stereophile.com/reference/23/index.html (http://www.stereophile.com/reference/23/index.html)
and a mishmash of over-specific and under-specified, like this
Sounds Like? An Audio Glossary
http://www.stereophile.com/content/sounds-...ossary-glossary (http://www.stereophile.com/content/sounds-audio-glossary-glossary)
(J Gordon Holt, to his credit, went out in a blaze of glory, scolding the 'audiophile' press for eschewing bias-controlled comparison)
This suggests to me that SO consider the audible effects of at least some nonlinear distortions on loudspeaker preference to be both real and worthy of further investigation.
Yes, he did say that, along with this: http://seanolive.blogspot.com/2009/01/what...ations-are.html (http://seanolive.blogspot.com/2009/01/what-loudspeaker-specifications-are.html)
1) The perception of loudspeaker sound quality is dominated by linear distortions, which can be accurately quantified and predicted using a set of comprehensive anechoic frequency response measurements.
2) Both trained and untrained listeners tend to prefer the most accurate loudspeakers when measured under controlled double-blind listening conditions.
3) The relationship between perception and measurement of nonlinear distortions is less well understood and needs further research. Popular specifications like Total Harmonic Distortion (THD) and Intermodulation Distortion (IM) do not accurately reflect the distortion’s audibility and effect on the perceived sound quality of the loudspeaker.
I don't think there was any misinterpretation by Earl, who was there...and has done his own investigations.
But hey, both SO and jj were at one time members here. Perhaps Arny could ask them directly, to avoid any risk of "inversion" (despite verbatim quotation).
This statement: The perception of loudspeaker sound quality is dominated by linear distortions, which can be accurately quantified and predicted using a set of comprehensive anechoic frequency response measurements (see my previous posting here)
Is true but does not exclude the influence of audible nonlinear distortion. Linear distortion has similar effects at any listening level. Nonlinear distortions of the most common kinds are by definition highly level dependent.
There's a door that people like Rich B could come in through if they could master the vocabulary, which is to say that just because a speaker measures well exclusive of nonlinear distortion doesn't mean that their actual audible nonlinear distortion doesn't affect listening enjoyment. Arguing that nonlinear distortion is never audible isn't realistic in a world that has 3.5" subwoofers that are speced to have response down to 25-30 Hz and also specs peak listening levels at 105 dB SPL.
Really, where the Higher Authorities who do *research* have the advantage is, they can describe the correlations between actual measured properties and actual listener responses.
This is partly true, but I think my question is reasonable: how do we prove that the research isn't 'circular' ie. that the choice of the listening material and 'subjects' is always going to be subjective.
You, and most of us, are typically just flailing about, trying to describe effects whose causes we don't know or just guess at... some of which effects may even be imaginary. At the extreme we have the 'high end' reviewers who tend towards the flowery prose of wine tasting.
But my approach is not to believe the Higher Authorities, or the reviewers. Nor is it to pay attention only to 'measurements'. My approach would be to start with a clean slate: what system is likely to give the most accurate reproduction? Researching existing technology using listening panels and so on is still "flailing about" as I see it. So I would dismiss vinyl and tape out of hand, regardless of the reactions some Higher Authority found it evoked in their choice of listeners listening to someone's choice of music. Digital is better by design, and I don't need to hear the alternatives. My attitude to speakers is the same: DSP active. No need to test them against anything else. They are simply more accurate.
I see two high ends in audio. One actually makes some sense - it is composed of very high performing stuff.
That sounds very vague. How do you know when high-end is actually high-end?
The other high end is just plain whatever the traffic provided by easily bullied, poorly informed people with more money than brains waste their obviously too easily obtained money on in order to get bragging rights.
So high-end is, on one hand, composed of high performing stuff, and on the other hand, high-end is just expensive. So you think a high price is somehow bullying people and means that people are poorly informed. Just because you can't afford the high-end doesn't make it a waste. What system do YOU have?
Paradigm - high performing high end. Wilson Audio - for the bragging rights market.
Based on what????? You haven't shown that Paradigm is even high performing. You just seem highly prejudiced against high-end brands and have clear biases towards Paradigm for some reason!
Really, where the Higher Authorities who do *research* have the advantage is, they can describe the correlations between actual measured properties and actual listener responses.
This is partly true, but I think my question is reasonable: how do we prove that the research isn't 'circular' ie. that the choice of the listening material and 'subjects' is always going to be subjective.
Is that your conclusion after having read the methods sections of the research papers ?
You, and most of us, are typically just flailing about, trying to describe effects whose causes we don't know or just guess at... some of which effects may even be imaginary. At the extreme we have the 'high end' reviewers who tend towards the flowery prose of wine tasting.
But my approach is not to believe the Higher Authorities, or the reviewers. Nor is it to pay attention only to 'measurements'. My approach would be to start with a clean slate: what system is likely to give the most accurate reproduction? Researching existing technology using listening panels and so on is still "flailing about" as I see it. So I would dismiss vinyl and tape out of hand, regardless of the reactions some Higher Authority found it evoked in their choice of listeners listening to someone's choice of music. Digital is better by design, and I don't need to hear the alternatives. My attitude to speakers is the same: DSP active. No need to test them against anything else. They are simply more accurate.
When you prioritize ' the most accurate reproduction' you are prioritizing *measurements*, whether you realize it or not. And the 'higher authorities' that made them. So who, exactly, are you rebelling against?
Define "High End"?
Rich, you were supposed to tell us what a "high end" speaker is! Help me out man, I did my best with the whole FR thing for you, which you hopefully now understand.
Would a Revel Salon qualify?
High end : exclusive, expensive, superior construction, better finishes, more sophisticated drivers, extended frequency extension.
Is that your conclusion after having read the methods sections of the research papers ?
My attitude is that simply having, and stating a methodology doesn't automatically make it science, or make it objective. If we are talking about listening to music, then it is automatically subjective to some extent. If a research paper states "A selection of typical music tracks was chosen at random", or "The listeners were allowed to make their own music selections" it doesn't make it science. I was just reading a piece by Higher Authority Sean Olive about preferring the use of 'trained' listeners for audio evaluation. His organisation trains its listeners using a special program of sessions where listeners learn to identify characteristics of processed "music". It seems to me, there are questions prompted by this: no matter how scientific it appears, there is an element of circularity. (As an example)
When you prioritize ' the most accurate reproduction' you are prioritizing *measurements*, whether you realize it or not. And the 'higher authorities' that made them. So who, exactly, are you rebelling against?
But even if I was prioritising "warm fuzzy tone" it could still be expressed as measurements, so whatever I prioritise, it can always be interpreted as measurements of some form. I didn't realise that we had to thank the Higher Authorities for sanctioning 'accuracy' for our use!
Edit: it's like the endless arguments about which is better: MP3 or AAC? etc. etc. Going on about the science of perceptual coding, masking etc. The answer is: you're an audiophile. You don't have to listen to any of them! I feel the same about discussions over whether it's best to use a second order crossover and reverse the tweeter, or blah blah blah. Or use this oil on our bearings. Or how best to clean an LP. What do we do when tape sheds oxide? The answer is, we don't need to worry about any of it! Science and technology have come to our aid, and we don't need to prove the new technology is better with subjective listening tests. We can just examine how it works, and confirm its superiority by looking at a few measurements done by somebody else, and then concentrate on the really interesting stuff.
My attitude is that simply having, and stating a methodology doesn't automatically make it science, or make it objective.
What it does is to make it possible for someone else to repeat the same experiment and either verify or refute the results (scientific method 101).
My attitude is that simply having, and stating a methodology doesn't automatically make it science, or make it objective.
What it does is to make it possible for someone else to repeat the same experiment and either verify or refute the results (scientific method 101).
IMO there's a justified bit of skepticism whether "listener's preference" can be reliable. Another method would be to compare a speaker to a reference sound (another speaker or sound source), which would come closer to our beloved ABX testing. I also wonder if preference changes a lot over time (years, decades, centuries, like fashion).
My attitude is that simply having, and stating a methodology doesn't automatically make it science, or make it objective.
What it does is to make it possible for someone else to repeat the same experiment and either verify or refute the results (scientific method 101).
You miss my point: when it comes to questions of whether it is possible to hear certain types of distortion when reproducing musical signals, the choice of which musical signals we use may affect the outcome of the experiment. Repeating the same flawed experiment and getting the same result doesn't mean that the results are then 'scientific' and universally applicable to all listeners with all types of music. People who believe they understand 'scientific method 101' are the ones who often confuse cause with correlation, and so on, and can't spot the subjectivity in a part of an experiment. If the 'audio scientists' and 'peers' reviewing their work, all have standard audiophile musical taste, then I can see how they could all conclude that if distortion X cannot be picked up by listeners against a wide variety [of mono, light jazz-based] music [played at a moderate volume on a passive speaker system] then it can be declared benign. Or rather that "their findings suggest that relatively high levels of distortion X are completely inaudible in musical signals". Which we can then parrot as a copper-bottomed universally-applicable scientific truth in this forum ad nauseam!
Or rather that "their findings suggest that relatively high levels of distortion X are completely inaudible in musical signals".
There are a few interesting online listening tests on the Klippel website (http://www.klippel-listeningtest.de/lt/), experts in loudspeaker (unit) measurements. The results seem to indicate that (anonymous?) listeners can hear loudspeaker distortion far below real speaker levels. The tests use artificially created distortion at various levels to simulate real speaker behavior. This method sounds plausible, but I don't know if it's widely accepted.
(http://www.klippel-listeningtest.de/lt/img/example_result.jpg)
btw, I've always wondered about the high score at the lowest distortion, to the right. Golden ears, cheating ?
btw, I've always wondered about the high score at the lowest distortion, to the right. Golden ears, cheating ?
It could be that a small number of people repeat the test over and over and home in on the distortion..? I found that my 'performance' went up once I realised I could press 'play' for each sample without having to go via 'stop', meaning I could compare the first couple of seconds immediately without a gap. That way it was quicker and avoided 'fatigue' - which I was definitely beginning to feel. But then, the test for me became that of identifying distortion within a two second sample of a particular song in mono at moderate volume using the cheapo headphones at hand - and may have ended up being that for most people. Could the results of such an experiment then be extrapolated to conclude that distortion type X is audible or inaudible at Y% in "music"? I imagine it could be! But whether that would be justified would be a different matter.
Is that your conclusion after having read the methods sections of the research papers ?
My attitude is that simply having, and stating a methodology doesn't automatically make it science, or make it objective. If we are talking about listening to music, then it is automatically subjective to some extent. If a research paper states "A selection of typical music tracks was chosen at random", or "The listeners were allowed to make their own music selections" it doesn't make it science. I was just reading a piece by Higher Authority Sean Olive about preferring the use of 'trained' listeners for audio evaluation. His organisation trains its listeners using a special program of sessions where listeners learn to identify characteristics of processed "music". It seems to me, there are questions prompted by this: no matter how scientific it appears, there is an element of circularity. (As an example)
I see a big misunderstanding about the nature of scientific investigations, especially in the area of relevance.
If something is investigated and all possible controls are thrown to the wind (e.g. your typical audiophile so-called listening test) then there's not a chance of it being scientific. The so-called data is actually noise. Flee the scene and save your sanity!
If something is investigated and reasonable controls are used and documented then it is scientific, but it still needs to be investigated further for its relevance and applicability to a particular question of interest. A positive outcome for that evaluation is not necessarily a slam dunk. The failure of a particular investigation to be applicable to every question that may exists in someone's mind does not necessarily detract from its basic scientific nature.
Human bias is impossible to totally remove from any non-trivial investigation, but knowing what those biases are can be very helpful.
You miss my point: when it comes to questions of whether it is possible to hear certain types of distortion when reproducing musical signals, the choice of which musical signals we use may affect the outcome of the experiment.
The above completely understates the true situation. The observation is in fact a truism. It has been known for decades that the choice of which musical signals we use profoundly affects the outcome of the experiment. Conversations about choice of musical signals fill the annals of subjective audio tests.
The fact that this is a point that someone seems to want to argue this only shows a lack of proper background in doing subjective experiments related to audio.
Repeating the same flawed experiment and getting the same result...
The error here is presuming that the fact that the choice of which musical signals we use may affect the outcome of the experiment represents some kind of a serious flaw. It is just another influence that needs to be managed. Examples of effective management of this kind of influences are in the literature of audio science including HA.
I see a big misunderstanding about the nature of scientific investigations, especially in the area of relevance.
If something is investigated and all possible controls are thrown to the wind (e.g. your typical audiophile so-called listening test) then there's not a chance of it being scientific. The so-called data is actually noise. Flee the scene and save your sanity!
If something is investigated and reasonable controls are used and documented then it is scientific, but it still needs to be investigated further for its relevance and applicability to a particular question of interest. A positive outcome for that evaluation is not necessarily a slam dunk. The failure of a particular investigation to be applicable to every question that may exists in someone's mind does not necessarily detract from its basic scientific nature.
Human bias is impossible to totally remove from any non-trivial investigation, but knowing what those biases are can be very helpful.
No, I don't think I misunderstand. Whether or not the Higher Authorities (and we) are well aware of the non-slam dunk-ness of the work, the people who quote their 'findings' are often not.
No, I don't think I misunderstand. Whether or not the Higher Authorities (and we) are well aware of the non-slam dunk-ness of the work, the people who quote their 'findings' are often not.
Since 99.9% or more of the so-called findings are based on totally uncontrolled sighted evaluations, the people proffering them obviously don't understand how irrelevant their so-called findings are to any reasonable question about audio. They obviously think they are acting in good faith. They even get a little testy when the true relevant facts are pointed out to them. ;-)
When did you discover this?
I've known it for about 40 years.
Can we move on from obsessing over truisms?
The observation is in fact a truism.
Well at least I've said something we can all agree on!
The fact that this is a point that someone seems to want to argue this only shows a lack of proper background in doing subjective experiments related to audio.
What I am wanting to argue on, is not the validity of subjective audio experiments, but the fact that some of them are deemed necessary at all. "Should we add distortion to audio signals deliberately? Let's do an experiment!". "Should we place stones at the corners of our amplifiers? Let's do an experiment!". Both are 'science' if we say they are. But deciding on whether to take any notice of them can be done without wading through the pages of the dubious 'science' and statistics that they would produce.
The error here is presuming that the fact that the choice of which musical signals we use may affect the outcome of the experiment represents some kind of a serious flaw. It is just another influence that needs to be managed. Examples of effective management of this kind of influences are in the literature of audio science including HA.
You're implicitly assuming that what goes on in audiophile trials is 'science'. In a way the definition is no better than the definition of art. "It is what we say it is". If you pick holes in our experiment then that is merely science in action. And if we then pick holes in your experiment then that, too, is also science in action. Everything is science as long as we say it is (and define our assumptions, describe our method blah blah).
Who can say that an "influence has been managed effectively"? That sounds very subjective!
Can we move on from obsessing over truisms?
Sorry. This branch of the discussion stemmed from my earlier objections to discussions that do nothing but quote from Higher Authorities - which are, in themselves, nothing but a form of endless regurgitation of truisms.
True and a pity, suggesting that Atkinson sees no value to any measurements of nonlinear distortion due to speakers. IOW he thinks that they all sound the same in this regard.
The measurements at the site I referenced suggest otherwise.
Rather than continue to play Arnieball, which a lot like Amirball aka Calvinball, why don't you just spell out what relevance the Soundstage NRC anechoic THD steady state measurements (you posted above) have perceptually?
Maybe once we establish that, we can see why you're bashing JA for
not performing them...with his quasi-anechoic measurement setup.
"One form of nonlinearity, crossover distortion, is detectable at .1% THD"
In fact crossover distortion is not detectable at 0.1% THD with normal musical recordings as the sound source.
I know of no proper DBTs ever being done to produce evidence about this claim, but if you download this file, you can use the FOOBAR2000 ABX tool to listen for yourself and present your results if you dare:
https://www.dropbox.com/s/9ebrlo639h99git/c...20joni.zip?dl=0 (https://www.dropbox.com/s/9ebrlo639h99git/crossover%20joni.zip?dl=0)
I have a better idea Amir, excuse me...Arnie. Why don't
you have your friend Earl take the test and verify
his threshold claim. You know, the guy whose position I had inverted, not Arnie, in Arnieworld.
cheers,
AJ
And I'm not understanding the second part or what you wrote --Geddes was there? *Where*?
C'mon now, the link I provided!
Here again: http://seanolive.blogspot.com/2009/01/what...ations-are.html (http://seanolive.blogspot.com/2009/01/what-loudspeaker-specifications-are.html)
The panelists included myself, Steve Temme (Listen Inc.), Dr. Earl Geddes (GedLee), Laurie Fincham (THX), Mike Klasco (Menlo Scientific), and Dr. Floyd Toole (former VP Acoustic Engineering at Harman), who served as the panel moderator. After about 30 minutes, a consensus was quickly reached on the following points:
If he only means typical THD and IM specs, that's not quite the same claim. And I don't think Sean would phrase it as strongly as Earl does..
Well, rather than continue the guesswork, we could ask him.
My reading is that of the above (THD/IM). Both think non-linear is possibly worthy of further investigation.
So exactly what is JA to be bashed for, for
not performing "distortion" measurements (my contention with Arnmir)?? What distortion measurements? On what basis?
cheers,
AJ
I've always wondered about the high score at the lowest distortion, to the right. Golden ears, cheating ?
Hah. I took that test myself many, many moons ago, managed to crack into the -40s IIRC (I'll have to dig up the screenshot). I was astounded others could hear well over 10db lower!!....until I realized there was a glitch in the tracks that allowed identification. I think it was pulled shortly after. Is it back now?
Btw, all of those online tests can be gamed. As the AVS wrecks demonstrated. Take the outlier results with a grain of salt.
High end : exclusive, expensive, superior construction, better finishes, more sophisticated drivers, extended frequency extension.
Whew!! Thank goodness sound has nothing to do with it! Hey, I have a chance now
What about looks? Or is that covered by finishes? I would have thought that would be #1 !!
Thanks I appreciate your input.
cheers,
AJ
Is that your conclusion after having read the methods sections of the research papers ?
My attitude is that simply having, and stating a methodology doesn't automatically make it science, or make it objective. If we are talking about listening to music, then it is automatically subjective to some extent. If a research paper states "A selection of typical music tracks was chosen at random", or "The listeners were allowed to make their own music selections" it doesn't make it science. I was just reading a piece by Higher Authority Sean Olive about preferring the use of 'trained' listeners for audio evaluation. His organisation trains its listeners using a special program of sessions where listeners learn to identify characteristics of processed "music". It seems to me, there are questions prompted by this: no matter how scientific it appears, there is an element of circularity. (As an example)
Olive has used both trained and untrained listeners (http://seanolive.blogspot.com/2008/12/loudspeaker-preferences-of-trained.html) -- and found that their preferences converge. Trained listeners are simply more reliable -- i.e., their individual perforamance from trial does trial does not vary widely, as can happen for untrained -- and efficient testees. You can get a robust answer *faster* using trained listeners.
As for training to hear artifacts or 'characteristics', that is absolutely common in perceptual testing. It , again , is there to *increase* discriminatory power in the testees and thereby increase the detective power of the test.
Have you actually read the research papers? You write as if none of the points you put forward had ever been considered by the 'Higher Authorities'
Can we move on from obsessing over truisms?
Sorry. This branch of the discussion stemmed from my earlier objections to discussions that do nothing but quote from Higher Authorities - which are, in themselves, nothing but a form of endless regurgitation of truisms.
And your alternative to citing research is what? "I heard it, therefore it is real?"
And I'm not understanding the second part or what you wrote --Geddes was there? *Where*?
C'mon now, the link I provided!
Here again: http://seanolive.blogspot.com/2009/01/what...ations-are.html (http://seanolive.blogspot.com/2009/01/what-loudspeaker-specifications-are.html)
The panelists included myself, Steve Temme (Listen Inc.), Dr. Earl Geddes (GedLee), Laurie Fincham (THX), Mike Klasco (Menlo Scientific), and Dr. Floyd Toole (former VP Acoustic Engineering at Harman), who served as the panel moderator. After about 30 minutes, a consensus was quickly reached on the following points:
My bad. If Earl accepts that view of nonlinear distortion -- that its role is still unclear rather than negligable -- as the 'consensus', then all I can fault him for is phrasing his opinion a bit more dogmatically than that, on other forums.
If he only means typical THD and IM specs, that's not quite the same claim. And I don't think Sean would phrase it as strongly as Earl does..
Well, rather than continue the guesswork, we could ask him. My reading is that of the above (THD/IM). Both think non-linear is possibly worthy of further investigation.
Might do!
So exactly what is JA to be bashed for, for not performing "distortion" measurements (my contention with Arnmir)?? What distortion measurements? On what basis?
I'm running the risk of GreenMarker censure for quoting a 'higher authority' yet again, but: here's Olive expanded on this matter in comments on his blog:
[N]one of the current loudspeaker specifications are very relevant -- but there is enough science to fix this.
The relevant specifications would include measures of a) frequency response performance that characterize quality of direct,early, late reflected sounds b) a perceptual meaningful measure of nonlinear distortion and maximum SPL c) perhaps spatial quality (related to directivity). I think a) is already possible, c) needs some work, and b) needs even more work.
I'm not sure I would fault JA specifically on this since few (no?) others are doing those B and C sorts of measurements, and in any case the measurements themselves 'need work'.
But really , do you really want to encourage more Arny/JA squabbling?
btw, I've always wondered about the high score at the lowest distortion, to the right. Golden ears, cheating ?
I'm sure there's some cheating and luck involved.
Case in point, I just did the "one tone" test and I passed the highest score (-60 dB = 0.1%) without a single error on "crappy" headphones with "crappy" onboard. I didn't cheat.
The distorted sample always sounded a bit higher in perceived frequency at the onset.
(-60 dB = 0.001% right?)
60dB is a ratio of 1/1000 = 0.1%. Still, I'm surprised you can hear that!
Define "High End"?
Rich, you were supposed to tell us what a "high end" speaker is! Help me out man, I did my best with the whole FR thing for you, which you hopefully now understand.
Would a Revel Salon qualify?
High end : exclusive, expensive, superior construction, better finishes, more sophisticated drivers, extended frequency extension.
Which are merely claimed and which are actually delivered?
Of course * 100 for percent...
I'm not sure if I was lucky, will try again later. Being 4 times lucky in a row (at the lowest levels of distortion) is not that unlikely.
Define "High End"?
Rich, you were supposed to tell us what a "high end" speaker is! Help me out man, I did my best with the whole FR thing for you, which you hopefully now understand.
Would a Revel Salon qualify?
High end : exclusive, expensive, superior construction, better finishes, more sophisticated drivers, extended frequency extension.
Which are merely claimed and which are actually delivered?
Like I said, you are highly prejudiced on this topic. You aren't exactly one to give a fair, honest assessment of the situation. I suspect your system is predominantly cheap components cobbled together. Anything expensive is just shot down because you can't afford it. But please, if I'm wrong, then please list your components for me so I can see who this
grand judge is in a better light.
Let's see if the Paradigm Studio can match this :
(http://www.highendpalace.com/CES1032.jpg)
(http://www.highendpalace.com/581e.jpg)
(http://www.nolaspeakers.com/products/gifs/exoticagr.jpg)
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
Stereophile has reviewed very similar EgglestonWorks speakers. A speakers that dips to almost 2 ohms in the bass, hast an almost 10 dB high peak in the bass, really weird/bad crossover, uneven directionality ...
... in audiophile lalaland for only $20,000 a pair.
A speakers performance is not defined by price, number of drivers and "high-end" label, you know?
This, btw, is not much different for the other speaker you posted. Just check out this quote:
The Diva Utopia Be—and more particularly the Center Utopia Be—produced a mixed set of measurements. This isn't unusual in high-end speakers.
In the Stereophile tester's room they measured:
... 50Hz room-mode peak at +12dB. Further up the band, however, a deep -10dB suckout from 80 to 120Hz is likely to have much more serious subjective consequences.
~50 Hz is also where this speaker dips dangerously low to almost 2 ohms.
Too bad they never made up to date measurements.
... in audiophile lalaland for only $80,000 for a set.
S&V also did pseudo-anechoic measurements. Weird directivity causes a couple of decibels drop from 2 to 3 kHz, much less at 6 to 8 kHz however. Fun.
Low frequency response was not great for any of those speakers I could find measurements on. Distortion measurements are not available, but I doubt that they would be less horrible.
If Earl accepts that view of nonlinear distortion -- that its role is still unclear rather than negligable -- as the 'consensus', then all I can fault him for is phrasing his opinion a bit more dogmatically than that, on other forums.
Well, my channeling skills suck, but I suspect he's a "Show me, don't tell me" guy. Make him a believer by presenting data, but until then....
I'm not sure I would fault JA specifically on this since few (no?) others are doing those B and C sorts of measurements, and in any case the measurements themselves 'need work'.
Bingo. I mean, sure bash the guy over DBTs and audiophoolery. But his measurement regime is pretty solid, I actually prefer some of how he presents to Soundstage (NRC). But steady state THD, or "to be investigated" non-linear? C'mon now..
But really , do you really want to encourage more Arny/JA squabbling?
I'm entertained by his squabbles either way, whether he's right (DBTs, audiojewelry, et al), or wrong (Bi-amp, speaker distortion, etc.). Especially the Amir ones.
cheers,
AJ
I've always wondered about the high score at the lowest distortion, to the right. Golden ears, cheating ?
Hah. I took that test myself many, many moons ago, managed to crack into the -40s IIRC (I'll have to dig up the screenshot). I was astounded others could hear well over 10db lower!!....until I realized there was a glitch in the tracks that allowed identification. I think it was pulled shortly after. Is it back now?
Btw, all of those online tests can be gamed. As the AVS wrecks demonstrated. Take the outlier results with a grain of salt.
Well, so much for my memory (2007/2008?)
.
The "tell" was a faint click IIRC in one of the tracks. Now the spike at top of the bell curve made sense. I emailed them and the test disappeared shortly after.
May have to try the new one some day. But quite frankly, I have bigger fish to fry.
cheers,
AJ
Let's see if it can match this in dynamics, low frequency response and low distortion output.
Where can those measurements be found for the pictures you are listing??
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
Please provide reliable evidence that those alleged benefits exist.
Standard technical tests will suffice.
What I see in those pictures could easily be poorly integrated collections of cheaply made drivers with a lot of bling.
Low frequency response was not great for any of those speakers I could find measurements on. Distortion measurements are not available, but I doubt that they would be less horrible.
The Eggleston Works speakers use 2 12" woofers per speaker in large cabinets. That's 4 x 12" woofers. You say the low frequency response was not great? Prove it.
A speakers performance is not defined by price, number of drivers and "high-end" label, you know?
Then what is it defined by? Cheap cabinets, cheap drivers, cheap crossovers, slapped together by super glue?
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
Please provide reliable evidence that those alleged benefits exist.
Standard technical tests will suffice.
What I see in those pictures could easily be poorly integrated collections of cheaply made drivers with a lot of bling.
The post above is just reflecting the biases and prejudices of the individual. Show me your list of equipment. I've asked you several times and not once have you bothered to do so. I want to see what system your audio journey has led you to.
Double post.
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
Please provide reliable evidence that those alleged benefits exist.
Standard technical tests will suffice.
What I see in those pictures could easily be poorly integrated collections of cheaply made drivers with a lot of bling.
The post above is just reflecting the biases and prejudices of the individual.
No, its about an agressive poster's inability to provide reliable evidence to back up his counter-scientific assertions.
Show me your list of equipment.
(1) The equipment I happen to possess or use at some time is irrelevant to a discussion of basic science.
For example yesterday I almost bought a Bose Acoustimass system just to see what I sounds like and maybe take it apart to see what's inside. So what?
Right now I'm listening to my favorite 2-channel system: A Sansa Fuze driving a Topping NX1 driving a pair of Sony XBA1 earphones. So what?
(2) From time to time I've mentioned online some of the equipment that I'm listening to. I just mentioned that I sometimes listen to KEF Q15s and provided a link to technical measurements of them. So what?
I've asked you several times and not once have you bothered to do so. I want to see what system your audio journey has led you to.
Its not a matter of bothering or not, its a matter of not exposing myself further to your prejudices.
You should talk about not bothering? You can't even be bothered to properly identify the speakers you seem to want to force us worship based on just your say-so.
Low frequency response was not great for any of those speakers I could find measurements on. Distortion measurements are not available, but I doubt that they would be less horrible.
The Eggleston Works speakers use 2 12" woofers per speaker in large cabinets. That's 4 x 12" woofers. You say the low frequency response was not great? Prove it.
I read on the internet that those 12" drivers have Xmax < 1 mm and come out of one of the sleaziest and most careless speaker driver factories on the Pacific Rim. ;-)
A speakers performance is not defined by price, number of drivers and "high-end" label, you know?
Then what is it defined by? Cheap cabinets, cheap drivers, cheap crossovers, slapped together by super glue?
AFAIK you've just described any of the speakers whose images you put up for us to worship based on just your say-so.
Among rational people speaker performance can be identified by means of good performance in proper technical tests. I provided links to some reliable evidence about the SQ of speakers that AJ & I own. Where is your corresponding evidence?
It is well known that high prices are proof of nothing but high prices and everything else needs to be judged on a case-by-case basis using reliable evidence.
Where's the evidence that the speakers you have put up images of for us to worship are anything but vastly overpriced junk?
High end : exclusive, expensive, superior construction, better finishes, more sophisticated drivers, extended frequency extension.
...
Let's see if the Paradigm Studio can match this :
...
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
A speakers performance is not defined by price, number of drivers and "high-end" label, you know?
A speakers performance is not defined by price, number of drivers and "high-end" label, you know?
Then what is it defined by? Cheap cabinets, cheap drivers, cheap crossovers, slapped together by super glue?
Wow factor, apparently. And as a recovering audiophile () I'd say that's just it. We are like guys in a bar talking about women: Just look at the high-end drivers on
that one.
I hope this is not offensive: I'm thinking on my feet, introspection, and talking about myself as much as others.
Yes, I got some "wow" from the photographs. Then I had an interesting thought, which was that one doesn't see stuff like this (not that I've got closer than photos) in studio control roome; not even the big ones with equally big (and quite possibly much more expensive) monitor setups.
Of course, in the home, speakers
do have to be acceptable pieces of furniture. It helps if they are beautiful, especially if big. It's one of those
preference things.
High end : exclusive, expensive, superior construction, better finishes, more sophisticated drivers, extended frequency extension.
...
Let's see if the Paradigm Studio can match this :
...
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
A speakers performance is not defined by price, number of drivers and "high-end" label, you know?
A speakers performance is not defined by price, number of drivers and "high-end" label, you know?
Then what is it defined by? Cheap cabinets, cheap drivers, cheap crossovers, slapped together by super glue?
Wow factor, apparently. And as a recovering audiophile () I'd say that's just it. We are like guys in a bar talking about women: Just look at the high-end drivers on that one.
I hope this is not offensive: I'm thinking on my feet, introspection, and talking about myself as much as others.
Yes, I got some "wow" from the photographs. Then I had an interesting thought, which was that one doesn't see stuff like this (not that I've got closer than photos) in studio control roome; not even the big ones with equally big (and quite possibly much more expensive) monitor setups.
Of course, in the home, speakers do have to be acceptable pieces of furniture. It helps if they are beautiful, especially if big. It's one of those preference things.
Abbey Road Studios use domestically available (and rather nicely finished) B&W speakers and Classé power amps. They may be an exception, but they are a major recording studio.
(http://www.grahams.co.uk/audio-images/ICON_CLASSEABBEY.jpg)
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
Amazing that when price and dimensions are no object, they still feel the need to add bass resonators to some of them.
Then what is it defined by?
Sound and corresponding measurements of soundwave attributes. You have made a very specific claim about a set of those measurable attributes: "Let's see if it can match this in
dynamics, low frequency response and low distortion output."
Your "pictures" show nothing of the sort. Rich, the measurements of the 3 speaker pictured above, that show these attributes? 2 appear to be 1" dome direct radiators, just like the Paradigm, so this ought to be interesting when you present your data.
If you have polar response data, which is critical to what your perceive at the LP (hopefully you finally got this! ) for that dual tower festooned with drivers speaker, by all means present it.
It ought to be fascinating, much like your visual assessments of loudspeaker "performance".
cheers,
AJ
Low frequency response was not great for any of those speakers I could find measurements on. Distortion measurements are not available, but I doubt that they would be less horrible.
The Eggleston Works speakers use 2 12" woofers per speaker in large cabinets. That's 4 x 12" woofers. You say the low frequency response was not great? Prove it.
(http://www.stereophile.com/images/archivesart/egga2fig2.jpg)
That's for the Andra II (nearfield response of woofers = red). Feel free to show us other measurements.
You call that a great low frequency response?
Good frequency response is not just defined by the bandwidth. Eggle uses this huge peak so they can trick customers into believing that their speaker is flat down to ~35 Hz.
If they were honest they'd have to specify the frequency response as 100 Hz - 18 kHz +/- 3 dB, or 80 Hz - 20 kHz +/- 5 dB, still ignoring the 5 dB dip at 1 kHz as shown in Stereophile's anechoic measurements.
A speakers performance is not defined by price, number of drivers and "high-end" label, you know?
Then what is it defined by? Cheap cabinets, cheap drivers, cheap crossovers, slapped together by super glue?
So you really haven't learned anything at all from the "Audio Hardware with Blind Tested Sonic Signature, specifically speakers" thread?
You didn't read anything we pointed you to?
I'm not going to repeat myself.
This reminds me so of "high-end" devices, where the designers (as in fashion) really think that putting $10 instead of $1 chips into their crap circuits will drastically improve the sound ...
Show me your list of equipment. I've asked you several times and not once have you bothered to do so. I want to see what system your audio journey has led you to.
This is excellent audiophile logic, congrats, you've passed.
Unfortunately, amongst rational, educated people, it's known as a Red Herring. Rather than have my sucky explanation, do you know how to do the Googles on the internets thing?
The Eggleston Works speakers use 2 12" woofers per speaker in large cabinets. That's 4 x 12" woofers. You say the low frequency response was not great? Prove it.
Rich, it looks like you made an excellent audiophile knowledge choice with your Eggleston pic!!
The Savoy Signature (http://egglestonworks.com/products/the-savoy-signature/) claims Frequency Response: -3dB @
18Hz . Definitely spanks the Paradigm!...on paper and audiophile mind at least. Unfortunately, no measurements, but wait, do not despair, because we do have measurements of another to get an idea of actual performance.
EgglestonWorks Andra II loudspeaker (http://www.stereophile.com/content/egglestonworks-andra-ii-loudspeaker-specifications), with even better
17hz extension.
As you can clearly see here:
cheers,
AJ
Don't you see how they start measuring at 1 kHz at -10 dB (only 5 dB away from the actual dip) and go from there?
17 Hz to 24 kHz, -3 dB(/+18 dB offset by -5 dB at 1 kHz)
Don't you see how they start measuring at 1 kHz at -10 dB (only 5 dB away from the actual dip) and go from there?
Quite so. By such standards of injurneering excellence, these Egglestons (http://www.stereophile.com/floorloudspeakers/237/index.html) would be flat to 10Hz referenced to 3Khz:
cheers,
AJ
The Eggleston Works speakers use 2 12" woofers per speaker in large cabinets. That's 4 x 12" woofers. You say the low frequency response was not great? Prove it.
(1) we have reliable evidence that just packing more bigger woofers into a too-small cabinet is not the optimal route to good bass.
http://www.stereophile.com/content/egglest...er-measurements (http://www.stereophile.com/content/egglestonworks-andra-ii-loudspeaker-measurements)
(http://www.stereophile.com/images/archivesart/egga2fig3.jpg)
Looking around a little I found the probable tech specs for the woofers in the above system:
http://www.dynaudio.com/int/pdf/DYN_Automo...rochure_INT.pdf (http://www.dynaudio.com/int/pdf/DYN_Automotive_Brochure_INT.pdf)
My first question was: "How did anybody mess up the response of such nicely speced drivers." (Esotar² 1200)
An answer is right there in the driver spec sheet: Vas = 163 liters = 5.5 cubic feet. Rule of thumb is that a speaker is well matched to an enclosure with volume equal to about 3 times driver Vas or in this case 15 cubic feet. Twice that for two drivers which is the case here or about 30 cubic feet.
The speaker enclosure Dimensions are given as 46" H by 15" W by 18" D. but enclosure wall thickness is about 1.6" so we have to take 3 inches off of every measurement giving 43" x 12" x 15". That nets out to about 5.5 cubic feet or only 25% of optimum.
So the double thickness MDF is a nice touch (but naively low tech given that far more stiffness can be obtained with less materials and weight using effectively designed cross bracing). It can't help making a rookie mistake like putting the wrong drivers in a too-small box. It could be largely circumvented with some well-designed equalization, but in the high end audio world, equalization is usually thought to be some kind of evil process.
Abbey Road Studios use domestically available (and rather nicely finished) B&W speakers and Classé power amps. They may be an exception, but they are a major recording studio.
(http://www.grahams.co.uk/audio-images/ICON_CLASSEABBEY.jpg)
Yes, there are exceptions. Abbey Road famously like their B&Ws.
There are also crossover companies that sell into both pro and consumer markets, where the consumer model may simply be passive rather than active, or might be identical but clad in fancier woodwork for the living room look.
If you look at the shape, and consider that the woofers probably use a separate chamber, you get to a much smaller volume. There's a vent in the back.
(http://www.ultraaudio.com/features/pics/200912_ew2.jpg)
Also, there are two woofers. The second one seems to be acoustically in series behind the visible one. This reduces Vas, but also lowers efficiency.
Weird.
Can we move on from obsessing over truisms?
Sorry. This branch of the discussion stemmed from my earlier objections to discussions that do nothing but quote from Higher Authorities - which are, in themselves, nothing but a form of endless regurgitation of truisms.
And your alternative to citing research is what? "I heard it, therefore it is real?"
Not even that. These days "I read it on the web" suffices as uncontestable revealed truth.
These days "I read it on the web" suffices as uncontestable revealed truth.
Not all that much different from "I read it in a book".
"I have experience: I once even did it for my job" is not incontestable revealed truth either.
These days "I read it on the web" suffices as uncontestable revealed truth.
Not all that much different from "I read it in a book".
"I have experience: I once even did it for my job" is not incontestable revealed truth either.
So where are you getting your info from mate?
Yes, there are exceptions. Abbey Road famously like their B&Ws.
Great choice when you have an aircraft carrier sized mixing board stretching across the room between you and the speakers. I bet that sounds really good playing the Beattles. "Accurate" too.
cheers,
AJ
These days "I read it on the web" suffices as uncontestable revealed truth.
Not all that much different from "I read it in a book".
I'm not sure about that at all.
Lots of crap on the web that would never ever make it into a book.
Lots of very good books: Toole, Winer, Dickason, Pohlman...
"I have experience: I once even did it for my job" is not incontestable revealed truth either.
Particularly true for people with IT, recording, production, and live sound experience. Some call it engineering, and I've done it all. Speaking from those experiences, its not really engineering.
Some of the most mislead people on the web are EE's.
I'm not sure about that at all.
Lots of crap on the web that would never ever make it into a book.
Possibly, but there
are published books on astrology, phrenology, crystal healing etc. Publication is not a guarantee of incontestability.
Science can be difficult to distinguish from pseudoscience, and I just think that some of the audio 'science' is getting into the pseudo category. What I mean by that is that as soon as we try to investigate whether a human listener can hear distortion at some level in "music" we have crossed the line; aesthetic judgements (which this has then become, no matter how much we try to avoid it) are outside the scope of science.
If the idea is that distortion levels can be higher in "music" and not, say, sine waves, then a difficulty is that I could come along and create a piece of music comprising four minutes thirty three seconds of 1 kHz sine wave (and "publish" it on Spotify etc. :-) ). If our would-be scientist then adds the qualifier "normal" or "typical" (music) then I think the conclusions of the experiment must always be open to question because of the possibility of cultural bias etc.
Basically, even if a scientific procedure is followed, an experiment may not be science if what is being investigated is outside the scope of science. Possibly a good topic for another thread!
Let's see if the Paradigm Studio can match this :
<jewelry deleted>
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
Yeah, let's! I would *love* to see a decent set of measurement of these amusing examples of audiophile eyecandy
FWIW, the loudspeakers models RichB has shown are:
Eggleston Savoy Signature ($55,000/pr)
JM Lab Grande Utopia BE ($80,000/pr)
Nola Grand Reference VI Gold ($298,000/pr)
The question is, of course, how much do you have to spend to get audio performance as good or better than these? To answer that we have ot know how they perform.
Neither Eggleston nor Nola has any models measured at Soundstage.
There's two JM lab models measured at Soundstage -- the measurements are old and not as comprehensive as later ones. The 'Mini Utopia' has quite nice performance but certainly is no bass monster (steep falloff below 50 Hz) nor is the perfamnce matchless.
http://www.soundstagemagazine.com/measurem...lab_miniutopia/ (http://www.soundstagemagazine.com/measurements/jmlab_miniutopia/)
As for the JM CHorus 706 (a bookshelf speaker) , well, hmm, I wouldn't exactly brag about that one
http://www.soundstagemagazine.com/measurem...mlab_chorus706/ (http://www.soundstagemagazine.com/measurements/jmlab_chorus706/)
There are something like *10* Paradigm models measured by Soundstage , some perform better than others, so which model are we talking about here as the reference?
http://www.soundstagenetwork.com/index.php (http://www.soundstagenetwork.com/index.php)
RichB, have you investigated Sean Olive's loudspeaker comparison work *at all*? Are you aware that one of his experiments included a top-rated loudspeaker of one of the audiophile mags -- and it didn't do very well at all either objectively or subjectively, compared to some much less pricey models? Do you find that incredible?
Yes, there are exceptions. Abbey Road famously like their B&Ws.
Great choice when you have an aircraft carrier sized mixing board stretching across the room between you and the speakers. I bet that sounds really good playing the Beattles. "Accurate" too.
cheers,
AJ
Yeah, I don't get that setup *at all*, unless there is some provision for moving the mixing board out of the way, or levitating the loudspeakers.
Yes, there are exceptions. Abbey Road famously like their B&Ws.
In that case there are many exceptions. Just to name a few:
Bob Ludwig mastering (Egglestone) (http://upload.wikimedia.org/wikipedia/commons/4/40/BobLudwigGateway.jpg)
Wisseloord mastering (Egglestone) (http://www.wisseloord.nl/studios/mastering/)
Soundmirror (B&W e.a.) (http://www.soundmirror.com/studio/)
Teldex Berlin (B&W) (http://www.teldexstudio.de/en/studios/postproduction.html)
Polyhymnia (B&W) (http://polyhymnia.com/studios-equipment/our-studios-in-baarn/)
Sterling Sound (B&W) (http://sterling-sound.com/wp-content/uploads/IMG_4596QBacon.jpg)
Emil Berliner Studios (B&W) (http://www.emil-berliner-studios.com//img/header/header_regie12_01.jpg)
Science can be difficult to distinguish from pseudoscience, and I just think that some of the audio 'science' is getting into the pseudo category. What I mean by that is that as soon as we try to investigate whether a human listener can hear distortion at some level in "music" we have crossed the line; aesthetic judgements (which this has then become, no matter how much we try to avoid it) are outside the scope of science.
What line does that cross? Are you saying that audio reproduction is arbitrary and without any goal?
Do you know what high fidelity stands for? Flat frequency response, minimal amount of noise and distortion, ...
This sounds more like an anti-science stance than an attempt to distinguish science from pseudoscience.
If the idea is that distortion levels can be higher in "music" and not, say, sine waves, then a difficulty is that I could come along and create a piece of music comprising four minutes thirty three seconds of 1 kHz sine wave (and "publish" it on Spotify etc. :-) ). If our would-be scientist then adds the qualifier "normal" or "typical" (music) then I think the conclusions of the experiment must always be open to question because of the possibility of cultural bias etc.
Basically, even if a scientific procedure is followed, an experiment may not be science if what is being investigated is outside the scope of science. Possibly a good topic for another thread!
What instrument produces a single clean sine wave? All bets are off as soon as you start generating artificial sounds ... you can even make phase shifts audible that are completely inaudible with music.
Science investigates both, the effect of distortion on pure tones and on music, be it jitter, harmonics distortion ... It's all science.
The only problem I see is using single tone results and claim that you can hear the same with real music. That is pseudoscience, or audiophile FUD.
Let's see if the Paradigm Studio can match this :
<jewelry deleted>
Let's see if it can match this in dynamics, low frequency response and low distortion output. There are many, many examples out there.
Yeah, let's! I would *love* to see a decent set of measurement of these amusing examples of audiophile eyecandy
FWIW, the loudspeakers models RichB has shown are:
Eggleston Savoy Signature ($55,000/pr)
JM Lab Grande Utopia BE ($80,000/pr)
Nola Grand Reference VI Gold ($298,000/pr)
The question is, of course, how much do you have to spend to get audio performance as good or better than these? To answer that we have ot know how they perform.
Neither Eggleston nor Nola has any models measured at Soundstage.
There's two JM lab models measured at Soundstage -- the measurements are old and not as comprehensive as later ones. The 'Mini Utopia' has quite nice performance but certainly is no bass monster (steep falloff below 50 Hz) nor is the perfamnce matchless.
http://www.soundstagemagazine.com/measurem...lab_miniutopia/ (http://www.soundstagemagazine.com/measurements/jmlab_miniutopia/)
As for the JM CHorus 706 (a bookshelf speaker) , well, hmm, I wouldn't exactly brag about that one
http://www.soundstagemagazine.com/measurem...mlab_chorus706/ (http://www.soundstagemagazine.com/measurements/jmlab_chorus706/)
There are something like *10* Paradigm models measured by Soundstage , some perform better than others, so which model are we talking about here as the reference?
http://www.soundstagenetwork.com/index.php (http://www.soundstagenetwork.com/index.php)
RichB, have you investigated Sean Olive's loudspeaker comparison work *at all*? Are you aware that one of his experiments included a top-rated loudspeaker of one of the audiophile mags -- and it didn't do very well at all either objectively or subjectively, compared to some much less pricey models? Do you find that incredible?
Yes I am aware of Sean Olives work and not surprisingly, his speakers, endorsed by his company, were preferred the most.
Science can be difficult to distinguish from pseudoscience
Personally I think people familiar with science can easily tell the difference.
I just think that some of the audio 'science' is getting into the pseudo category.
You can indeed find plenty of experiments done by people who pretend or think they are doing science (and present it as such) but neglect to use proper controls and misinterpret the results.
What I mean by that is that as soon as we try to investigate whether a human listener can hear distortion at some level in "music" we have crossed the line; aesthetic judgements (which this has then become, no matter how much we try to avoid it) are outside the scope of science.
No. Whether the presence of a given type of distortion is audible under certain conditions is not an aesthetic judgment. Whether a given type of audible distortion is tolerable or perhaps desirable is an aesthetic judgment.
If the idea is that distortion levels can be higher in "music" and not, say, sine waves, then a difficulty is that I could come along and create a piece of music comprising four minutes thirty three seconds of 1 kHz sine wave (and "publish" it on Spotify etc. :-) ). If our would-be scientist then adds the qualifier "normal" or "typical" (music) then I think the conclusions of the experiment must always be open to question because of the possibility of cultural bias etc.
If you know what you are testing for, then you try to find a worst case signal to determine audibility. If something is only audible under certain conditions, and you can relate those conditions to more "typical" conditions in the real world, you can then attempt to make a judgment about audibility under "typical" conditions. You also can identify what may be the "exceptional" cases.
What's important is not the semantics of calling something "typical" or "exceptional" as much as learning how different specific conditions affect audibility.
Basically, even if a scientific procedure is followed, an experiment may not be science if what is being investigated is outside the scope of science. Possibly a good topic for another thread!
But if we accept this premise, then people who lack a solid understanding of acoustics, analog electronics, digital signal processing, the physiology of human hearing, et al. will just argue that the things they don't know are "outside the scope of science" without bothering to do any reading. And if one is not rather familiar with the research that has been done, then one is in no position to evaluate the veracity of that research.
It is not uncommon to see arguments that "knowledge is worthless" when people who lack a certain background find themselves arguing with people with much more extensive knowledge. They are at a distinct disadvantage in the argument and it's much easier for them to just dismiss all that knowledge than it is to do some reading, admit their ignorance, and/or (gasp) risk losing the argument.
Yes I am aware of Sean Olives work and not surprisingly, his speakers, endorsed by his company, were preferred the most.
Excuse the expression, but you're either a troll or a moron.
I read on the internet that those 12" drivers have Xmax < 1 mm and come out of one of the sleaziest and most careless speaker driver factories on the Pacific Rim. ;-)
Please document a source for this outrageous claim.
AFAIK you've just described any of the speakers whose images you put up for us to worship based on just your say-so.
Perhaps you are irrational, because the speakers I referenced above have anything but poor build quality. You seem to be casting aspersions here, which is intellectually dishonest behaviour. No one who is reasonable or rational for that matter would make any claim that the build quality was shoddy, or that the drive units were inadequate, or that it was just overpriced junk.
You pretty much tip your biases and prejudices by spewing such nonsense. I asked you to cite your equipment so that I could judge your equipment and choices as you seem so very confident to judge the systems I've referenced. Seems only fair.
Among rational people speaker performance can be identified by means of good performance in proper technical tests.
Good performance in a technical test, according to your biblical definition of what constitutes 'good performance' is no guarantee of good sound quality within the room.
It is well known that high prices are proof of nothing but high prices and everything else needs to be judged on a case-by-case basis using reliable evidence.
Right, and you've provided no reliable evidence in support of your bogus claims, especially the claim about <1mm Xmax drivers. You are highly biased or conditioned, if you will, to think that expensive speakers are just overpriced junk. It's difficult to argue with someone whose prejudices override any reasonable or rational approach to honest discussion.
Where's the evidence that the speakers you have put up images of for us to worship are anything but vastly overpriced junk?
You are claiming that the speakers are vastly overpriced junk, phrased as a question, thinking I'm going to take the bait. First show that they are overpriced junk. All you've done is spew nonsense here.
Yes I am aware of Sean Olives work and not surprisingly, his speakers, endorsed by his company, were preferred the most.
Excuse the expression, but you're either a troll or a moron.
Good argument. Just spew insults when you have nothing salient to offer. Sean Olive tests the speakers in his own facility, using techniques best served to bring about a result that people prefer, and lo and behold, out of all the speakers tested, his speakers were preferred.
Nothing at all suspicious about that. You're either just plain naive, or just in denial.
Yes, there are exceptions. Abbey Road famously like their B&Ws.
In that case there are many exceptions. Just to name a few:
Bob Ludwig mastering (Egglestone) (http://upload.wikimedia.org/wikipedia/commons/4/40/BobLudwigGateway.jpg)
Wisseloord mastering (Egglestone) (http://www.wisseloord.nl/studios/mastering/)
Soundmirror (B&W e.a.) (http://www.soundmirror.com/studio/)
Teldex Berlin (B&W) (http://www.teldexstudio.de/en/studios/postproduction.html)
Polyhymnia (B&W) (http://polyhymnia.com/studios-equipment/our-studios-in-baarn/)
Sterling Sound (B&W) (http://sterling-sound.com/wp-content/uploads/IMG_4596QBacon.jpg)
Emil Berliner Studios (B&W) (http://www.emil-berliner-studios.com//img/header/header_regie12_01.jpg)
All these studios are using vastly overpriced junk, according to a few know-it-alls here. You know, it really wouldn't surprise me if Arnold, or some others here, were using cheap $200 junk speakers connected to a cheap AVR, in an untreated room.
Excuse the expression, but you're either a troll or a moron.
Good argument. Just spew insults when you have nothing salient to offer. Sean Olive tests the speakers in his own facility, using techniques best served to bring about a result that people prefer, and lo and behold, out of all the speakers tested, his speakers were preferred.
Nothing at all suspicious about that. You're either just plain naive, or just in denial.
Everyone here can see that you are completely ignorant of the research, including the research they did back at the NRC, not Harman.
Then you don't seem to understand the idea of a double blind test, like at all.
You also don't seem to get that there is research without Harman employees, of non Harman products, ... and even if you exclude all Harman products from their recent research, there's still the same conclusion.
You also seem to be ignorant of the fact that Harman products did not always score well in the tests. The fact that Olive gave us enough information to e.g. identify the anonymized products in their headphone comparison, where a premium AKG product ranked worse than a Bose noise-canceling hp, also doesn't fit your little whacky conspiracy theory.
You just made a complete ass out of yourself without even noticing it.
Don't worry, you'll never hear from me again. Let us know when your brain has overcome the conspiracy theories though.
I read on the internet that those 12" drivers have Xmax < 1 mm and come out of one of the sleaziest and most careless speaker driver factories on the Pacific Rim. ;-)
Please document a source for this outrageous claim.
Is that an outrageous claim? Given the technical malfeasance that is so obvious in the design of the Eggonyourface speakers, (or is that Eggleston?) why wouldn't some of the other mystery meat you referenced be equally defective?
I read on the internet that those 12" drivers have Xmax < 1 mm and come out of one of the sleaziest and most careless speaker driver factories on the Pacific Rim. ;-)
Please document a source for this outrageous claim.
Is that an outrageous claim? Given the technical malfeasance that is so obvious in the design of the Eggonyourface speakers, (or is that Eggleston?) why wouldn't some of the other mystery meat you referenced be equally defective?
The absence of any reliable evidence from you says it all. All talk, no substance.
It is not uncommon to see arguments that "knowledge is worthless" when people who lack a certain background find themselves arguing with people with much more extensive knowledge. They are at a distinct disadvantage in the argument and it's much easier for them to just dismiss all that knowledge than it is to do some reading, admit their ignorance, and/or (gasp) risk losing the argument.
Yes, if someone did say that "knowledge was worthless", you might, indeed think that. (But it would be very bad manners to say it to them! I'm sure you wouldn't do that. One might assume that the argument was lost as soon as that happened.)
"Extensive knowledge" can cover a multitude of different levels. The extensive knowledge obtained "on the shop floor" is not necessarily a substitute for the extensive knowledge obtained through reading text books, which is not a substitute for an inquiring mind. Extensive 'shop floor' knowledge of scientific procedure cannot tell you whether questions are beyond the scope of science - such as "how do we define good art?" or "is there a God?". There is a common misconception that if 'scientific procedure' is followed then what results is 'science'. But this is not true. We could follow scientific procedure to attempt to define what is funny. It might be interesting, and maybe not worthless knowledge, but it wouldn't be science.
Utopia BE measurements have already been mentioned:
(http://www.stereophile.com/images/archivesart/GubFig1.jpg)
(in room, 5 dB/div)
Here's a JBL speaker in the same room:
(http://www.stereophile.com/images/archivesart/GubFig5.jpg)
Nobody seems to be able to (or be willing to) measure the Nola speakers. Understandable considering the price and "reputation" involved.
Stereophile did measurements of the Nola Contender however (green/blue are the ports):
(http://www.stereophile.com/images/112Nolfig3.jpg)
I don't know how you can mess up a 3 driver speaker (2 woofers, 1 tweeter) like that.
What instrument produces a single clean sine wave? All bets are off as soon as you start generating artificial sounds ... you can even make phase shifts audible that are completely inaudible with music.
You've just mentioned one of the things that started me down this strand of thought: 'science' tells us that "phase is inaudible with music". If so, then this conveniently lets passive speakers off the hook in terms of their phase distortion. But there are so many questions to ask. Like when we say "inaudible" are we meaning in timbre only? Or do we mean inaudible in terms of 'imaging'? Mono or stereo? And then the vexed question of how we define "music". As I pointed out, music could comprise pure sine waves or other steady tones (such as produced by the glass harmonica, theremin, pipe organ) so the claim about inaudibility with music in terms of just timbre doesn't even stand up.
Basically, as soon as someone says "Phase is inaudible with music. End of." my sceptical antennae twitch! Much more information is needed, and even then it cannot tell the whole story.
Everything becomes audible if you overdo it, even phase shift with music. There is no single hard line for everyone and every type of music.
"Inaudible" as in you cannot hear a difference. I'm sure you have read this:
human-hearing-phase-distortion-audibility-part-2 (http://www.audioholics.com/room-acoustics/human-hearing-phase-distortion-audibility-part-2)
Phase shift of course needs to be identical in both channels.
What I'm saying is that you do not need to obsess about something like phase, but of course hi-fi reproduction is trying to minimize it, same as distortion.
If there any EEs here, I have a question. How do I know if my amplifier is clipping? Would I hear lots of distortion, or compression or are there any other symptoms I can expect?
What I've been told is that if my amp is not clipping then it is supplying all the power my speakers need. But perhaps I'm not aware that my amp is clipping because I'm not focusing on the clipping. In electrical terms, if the amp is not clipping then the power supplies are not being "drained"?
I'm sure you have read this:
Indeed, and I couldn't have linked to a better example of why my questions are relevant! It by no means says that phase is inaudible, even in music. It says that phase distortion is more audible in headphones than speakers, so are we saying that it is passive speakers' phase distortion that is masking the audbility of the phase shifts in the first place i.e. it makes no sense to attempt to assess the effect using passive speakers. And so on!
Edit:
What I'm saying is that you do not need to obsess about something like phase, but of course hi-fi reproduction is trying to minimize it, same as distortion.
Yes, that's reasonable, but if you truly believe that, the only way to (objectively) minimise all forms of distortion is to use digital sources, DSP and active speakers. Do you? (I do)
No, we are saying that a room messes up the phase so much, that those amounts of phase shifts are being masked.
The all-pass networks used covered a huge range, where passive speakers typically do not have a crossover or roll-off.
Our hearing just isn't that interested in absolute phase shifts (identical for each ear). Why would it?
edit: Yes, digital, DSP and active all the way for me!
No, we are saying that a room messes up the phase so much, that those amounts of phase shifts are being masked.
But the idea that the room causes phase shifts is a very frequency domain-centric interpretation. The room is contributing delayed reflections that, were the music steady tones, would indeed appear as phase shifts. But on time-varying signals our ears are simply picking up multiple 'echoes' in the time domain, and there is a completely consistent relationship between the frequency and time domains, even as the listener moves in the room. This is different from a speaker that has a flat frequency response, but arbitrary phase distortion.
Yes I am aware of Sean Olives work and not surprisingly, his speakers, endorsed by his company, were preferred the most.
If Olive did his homework and developed reliable evidence about what people generally desire and prefer in terms of speaker performance, and then the organization that he led failed to act on that knowledge and build speakers accordingly, what sort of fools would they be?
Technical work by people like Olive, Toole, Greisinger, and Keele is published in the annals of highly respected professional technical organizations such as the AES.
I did a search in the AES and IEEE databases on Eggleston and come up dry. I wondre why?
Seems to me like Eggleston needs to catch up on his reading:
"Closed box speaker system analysis" JAES vol 20 number 5, 1972.
If the speakers themselves masked the phase shift, then we wouldn't see much better audibility of phase shifts with speakers in anechoic conditions, similar to headphones.
Also, phase shift is additive and therefore you should still hear the filters being enabled, adding to the phase shift of the speaker.
If the speakers themselves masked the phase shift, then we wouldn't see much better audibility of phase shifts with speakers in anechoic conditions, similar to headphones.
Also, phase shift is additive and therefore you should still hear the filters being enabled, adding to the phase shift of the speaker.
Well, I am simply taking you at your word and minimising all forms of distortion as best I can. I am not looking for reasons why human hearing might be insensitive to phase shifts or nonlinear distortion, or even putting together scientific experiments to prove it (although it is a "truism" to say that science never proves anything, except in mathematics).
If there any EEs here, I have a question. How do I know if my amplifier is clipping?
My formal education and work experience includes undergraduate and post graduate work in all topics commonly related to a degree in Electrical Engineering, so calling myself an EE requires a certain amount of humility which I can muster from time to time. ;-)
Clipping is always detectable via conventional signal analysis using an oscilloscope, or FFT analyzer, etc. even in amounts that are far smaller than any that are audible.The best way to measure clipping is to record the electrical output of the power amplifier and go back and analyze the recording which I have done many times.
Would I hear lots of distortion, or compression or are there any other symptoms I can expect?
In small amounts, clipping is inaudible. In large amounts it alters first the harmonic structure and then the dynamics of music and dialogue. The most audible effects of clipping are usually intermodulation distortion because the artifacts it creates are not necessarily related to the harmonic structure of the signal.
What I've been told is that if my amp is not clipping then it is supplying all the power my speakers need.
That is false because the amount of power that your speakers need stands apart from equipment effects such as clipping.
The amount of power that your speakers need to please you is dependent on things like:
(1) First and foremost your preferences for loudness
(2) The actual efficiency of your speaker system.
(3) The distance from the listening point to the speakers
(4) Effects of the room such as cabin gain and reinforcement by reflections and canclellation off of surfaces in the room.
(5) A system with an effective subwoofer system is far less prone to have audible distortion due to amplifier clipping or nonlinearities in the speakers.
etc.
What I would tell you is that If you are pleased with your system's loudness and your amplifiers are not clipping then there is a very high probability that your amplifiers have sufficient power.
But perhaps I'm not aware that my amp is clipping because I'm not focusing on the clipping.
That is true.
In electrical terms, if the amp is not clipping then the power supplies are not being "drained"?
The most common source of clipping in actual use is that the amp is being called on to provide more energy at any particular moment than it is capable of providing without clipping. The reasons for that vary with the situation, but in general clipping is caused by the limited ability of the amplifier to produce the required current and voltage. The most commonly exceeded limit relates to voltage.
Voltage limiting in power amps is usually caused by the amplifier circuitry requiring more voltage than the power supplies can supply. Due to the high crest factor of music and dialog and legal requirements for amplifier power ratings to be advertised in the US, most amplifiers clip before their power supplies are significantly drained.
That is false because the amount of power that your speakers need stands apart from equipment effects such as clipping.
How is it false though? If the amp is not clipping, which you mean to say is not running out of voltage and current, then the amp is supplying all the power the speaker needs at that moment in time.
I should have said at my listening levels, because I turn the volume to max and it might start clipping. But at my listening levels, if I do not notice audible distortion or compression, that is a tell tale sign that my amplifier is not being limited in its power delivery?
In actual playback, if I don't notice a problem then I guess I don't have one.
No, we are saying that a room messes up the phase so much, that those amounts of phase shifts are being masked.
But the idea that the room causes phase shifts is a very frequency domain-centric interpretation.
I don't know where one might get that idea or the seemingly implied idea that frequency-domain analysis would conceal relevant facts.
The time domain and frequency domain are perfectly related to each other by a number of different mathematical means, the most common of which is Fourier analysis. If an artifact exists in the time domain then it also exists in the frequency domain.
The room is contributing delayed reflections that, were the music steady tones, would indeed appear as phase shifts.
Steady tones are not required to identify or quantify phase shifts. Using them may make it easier to measure using simple or ancient traditional means, but that does not mean that extant and widely used means can't be used to identify and quantify them from non-steady signals such as impulses.
But on time-varying signals our ears are simply picking up multiple 'echoes' in the time domain,
The ear/brain is also capable of identifying phase shifts in the lower audible range, and it is also (far more) capable of identifying the effects of phase shifts over the entire audible range.
...there is a completely consistent relationship between the frequency and time domains, even as the listener moves in the room.
Agreed and that shows up in proper measurements and listening tests with steady or musical (non steady) signals.
This is different from a speaker that has a flat frequency response, but arbitrary phase distortion.
How?
The (to quote you) "... completely consistent relationship between the frequency and time domains..." is not an isolated mathematical concept but one that is pervasive and one that effects what we measure and what we hear.
That is false because the amount of power that your speakers need stands apart from equipment effects such as clipping.
How is it false though? If the amp is not clipping, which you mean to say is not running out of voltage and current, then the amp is supplying all the power the speaker needs at that moment in time.
The speaker does not run in a vacuum, it exists to meet the needs of the listener. If the speaker is not cleanly reproducing sound at a loud enough level to please the listener then the system is failing whether it is clipping or not.
I should have said at my listening levels,
Well yes!
because I turn the volume to max and it might start clipping. But at my listening levels, if I do not notice audible distortion or compression, that is a tell tale sign that my amplifier is not being limited in its power delivery?
No for the reason I stated. What is unclear about:
"Clipping is always detectable via conventional signal analysis using an oscilloscope, or FFT analyzer, etc. even in amounts that are far smaller than any that are audible."
Just because you don't hear clipping doesn't mean that your amplifier isn't clipping. Just because you don't hear clipping today doesn't mean that you won't notice it tomorrow.
In actual playback, if I don't notice a problem then I guess I don't have one.
Well, you don't have a problem today. What about tomorrow?
It is possible and even easy to build an audio system that can never be clipped. However nobody can guarantee that it won't be dissatisfactory tomorrow because it isn't loud enough.
Back to distortion, look at this Focal bookshelf (http://img2.audio.de/Messwerte-f960x960-ffffff-C-ae9b3ed9-110025964.jpg).
In the best case THD is down probably -50 to -60 dB. That's over 0.3% to 0.1%. In the bass it's more like 1% to 2%.
A random floorstanding speaker, same story (http://img1.audio.de/Heco-Celan-GT-902-r960x600-C-8d769dc2-49906877.jpg). (93 dB)
A €25,000 floorstanding speaker, worse in the midrange (http://img1.audio.de/KEF-Blade-f960x960-ffffff-C-443ac08f-71243221.jpg). (91 dB)
A much cheaper floorstanding, cleaner (http://img2.audio.de/image-f960x960-ffffff-C-af5c7254-71886868.jpg). (96 dB according to manufacturer)
As you can see, these speakers don't need much power to reach high SPL and high distortion. Less than 20W @ 4ohm for 100 dB SPL.
Keep in mind that this is just harmonic distortion. As soon as you play real music you will get loads of intermodultion products.
I don't know how you can mess up a 3 driver speaker (2 woofers, 1 tweeter) like that.
It is very simple to create a speaker that is a mess, even out of good components if you apply certain kinds of high end audio "Wisdom". I've seen it done many times. I think I caught an example of it with the Eggleston, for example.
(1) Develop the speaker entirely by ear, don't use any of the common and readily available technological aids. Measurements are for the naive.
(2) Avoid studying the widely available technical articles by Olive, Toole, etc. because they are sold out to the mid-fi audio establishment. Never study any relevant IEEE or JAES papers for the same reason. If the math is over your head never study math and science until it isn't over your head, that's for eggheads.
(3) Avoid using accepted technology for designing the enclosure, picking drivers and designing the crossover.
(4) Focus on the web buzz about technical features and follow the latest opinions of high end reviewers and internet trolls.
(5) Mimic existing high end loudspeaker designs.
(6) Don't start out simple and work your way up. Never work in private. That will only delay your entry into the pantheon of high end speaker designers. Your first effort needs to advance the state of the loudspeaker design art or you are a fool.
No for the reason I stated. What is unclear about:
"Clipping is always detectable via conventional signal analysis using an oscilloscope, or FFT analyzer, etc. even in amounts that are far smaller than any that are audible."
Nothing unclear per se, but I don't have the expertise to do diagnostic testing, so I can't run these tests. I rely on audible sound to detect whether there is a problem.
Just because you don't hear clipping doesn't mean that your amplifier isn't clipping. Just because you don't hear clipping today doesn't mean that you won't notice it tomorrow.
But now I'm uncertain as to whether I would need a more powerful amplifier or not. If I'm not hearing audible distortion, then I would think I'm in the clear. But you are saying that it might be there and I'm not aware of it.
So then when is it a good idea to upgrade your amplifier? Is there no sure-fire way to assess this problem purely by listening, or must it be confirmed always by testing on the bench?
In actual playback, if I don't notice a problem then I guess I don't have one.
Well, you don't have a problem today. What about tomorrow?
I don't listen at ridiculous volumes. With an SPL meter, I think the loudest peaks are around 95-96 dB. Average around 65-75 dB.
It is possible and even easy to build an audio system that can never be clipped. However nobody can guarantee that it won't be dissatisfactory tomorrow because it isn't loud enough.
When people go out and buy bigger amps, I've seen you criticise them for making choices that wouldn't necessarily benefit their bottom-line performance. It's all over AVS.
What people on AVS say is that if the amp isn't audibly clipping, there isn't an audible problem and power is not being limited to the speakers. So since I'm not experiencing audible distortion that I can detect, I am thinking that I'm in the clear and that my amp is big enough for my listening requirements. Do I have reason to think that my assumptions are misguided?
Back to distortion, look at this Focal bookshelf (http://img2.audio.de/Messwerte-f960x960-ffffff-C-ae9b3ed9-110025964.jpg).
In the best case THD is down probably -50 to -60 dB. That's over 0.3% to 0.1%. In the bass it's more like 1% to 2%.
A random floorstanding speaker, same story (http://img1.audio.de/Heco-Celan-GT-902-r960x600-C-8d769dc2-49906877.jpg). (93 dB)
A €25,000 floorstanding speaker, worse in the midrange (http://img1.audio.de/KEF-Blade-f960x960-ffffff-C-443ac08f-71243221.jpg). (91 dB)
A much cheaper floorstanding, cleaner (http://img2.audio.de/image-f960x960-ffffff-C-af5c7254-71886868.jpg). (96 dB according to manufacturer)
As you can see, these speakers don't need much power to reach high SPL and high distortion. Less than 20W @ 4ohm for 100 dB SPL.
Keep in mind that this is just harmonic distortion. As soon as you play real music you will get loads of intermodultion products.
Please forgive me, but can you explain how I am supposed to read the graphs you provided? These are distortion graphs, or frequency response graphs?
No for the reason I stated. What is unclear about:
"Clipping is always detectable via conventional signal analysis using an oscilloscope, or FFT analyzer, etc. even in amounts that are far smaller than any that are audible."
Nothing unclear per se, but I don't have the expertise to do diagnostic testing, so I can't run these tests. I rely on audible sound to detect whether there is a problem.
Why limit yourself that way?
Just because you don't hear clipping doesn't mean that your amplifier isn't clipping. Just because you don't hear clipping today doesn't mean that you won't notice it tomorrow.
But now I'm uncertain as to whether I would need a more powerful amplifier or not. If I'm not hearing audible distortion, then I would think I'm in the clear. But you are saying that it might be there and I'm not aware of it.
So then when is it a good idea to upgrade your amplifier? Is there no sure-fire way to assess this problem purely by listening, or must it be confirmed always by testing on the bench?
You say below that you've read what I've written on AVS. I've written there many times about how to evaluate this problem, right? Why are you asking questions here that I've answered there many, many times?
In actual playback, if I don't notice a problem then I guess I don't have one.
Well, you don't have a problem today. What about tomorrow?
I don't listen at ridiculous volumes. With an SPL meter, I think the loudest peaks are around 95-96 dB. Average around 65-75 dB.
Same question - if you've read my answers on AVS why ask the same questions I answered there, here?
It is possible and even easy to build an audio system that can never be clipped. However nobody can guarantee that it won't be dissatisfactory tomorrow because it isn't loud enough.
When people go out and buy bigger amps, I've seen you criticise them for making choices that wouldn't necessarily benefit their bottom-line performance. It's all over AVS.
My criticism was not as broad and non-specific as you seem to be pretending here.
Same question - if you've read my answers to these questions on AVS why ask them here? Why ignore the additional details that are relevant?
What people on AVS say is that if the amp isn't audibly clipping, there isn't an audible problem and power is not being limited to the speakers. So since I'm not experiencing audible distortion that I can detect, I am thinking that I'm in the clear and that my amp is big enough for my listening requirements. Do I have reason to think that my assumptions are misguided?
It's all about how sure you want to be. Why is that you quote what others have posted on AVS and throw it up at me when I answered it there, and then seem to pretend that I didn't post more specific answers?
Logic says that if someone willfully ignores what I wrote again and again on a certain topic at AVS, that is their life's pattern and they will do the same thing here. We have other examples here of people who post on AVS who are doing this right now. Therefore, there is no logical reason for me to allow myself to be bullied into repeating myself yet again. Now if there are specfiic questions that cite those statement, of course I would be pleased to expand on or explain. This is one reason why I have decreased my activities on certain forums - I grow tired of repeating the same-old, same-old again and again.
The time domain and frequency domain are perfectly related to each other by a number of different mathematical means, the most common of which is Fourier analysis. If an artifact exists in the time domain then it also exists in the frequency domain.
Thanks, but I know this. I am familiar with the use of Fourier transforms and the equivalence of the frequency and time domains - my own DSP-based audio system uses it extensively. But I also know that a room's contribution is not simply an indiscriminate "phase shift" regardless of whether the signal is a steady tone or a transient. Therefore it does not follow that it will automatically mask phase distortion in the speaker system.
This is different from a speaker that has a flat frequency response, but arbitrary phase distortion.
How?
Because the 'equation' between the original recording and what reaches the listener's ears is broken. The room didn't break it, because the delays, phase shifts and amplitude variations it introduced are all linked consistently. You may say "We can't possibly hear the difference. Let's do a scientific experiment to show it". But I'd rather just use technology (DSP) to bypass the question all together and have zero phase shift contributed by the speakers (or as close as possible).
(3) Avoid using accepted technology for designing the enclosure, picking drivers and designing the crossover.
That's the point I really don't get. There is free software for this. Data for a basic simulation is entered quickly.
If the simulation already comes up with dubious results then continue with that design and if the implementation is even worse, then continue and sell the product with the justification that "speaker design is art after all"?
... if there is a simulation at all.
Or do they blindly trust simplistic simulations without ever measuring the implementation?
It's all about how sure you want to be. Why is that you quote what others have posted on AVS and throw it up at me when I answered it there, and then seem to pretend that I didn't post more specific answers?
Logic says that if someone willfully ignores what I wrote again and again on a certain topic at AVS, that is their life's pattern and they will do the same thing here. We have other examples here of people who post on AVS who are doing this right now. Therefore, there is no logical reason for me to allow myself to be bullied into repeating myself yet again. Now if there are specfiic questions that cite those statement, of course I would be pleased to expand on or explain. This is one reason why I have decreased my activities on certain forums - I grow tired of repeating the same-old, same-old again and again.
I don't spend all my time reading all your posts, but I have read some of your replies and in many cases the implication seemed to be that listener did not need the additional amplification based on your reasoning.
So I'm asking you here to help me to reassure myself of this issue. When is it or is it not necessarily to buy a bigger amplifier? What are the tell-tale signs during playback that an amplifier is running out of power?
You asked me earlier why I limit myself to doing technical tests, but the answer should be self-evident. I'm not an EE, I don't have the tools to perform the testing and I lack the expertise to interpret the results, so I'm not in a position to test amplifier clipping.
My criticism was not as broad and non-specific as you seem to be pretending here.
Same question - if you've read my answers to these questions on AVS why ask them here? Why ignore the additional details that are relevant?
Because those details are vague and locked inside your brain. I lack the power to determine what you mean to say instead of what you say, and sure, I could run back and forth and ask Arnold 5 months ago on a thread what he *really* meant, but the past is the past, so I look to the present.
The only course of action I have is to ask a question in hopes that you will answer in a clear and concise manner with as little obfuscation as possible.
Back to distortion, look at this Focal bookshelf (http://img2.audio.de/Messwerte-f960x960-ffffff-C-ae9b3ed9-110025964.jpg).
In the best case THD is down probably -50 to -60 dB. That's over 0.3% to 0.1%. In the bass it's more like 1% to 2%.
What to expect in the bass range from a single 6" midrange-woofer and plain dome tweeter?
A random floorstanding speaker, same story (http://img1.audio.de/Heco-Celan-GT-902-r960x600-C-8d769dc2-49906877.jpg). (93 dd)
Obviously more bass drivers, but seemingly quite a bit cleaner in the bass range. About the same price, sorta like the Paradigms I mentioned.
A €25,000 floorstanding speaker, worse in the midrange (http://img1.audio.de/KEF-Blade-f960x960-ffffff-C-443ac08f-71243221.jpg). (91 dB)
Seemingly a good example of
not getting what you pay for! Oh, the cabinetry is avant-garde.
A much cheaper floorstanding, cleaner (http://img2.audio.de/image-f960x960-ffffff-C-af5c7254-71886868.jpg). (96 dB according to manufacturer)
As you can see, these speakers don't need much power to reach high SPL and high distortion. Less than 20W @ 4ohm for 100 dB SPL.
American Iron making me a little proud.
Yet another system based on a dome tweeter at the base of a waveguide. Note that the KEF Blade is one too, but something got lost when the price got jacked up more than an order of magnitude, I guess. ;-)
Keep in mind that this is just harmonic distortion. As soon as you play real music you will get loads of intermodultion products.
Both THD and IM can be caused by the same basic nonlinearity. Low THD suggests but does not guarantee low IM. However when the THD is low at frequencies that are well below any measurement bandpass limit, the relationship is usually pretty tight. In this case low THD strongly suggests low IM.
My criticism was not as broad and non-specific as you seem to be pretending here.
Same question - if you've read my answers to these questions on AVS why ask them here? Why ignore the additional details that are relevant?
Because those details are vague and locked inside your brain.
No they are not. They are right there in black and white.
I lack the power to determine what you mean to say instead of what you say, and sure, I could run back and forth and ask Arnold 5 months ago on a thread what he *really* meant, but the past is the past, so I look to the present.
I sense that you aren't trying, and are back trying to bully me into doing your bidding.
The only course of action I have is to ask a question in hopes that you will answer in a clear and concise manner with as little obfuscation as possible.
No, you have other far more reasonable courses of action, such as go on AVS, retrieve my recent posts, and actually read them this time.
The time domain and frequency domain are perfectly related to each other by a number of different mathematical means, the most common of which is Fourier analysis. If an artifact exists in the time domain then it also exists in the frequency domain.
Thanks, but I know this. I am familiar with the use of Fourier transforms and the equivalence of the frequency and time domains - my own DSP-based audio system uses it extensively. But I also know that a room's contribution is not simply an indiscriminate "phase shift" regardless of whether the signal is a steady tone or a transient. Therefore it does not follow that it will automatically mask phase distortion in the speaker system.
It is a time delay that manifests itself as a phase shift. Sometimes it is easier to look at it as a time delay, and sometimes it is easier to look as a phase shift. If the phase shift is < 180 degrees over a wide range then looking at it as a phase shift is usually the more helpful approach.
This is different from a speaker that has a flat frequency response, but arbitrary phase distortion.
How?
Because the 'equation' between the original recording and what reaches the listener's ears is broken. The room didn't break it, because the delays, phase shifts and amplitude variations it introduced are all linked consistently.
Not so much. Consistently wrong and unhelpful is still wrong and unhelpful.
You may say "We can't possibly hear the difference. Let's do a scientific experiment to show it". But I'd rather just use technology (DSP) to bypass the question all together and have zero phase shift contributed by the speakers (or as close as possible).
Zeroing out phase shift is a well known fool's journey. Not because it is inaudible, but because it is an impossible goal. Moderation in all things!
(3) Avoid using accepted technology for designing the enclosure, picking drivers and designing the crossover.
That's the point I really don't get. There is free software for this. Data for a basic simulation is entered quickly.
If the simulation already comes up with dubious results then continue with that design and if the implementation is even worse, then continue and sell the product with the justification that "speaker design is art after all"?
... if there is a simulation at all.
Or do they blindly trust simplistic simulations without ever measuring the implementation?
IME there is nothing like getting 4 or 5 figures for a 2 or 3 figure loudspeaker to convince someone that they are doing everything right. ;-)
The Eggleston looks to me like a legacy design where the driver manufacturer moved on and left his builders behind, but they were getting a nice cash flow and didn't want to mess up a good thing.
I sense that you aren't trying, and are back trying to bully me into doing your bidding.
Perhaps I don't understand the meaning of your posts back when you posted and require clarification. Is it so difficult to ask you to be more clear about this particular topic? I asked you some clear questions hoping you would lay my fears to rest. Why you would ask me to grovel at your feet, I don't know.
Both THD and IM can be caused by the same basic nonlinearity.
Sure, but there can be port nonlinearity, doppler effects, and other nonlinearities that only show up with multiple frequencies.
What I'm trying to say is that you probably never need to worry about amplifier IMD (either), if the amp is not broken.
Back to distortion, look at this Focal bookshelf (http://img2.audio.de/Messwerte-f960x960-ffffff-C-ae9b3ed9-110025964.jpg).
In the best case THD is down probably -50 to -60 dB. That's over 0.3% to 0.1%. In the bass it's more like 1% to 2%.
A random floorstanding speaker, same story (http://img1.audio.de/Heco-Celan-GT-902-r960x600-C-8d769dc2-49906877.jpg). (93 dB)
A €25,000 floorstanding speaker, worse in the midrange (http://img1.audio.de/KEF-Blade-f960x960-ffffff-C-443ac08f-71243221.jpg). (91 dB)
A much cheaper floorstanding, cleaner (http://img2.audio.de/image-f960x960-ffffff-C-af5c7254-71886868.jpg). (96 dB according to manufacturer)
As you can see, these speakers don't need much power to reach high SPL and high distortion. Less than 20W @ 4ohm for 100 dB SPL.
Keep in mind that this is just harmonic distortion. As soon as you play real music you will get loads of intermodultion products.
Please forgive me, but can you explain how I am supposed to read the graphs you provided? These are distortion graphs, or frequency response graphs?
The 25000 speaker is the KEF Blade. So what exactly am I looking at with that graph? You say the distortion is worse in the mid-range? What about the frequency response and low-frequency dynamics? Is it not any good?
Normally when you look at a speaker line, you think the cheapest speaker within that line offers the most basic performance. That a more expensive speaker within the same line offers better performance until you reach the flagship. Is it possible that a lesser speaker in the same line offers improved measured performance?
Both THD and IM can be caused by the same basic nonlinearity.
Sure, but there can be port nonlinearity, doppler effects, and other nonlinearities that only show up with multiple frequencies.
AFAIK port nonlinearity shows up as THD as well as IM. IME it is easy enough to hear with a single tone.
Doppler is not a traditional nonlinearity (motor, suspension) and it is different enough from them that it deserves a discussion of its own. Yes. strictly speaking it is IM. But it is not a consequence of traditional nonlinearities.
The perception of Doppler is also different. For example, masking usually doesn't explain when it is audible and when it is not.
BTW I allowed for Doppler when I said: "Both THD and IM
can be caused by the same basic nonlinearity."
What I'm trying to say is that you probably never need to worry about amplifier IMD (either), if the amp is not broken.
...or clipping! ;-)
Zeroing out phase shift is a well known fool's journey. Not because it is inaudible, but because it is an impossible goal. Moderation in all things!
Oh well, I'll dial some phase distortion back in, then! After all the science, it comes down to the old audio saying: "Now don' ee be getting ideerzz in yurr 'ed about that there 'liminayding phase dizz'torshon. The old ways is best, an' allus shall bee."
Please forgive me, but can you explain how I am supposed to read the graphs you provided? These are distortion graphs, or frequency response graphs?
The 25000 speaker is the KEF Blade. So what exactly am I looking at with that graph? You say the distortion is worse in the mid-range? What about the frequency response and low-frequency dynamics? Is it not any good?
The sore thumb is the red line at the bottom of the page that peaks at 75 dB at about 700 Hz.
Looks to me like the tweeter is being over stressed, probably by a tweeter crossover with too gentle of a slope at its low end.
I think that AJ may have even alluded to this in his comments about the Q family because the Blade seems to be yet another Q-family member.
The Q family has another problem due to the flare of the waveguide for the tweeter being the woofer cone which is potentially moving fast and far enough to cause trouble. Another example of why subwoofers are a good idea.
Normally when you look at a speaker line, you think the cheapest speaker within that line offers the most basic performance. That a more expensive speaker within the same line offers better performance until you reach the flagship. Is it possible that a lesser speaker in the same line offers improved measured performance?
All things are possible, but your basic idea is a reasonable expectation.
One of the reasons for testing is to make sure that something unreasonable is not going on, because unreasonable stuff still happens.
A speaker that I thought sounded very nice to me was the KEF LS50. Please review the technical measurements and tell me if this is considered to be a good performer on the bench.
http://www.soundstagenetwork.com/index.php...&Itemid=153 (http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=941:nrc-measurements-kef-ls50-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153)
We could follow scientific procedure to attempt to define what is funny. It might be interesting, and maybe not worthless knowledge, but it wouldn't be science.
Luckily we are interested in whether audio gear produces audible artifacts, not whether it is funny.
A speaker that I thought sounded very nice to me was the KEF LS50. Please review the technical measurements and tell me if this is considered to be a good performer on the bench.
http://www.soundstagenetwork.com/index.php...&Itemid=153 (http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=941:nrc-measurements-kef-ls50-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153)
Yet another KEF Q-series device, but very pricey. Given your previous comments about your goals for SPL. not a bad choice. Needs a subwoofer.
A speaker that I thought sounded very nice to me was the KEF LS50. Please review the technical measurements and tell me if this is considered to be a good performer on the bench.
http://www.soundstagenetwork.com/index.php...&Itemid=153 (http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=941:nrc-measurements-kef-ls50-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153)
Yet another KEF Q-series device, but very pricey. Given your previous comments about your goals for SPL. not a bad choice. Needs a subwoofer.
I own Paradigm Studio 20 v5 bookshelves in one of my rooms. I do use a subwoofer to augment the low-end. I don't know how well they perform on the bench or if you guys would consider it to be a high fidelity speaker.
What instrument produces a single clean sine wave?
IME a well-played French Horn can come pretty close.
These are my speakers :
http://www.stereophile.com/content/paradig...-v5-loudspeaker (http://www.stereophile.com/content/paradigm-reference-studio-20-v5-loudspeaker)
I own Paradigm Studio 20 v5 bookshelves in one of my rooms. I do use a subwoofer to augment the low-end. I don't know how well they perform on the bench or if you guys would consider it to be a high fidelity speaker.
http://www.soundstagemagazine.com/measurem...gm_studio20_v3/ (http://www.soundstagemagazine.com/measurements/paradigm_studio20_v3/)
As has been said here before in the past week or so, significant performance differences between submodels are generally pretty small, often just cosmetic.
We could take this as clear evidence that some people think that 10 seconds worth of Googling is beneath their dignity.
These are my speakers :
http://www.stereophile.com/content/paradig...-v5-loudspeaker (http://www.stereophile.com/content/paradigm-reference-studio-20-v5-loudspeaker)
Street price seems to run from $600 to 900 each.
Begs comparison with:
http://www.stereophile.com/content/infinit...er-measurements (http://www.stereophile.com/content/infinity-primus-p162-loudspeaker-measurements)
Which I've paid as little as $100 each for.
Please compare and contrast. I'll give you that the Paradigms are prettier looking.
@LS50: Well, you can't expect undistorted low frequency output at high SPL with a small woofer in a small box and there's the problem that Arnold mentioned before (but I doubt the little woofer is moving much).
Almost all of the LF comes from the rear port.
With that in mind, the performance seems to be pretty good.
@LS50: Well, you can't expect undistorted low frequency output at high SPL with a small woofer in a small box and there's the problem that Arnold mentioned before (but I doubt the little woofer is moving much).
Almost all of the LF comes from the rear port.
With that in mind, the performance seems to be pretty good.
How do my speakers fare (Paradigm Studio 20 v5)?
Have you anyting else to do (http://www.hydrogenaud.io/forums/index.php?s=&showtopic=106825&view=findpost&p=874805)
Have you anyting else to do (http://www.hydrogenaud.io/forums/index.php?s=&showtopic=106825&view=findpost&p=874805)
I have the v5. I linked to a set of measurements of the v5 and I require your assistance to interpret what is being measured.
A speaker that I thought sounded very nice to me was the KEF LS50. Please review the technical measurements and tell me if this is considered to be a good performer on the bench.
http://www.soundstagenetwork.com/index.php...&Itemid=153 (http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=941:nrc-measurements-kef-ls50-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153)
Yet another KEF Q-series device, but very pricey. Given your previous comments about your goals for SPL. not a bad choice. Needs a subwoofer.
Lower than average sensitivity, though. From Stereophile, "85dB/2.83V/m."
KEF do an active speaker with what looks like similar drivers - the X300A - which is cheaper.
It was said earlier that when amps clip they run out of voltage. But what about current? I thought voltage x current = power and an amp that runs out of power begins to clip.
Another thing I wanted to know is what do manufactures mean when they specify an amp at a minimum recommended impedance. Like they'll say minimum 6 ohm or 6 ohm compatible. What does that mean? Or is that just a guideline? Does it matter when playing music?
It was said earlier that when amps clip they run out of voltage. But what about current? I thought voltage x current = power and an amp that runs out of power begins to clip.
The relationship between voltage, current, and impedance is defined by
Ohm's Law (http://en.wikipedia.org/wiki/Ohm's_law).
If your amp can't supply the current, the voltage will also drop. i.e. An oscilloscope measures voltage, and you'll see the voltage clipping if your amplifier clips due to lack of current.
Amplifiers are considered "constant voltage" devices. Obviously, the voltage isn't
truly constant because normal audio is a constantly-varying signal that depends on the program material and the volume setting. In this context, "constant voltage", means that the output voltage does not depend on impedance (or current). That is, you'll get the same voltage with 4-Ohm speakers, 8-Ohm speakers, or no speakers (infinite impedance), as long as the amplifier is operating properly within it's specified limits, etc.
Another thing I wanted to know is what do manufactures mean when they specify an amp at a minimum recommended impedance. Like they'll say minimum 6 ohm or 6 ohm compatible. What does that mean? Or is that just a guideline? Does it matter when playing music?
It's possible to damage an amplifier (by drawing too much power & current) if the impedance is too low. Or, you might just get distortion (clipping) at lower than rated power.
In accordance with Ohm's law, current is inversely proportional to impedance.
If you cut the impedance in half (with the same voltage) you'll get twice the current and twice the power. In fact, you'll find that many amplifiers are rated at twice the power with 4 Ohm speakers. If an amplifier isn't getting twice the power at 4-Ohms, it's being current-limited (at 4 Ohms). Or, it might be thermally limited with thermal protection kicking-in if you "pull" too much power with lower impedance.
BTW - If you double the voltage you also double the current (assuming constant impedance), which means doubling the voltage gives you 4 times the power.
If you wire a pair of speakers in parallel, you cut the impedance in half and double the power (if the amp can provide the current). For example, say you have a speaker connected and it's getting 100W. If you connect a 2nd speaker with the same impedance, it will also get 100W.
This is similar to plugging-in a 2nd light bulb to the wall socket... You get twice the current and consume twice the power. If you try to get too much current the breaker will "blow". And, you might get a slight voltage drop if you "pull" a lot of power from the socket.
High power car stereo amplifiers are sometimes rated down to 1 or 2 Ohms so you can connect two or more 4 Ohm woofers.
It was said earlier that when amps clip they run out of voltage.
Not at all. Trying to help some people seems to be like diving into a sewer because they incessantly misrepresent what you say. Are they merely incapable of remembering what was said 5 minutes ago, or are they being malevolent?
Here's what I actually said:
The most common source of clipping in actual use is that the amp is being called on to provide more energy at any particular moment than it is capable of providing without clipping. The reasons for that vary with the situation, but in general clipping is caused by the limited ability of the amplifier to produce the required current and voltage. The most commonly exceeded limit relates to voltage.
But what about current?
Just stated in clear English:
"...in general clipping is caused by the limited ability of the amplifier to produce the required current and voltage.
Translation from clear English to hopefully clearer English:
It is possible that the amplifier is clipping due to its limited ability to produce the required current or voltage or both.
I tried to be helpful by adding
"The most commonly exceeded limit relates to voltage"
which means that the limit could be due to a lack of voltage or current or both, but the most likely is a lacking in voltage.
Another thing I wanted to know is what do manufactures mean when they specify an amp at a minimum recommended impedance. Like they'll say minimum 6 ohm or 6 ohm compatible. What does that mean? Or is that just a guideline? Does it matter when playing music?
It is clearly a guideline. In general it does not matter when playing music for 2 reasons:
(1) The crest factor of music is high, and power amp ratings are based on signals with low crest factor. High crest factor signals are easy on amps.
(2) Most people's preferences (like apparently you!) for music are for less than 105 dB SPL peaks.
Just stated in clear English:
"...in general clipping is caused by the limited ability of the amplifier to produce the required current and voltage.
Translation from clear English to hopefully clearer English:
It is possible that the amplifier is clipping due to its limited ability to produce the required current or voltage or both.
I tried to be helpful by adding
"The most commonly exceeded limit relates to voltage"
which means that the limit could be due to a lack of voltage or current or both, but the most likely is a lacking in voltage.
Yes, but the emphasis was placed on voltage. I was trying to understand why you placed emphasis on voltage, as opposed to current. Like you say that clipping is most likely when the amp is lacking in voltage, so now why is that? As opposed to a lack of current?
Just stated in clear English:
"...in general clipping is caused by the limited ability of the amplifier to produce the required current and voltage.
Translation from clear English to hopefully clearer English:
It is possible that the amplifier is clipping due to its limited ability to produce the required current or voltage or both.
I tried to be helpful by adding
"The most commonly exceeded limit relates to voltage"
which means that the limit could be due to a lack of voltage or current or both, but the most likely is a lacking in voltage.
Yes, but the emphasis was placed on voltage. I was trying to understand why you placed emphasis on voltage, as opposed to current. Like you say that clipping is most likely when the amp is lacking in voltage, so now why is that? As opposed to a lack of current?
Using some simplifications in this explanation: transistors are current-operated devices and high-current versions are easily made and relatively cheap. However, the output transistors have to be rated to withstand the supply voltage of the output stage, which is what governs the absolute maximum voltage the amp can deliver. High-voltage transistors are expensive.
Plus, the power supply will need to be beefed up since higher voltage also means more current will be needed for a given impedance load (power goes up with the square of the voltage, given a fixed load impedance) so even more expense.
In other words, it's largely about money and compromise. Doubling the output voltage will quadruple the power and cost (maybe) six or more times as much.
Of course,
on the market price is no good indicator of either performance or power.
The Jeff Rowland Model 10 amp, for example, used a bunch of $6 ICs at its core for the amplification (150W into 8 ohms continuous). The retail price was close to $7,000. Same with the Concerta II.
I wonder if those reviews would have been so positive had the reviewers been told beforehand that the amplification is done by a bunch of LM3886 (http://www.ti.com/product/lm3886)'s.
Random example (http://www.audioasylum.com/forums/amp/messages/7578.html):
Tremendous transparency. Excellent musicality.... vocals were ooh so natural. Dynamics, I wasn't expecting this but these amps were really dynamic. Transient response was top-class, the snap of percussion was as good as I've heard in my (any?) system.
Yes, but the emphasis was placed on voltage. I was trying to understand why you placed emphasis on voltage, as opposed to current. Like you say that clipping is most likely when the amp is lacking in voltage, so now why is that? As opposed to a lack of current?
As has been pointed out, it is the natural tendency of a SS amp to deliver massive amounts of current to a low impedance load to the point of rapidly (fractions of a second) destroying itself though excess internal heating of its output devices. In the early days SS amps didn't manage this properly and many of them failed. Shorted speaker leads could be fatal. A rubbing voice coil could be fatal. Too many speakers in parallel could be fatal. This was quickly addressed by adding current limiting circuitry. It was then discovered that current limiting circuitry often created problems of its own by activating unexpectedly and by the very nature of its operation.
Clipping due to an amp running out of power supply voltage tends to be pretty clean - it turns music into flat topped waves or waves that have only a low frequency wave riding on them from the power line. The output devices are at the time in a low impedance state with a low voltage being dropped across them, and the load simply sees a signal that stops increasing at some arbitrary point which it sees very frequently because that what musical signals do. The amp and the load recover pretty quickly (microseconds).
Clipping due to current limiting can be messy. The output devices are put into a high impedance state and an attempt is made to cut current to the load. If the load is inductive (all speakers are at least somewhat inductive over much of the audio range) then the load reacts by increasing the voltage across it by drawing energy from its collapsing magnetic field to maintain the current flow constant. This inductive kick can reach 100's of volts (the same principle is exploited in automotive ignition systems) and destroy many components in both the amplifier and the speaker itself. Fortunately most amplifiers are designed to limit this kick to the supply voltage of the output stage, but you can see how messy this can be. The recovery time may stretch into milliseconds and it can be disruptive to the music.
In the 80s and 90s some high end designers built amplifiers with literally dozens of big, shiny output transistors per channel and these amplifiers were nearly impossible to cause to activate their protective circuits. They looked very impressive and in fact they had fantastic current reserves, but with ordinary speakers the shiny metal output transistor and massive heat sink bling did nothing for sound quality. If you had a truly pathological speaker, they were just what you needed.
Therefore clipping due to current limiting is avoided where possible by choice of output devices, which has been increasingly more economical due to the continuing advances in SS technology. Current/voltage capabilities that used to require a goodly number of devices with heavy duty metal cases in series/parallel can now be handled by a single far smaller device encapsulated in plastic. A single Y2K plastic transistor can approximate the current capacity of maybe a half-dozen ca. 1970 output devices.
Speaker technology has also improved. In the 70s and 80s a number of speakers such as the Infinity IRS went to marketwith wildly aggressive impedance curves. They either fried poorly protected amplifiers or activated their protection circuits at relatively low listening levels. In the 90s a number of manufacturers such as NHT recognized that amplifier-friendly speakers were less likely to be returned which made everybody much happier. Today most speaker designers know that low impedances and high reactance don't mix at the same frequency. Low impedance with modest reactance is not necessarily as troublesome as is high reactance and high impedance in combination.
Of course, on the market price is no good indicator of either performance or power.
The Jeff Rowland Model 10 amp, for example, used a bunch of $6 ICs at its core for the amplification (150W into 8 ohms continuous). The retail price was close to $7,000. Same with the Concerta II.
I wonder if those reviews would have been so positive had the reviewers been told beforehand that the amplification is done by a bunch of LM3886 (http://www.ti.com/product/lm3886)'s.
Random example (http://www.audioasylum.com/forums/amp/messages/7578.html):
Tremendous transparency. Excellent musicality.... vocals were ooh so natural. Dynamics, I wasn't expecting this but these amps were really dynamic. Transient response was top-class, the snap of percussion was as good as I've heard in my (any?) system.
In case anybody doubts this, here is a picture of the output circuit card of that amplifier:
(http://cdn.head-fi.org/9/95/350x700px-LL-95ae27d8_model10-2.jpeg)
Clearly shown are LM3886 amplifier chips. I believe that there were 8 per channel which can yield a fairly powerful amp. I see it is rated at 150/275 wpc 8/4 ohms.
http://jeffrowlandgroup.com/Docs/M10Manual.pdf (http://jeffrowlandgroup.com/Docs/M10Manual.pdf)
However the spec sheet for this chip:
http://www.ti.com/lit/ds/symlink/lm3886.pdf (http://www.ti.com/lit/ds/symlink/lm3886.pdf)
(page 11)
shows THD > 0.1% and rapidly rising above 10 KHz, which would be an embarrassment to any number of conventionally designed "PA" amplifiers that high end audio eggspurts love to look down on.
I also find the apparent absence of heat sink attaching screws to be a little appalling. The device is available with an electrically isolated package that would make the mica insulators superfluous. It seems to have mounting hardware it shouldn't need and lacks mounting hardware that it should need. Whatever!
6 per channel. Basically a beefier Gainclone with fancy case.
The performance doesn't seem to be bad and I don't even want to comment on that. But the price...
Therefore clipping due to current limiting is avoided where possible by choice of output devices, which has been increasingly more economical due to the continuing advances in SS technology.
Today most speaker designers know that low impedances and high reactance don't mix at the same frequency. Low impedance with modest reactance is not necessarily as troublesome as is high reactance and high impedance in combination.
A combination of the above two means that any modern day SS amp with 100 wpc and most modern speakers can happily coexist in a typical home listening room without having to lose sleep or obsessing over clipping. And up to a point, watts are cheap, one can even have a 150 wpc amp without spending silly money, which will comfortably drive most modern speakers that will fit in said room aesthetically and sonically.
Yes I am aware of Sean Olives work and not surprisingly, his speakers, endorsed by his company, were preferred the most.
Sean Olive tests the speakers in his own facility, using techniques best served to bring about a result that people prefer, and lo and behold, out of all the speakers tested, his speakers were preferred.
Nothing at all suspicious about that.
A speaker that I thought sounded very nice to me was the KEF LS50.
I own Paradigm Studio 20 v5 bookshelves in one of my rooms.
Pssst, hey Rich, just a little FYI to fill some ummm, small gaps in your knowledge.
The man whose character you impugn, actually worked at the NRC (under Toole IIRC) prior to working at the "preferred speaker company". Though less known, KEFs Eureka/Archimedes project on the other side of the pond found largely the same thing, independently.
Both speakers you like above, are cut from the exact same cloth, the one you own, literally from the SO book.
You're either just plain naive, or just in denial.
I have long remarked that I haven't seen any issue on audio boards that can't be explained by Dunning-Kruger.
Party on Garth...
cheers,
AJ
Oh and you 3 bling fashion speakers you "pictured", all would fail that school, rather badly.
Yes, I can tell from just the pictures.
However the spec sheet for this chip shows THD > 0.1% and rapidly rising above 10 KHz...
Which graph are we looking at..?
However the spec sheet for this chip shows THD > 0.1% and rapidly rising above 10 KHz...
Which graph are we looking at..?
Good point. Overlooked a leading zero. The THD graphs in
http://www.ti.com/lit/ds/symlink/lm3886.pdf (http://www.ti.com/lit/ds/symlink/lm3886.pdf) page 11, figure 16- 18 are just fine, but not exceptional.
The lesson there is that amp designers (both high end and otherwise) that make a big thing out of using all discrete power amps are sending up smoke signals.
Basically, as soon as someone says "Phase is inaudible with music. End of."
Not sure who said that, but it's largely irrelevant.
You wouldn't know this, but it's a logical fallacy to (attempt to) prove a negative.
IOW, "proving" phase inaudibility absolutely, is a fools errand.
Rather, the onus falls squarely on you GM, to prove "phase" audibility, with whatever you think you are doing with you limited knowledge set....and mic pointed straight in front of your speaker at some distance.
Why don't you explain how you did it here:
Lets see what "phase" you are talking about, what "distortion" exists in passives and exactly how GM is "hearing" this. Your methods should provide some fascinating reading.
Oh and by all means, don't be ashamed to post a pic of the studiophiles room too.
cheers,
AJ
^ Could definitely use some more screws on the front.
^ Could definitely use some more screws on the front.
They look tight. He's definitely no audiophile (http://www.normankoren.com/Audio/)
Basically, as soon as someone says "Phase is inaudible with music. End of." my sceptical antennae twitch! Much more information is needed, and even then it cannot tell the whole story.
A statement like "Phase is inaudible" is an excluded middle statement. Phase is audible under some conditions, not others.
Phase shifts applied differently to different channels gets to be easy to hear in modest amounts.
Phase shifts applied identically the same to different channels gets to be hard to hear even in seemingly huge amounts.
Since we generally don't listen to speakers with our ears pressed up against the speaker cones, there is always a lot of phase shift due to the distance, so we have to be careful about talking about phase shift versus group delay which takes the delay due to distance out of the discussion pretty well.
Then we need to distinguish between phase and polarity, given that they are often conflated and are really two different things.
Finally, phase shifts can be audible with certain pathological signals, while completely slipping by when the sounds are the usual music and drama.
Oversimplified statements have this nasty tendency to be always wrong, so falsifying them is often a cheap shot.
I'll tell my life a little, sorry for that.
when I started using EQ some years back , I soon realized I could somehow end up with changes in the "imaging", a guitar would move upfront a little simply because I played with the midrange or some stuff like that. not knowing what caused it, I looked for information about EQ and most people I talked to(pre-internet was a bitch) were bashing EQ about phase shift. not finding better and mostly not knowing anything, I spent some times thinking that phase shifts were the one cause of my changes in spatial cues. I still read that often, so I guess a lot of people listened to the same dudes ranting about how bad EQ was.
since I've learned a little more about psycho acoustic and human hearing in general, how high freqs are easy to pinpoint compared to lower ones, how the shape of our ears changes the frequency response as a way to know up from down and other funky tricks. I came to realize that as soon as the FR was changing it was bound to affect everything else in the way we perceive music.
and that's where I stand now, not 100% sure about much, but clearly a lot less obsessed with phase.
just to say that when we don't know much, we tend to pick the first presented thing that remotely looks like a solution and cling onto it. it does seem like phase(and lately jitter) is one of those thing. it's here, it does unclear stuff, maybe. let me blame it for whatever I don't understand.
^ Could definitely use some more screws on the front.
They look tight. He's definitely no audiophile (http://www.normankoren.com/Audio/)
Oh, that's a keeper. But it took me a minute to see that -- so just in case anyone else missed it on that busy page:
Loose screws? Sometime in the late 1980's I visited an audiophile in San Diego. When he opened the door he exclaimed, "Norm! I've made the most incredible discovery! A simple modification that really improves the sound— opens it up, sharpens the imaging, and deepens the soundstage. I've loosened the screws on all my chassis. You gotta hear it!" He learned this tweak from a friend of his who had made a small fortune with a "CD enhancer" spray called Finyl. At the time, the ads for Finyl claimed it would increase the number of bits of resolution. His system sounded terrible.
I'd heard rumors that some audiophiles had their screws loose, but I never took them literally until that evening.
I'll tell my life a little, sorry for that.
when I started using EQ some years back , I soon realized I could somehow end up with changes in the "imaging", a guitar would move upfront a little simply because I played with the midrange or some stuff like that. not knowing what caused it, I looked for information about EQ and most people I talked to(pre-internet was a bitch) were bashing EQ about phase shift. not finding better and mostly not knowing anything, I spent some times thinking that phase shifts were the one cause of my changes in spatial cues. I still read that often, so I guess a lot of people listened to the same dudes ranting about how bad EQ was.
since I've learned a little more about psycho acoustic and human hearing in general, how high freqs are easy to pinpoint compared to lower ones, how the shape of our ears changes the frequency response as a way to know up from down and other funky tricks. I came to realize that as soon as the FR was changing it was bound to affect everything else in the way we perceive music.
and that's where I stand now, not 100% sure about much, but clearly a lot less obsessed with phase.
just to say that when we don't know much, we tend to pick the first presented thing that remotely looks like a solution and cling onto it. it does seem like phase(and lately jitter) is one of those thing. it's here, it does unclear stuff, maybe. let me blame it for whatever I don't understand.
But if you have the opportunity to minimise phase distortion from your speakers..? Does it matter whether or not some research (funded by a passive speaker manufacturer maybe?) seems to shows that "phase isn't important (for timbre in mono)"? It
is an arbitrary distortion of the signal that can be minimised with DSP.
Are people really arguing that phase isn't important, or are is it that the very thought of using DSP is so abhorrent to them that passive speakers just
have to be adequate?
('Hilarious' that people can't see past the aesthetics of a pair of prototype speakers!)
But if you have the opportunity to minimise phase distortion from your speakers..? Does it matter whether or not some research (funded by a passive speaker manufacturer maybe?) seems to shows that "phase isn't important (for timbre in mono)"? It is an arbitrary distortion of the signal that can be minimised with DSP.
Are people really arguing that phase isn't important, or are is it that the very thought of using DSP is so abhorrent to them that passive speakers just have to be adequate?
Hey, I'm arguing all the way for higher fidelity, even if we're past some audible thresholds (true high end). I think most here are interested in what we can achieve technologically.
But claims about the audibility of various things, including phase, are often grossly exaggerated.
You gotta keep all the things in perspective. If you can easily correct the phase response then please do it, but its effects will likely be minimal to inaudible if there are problems with much more important things: the recording, the speakers, positioning, the room ...
I wouldn't tell someone with such problems that he has to fix the phase response.
Similarly I wouldn't tell someone to buy a better amp to reduce distortion by 0.5% when the magnitude response is all over the place.
('Hilarious' that people can't see past the aesthetics of a pair of prototype speakers!)
Was that directed at me? I thought it was pretty clear that my comment was not serious.
('Hilarious' that people can't see past the aesthetics of a pair of prototype speakers!)
Was that directed at me? I thought it was pretty clear that my comment was not serious.
Sorry xnor. I had a sense of humour lapse.
I screwed the fronts on (to recycled old speaker enclosures) simply so I could experiment with different configurations. However, interestingly, I notice that some manufacturers of BBC monitor type speakers deliberately screw the rear panels on rather than gluing them, as part of the thin wall 'lossy' philosophy.
Was that directed at me? I thought it was pretty clear that my comment was not serious.
Probably didn't like the phase of your response.
The "hilarious" part is GMs talk about "phase" when he can't define it and talk about "hearing" it, when he "fixed" that which "plagues passives", with his own active speakers.
If GM could show specifically what he did and how he determined it lifting veils and whatnot, it would seem more like a HA discussion than an AA one.
I think we both already know why he can't.
No books or internets.
Perhaps he's an anarchist?
cheers,
AJ
Hey, I'm arguing all the way for higher fidelity, even if we're past some audible thresholds (true high end). I think most here are interested in what we can achieve technologically.
I am not an engineer, but isn't the pursuit of something in audio engineering that is beyond audible thresholds, "bad" engineering by an engineer's definition? What's the point of it?
I am not an engineer, but isn't the pursuit of something in audio engineering that is beyond audible thresholds, "bad" engineering by an engineer's definition? What's the point of it?
Because there is no fixed audible threshold, and you want to build in as much margin as possible?
Some engineering, if it's 'only software' can be free to implement, in which case it's not necessarily bad engineering to throw it in, anyway.
I am not an engineer, but isn't the pursuit of something in audio engineering that is beyond audible thresholds, "bad" engineering by an engineer's definition? What's the point of it?
Depends on the extra cost needed to achieve the improvements, if any, and the budget. If some design lets you realize optimizations cheaply and easily, you'd be a fool not to make use of it.
Consider improvements beyond the established thresholds as safety margin, headroom, and finally plain overkill (for bragging rights).
Also, technological advancements that may seem like total overkill can trickle down so they become affordable for ordinary mortals, to the point where you achieve much better performance for less money.
Anyway, I've clearly rubbed a couple of people up the wrong way, for which I apologise. It's always going to be tricky when being enthusiastic about one thing (in my case DSP and active speakers) is implicitly a criticism of someone else's enthusiasms.
I'm wondering why audiophiles haven't long made the move to active speakers. Well, actually I'm not since they cling to all kinds of old stuff.
I guess the main problem is that they'd need 2 to 3 times the cables from their heavy aluminum blocks (amps) to the speakers, the interconnects, the digital cables ... considering the price of audiophile cables. That would probably cost more than their whole system does now.
I'm wondering why audiophiles haven't long made the move to active speakers. Well, actually I'm not since they cling to all kinds of old stuff.
I guess the main problem is that they'd need 2 to 3 times the cables from their heavy aluminum blocks (amps) to the speakers, the interconnects, the digital cables ... considering the price of audiophile cables. That would probably cost more than their whole system does now.
I'm sure you know that isn't necessary, a single low level signal cable is all that is needed if the amps are internal, as most commercial stuff (mainly pro) is. Plus an AC cord of course.
Here's one I just finished this week, for a Canadian music lover, not an audiophile of course.
(http://i216.photobucket.com/albums/cc73/AJinFLA/More%20Speaker%20pics/CustomM2A-rear.jpg)
Their choice for convenience to go active, not due to any irrational, misinformed opinions about passive networks.
I definitely understand why those with limited knowledge would fear them however.
cheers,
AJ
True audiophiles would never get rid of their external DACs and amps that they so carefully "matched" to their "chain". Especially not if the replacements are integrated. That's blasphemy!
True audiophiles would never get rid of their external DACs and amps that they so carefully "matched" to their "chain". Especially not if the replacements are integrated. That's blasphemy!
Indeed, read a review of a £1200 Chord Hugo and the reviewers couldn't understand why I wouldn't spend £1200 on it and was perfectly happy with my FiiO E7
I'm wondering why audiophiles haven't long made the move to active speakers. Well, actually I'm not since they cling to all kinds of old stuff.
I guess the main problem is that they'd need 2 to 3 times the cables from their heavy aluminum blocks (amps) to the speakers, the interconnects, the digital cables ... considering the price of audiophile cables. That would probably cost more than their whole system does now.
I would think *fewer* audio cables, thanks to internal amping. But more electrical power cables.
Honestly for me it's just simple laziness and inertia about how to add 'room correction' DSP and DPL II to an all-active loudspeaker system.
I'll probably get around to it some day ...
I'm wondering why audiophiles haven't long made the move to active speakers. Well, actually I'm not since they cling to all kinds of old stuff.
I guess the main problem is that they'd need 2 to 3 times the cables from their heavy aluminum blocks (amps) to the speakers, the interconnects, the digital cables ... considering the price of audiophile cables. That would probably cost more than their whole system does now.
I would think *fewer* audio cables, thanks to internal amping. But more electrical power cables.
Honestly for me it's just simple laziness and inertia about how to add 'room correction' DSP and DPL II to an all-active loudspeaker system.
I'll probably get around to it some day ...
I circumvented that problem by devising a high performance speaker-to-line level adaptor based on an inexpensive direct box.
(http://ecx.images-amazon.com/images/I/61nVeYVPy2L._SL1000_.jpg)
I replace the two TS jacks with dual banana plugs,
(http://thumbs2.ebaystatic.com/d/l225/m/mEok0INNZWqsvjfvgp7gkyw.jpg)
gut everything but the transformer, and add three power matching resistors.
The existing XLR jack matches the input of typical pro audio active monitors:
(http://www.basic-home-recording-studio.com/images/direct_box_passive.jpg)
Speaking of pro audio monitors, we finally got around to doing some response versus direction testing on LSR 308s.
The results were phenominal.
What is most interesting is that we can't confirm most of these:
(http://www.prosoundnetwork.com/portals/0/PRO_12_13_Final-185.jpg)
The on axis response is pretty close to what we observed and is really impressive, but so were the FR curves for the first 20 degrees off the vertical axis, and 60 degrees off the horizontal axis. We did not seem to see the slow progression in treble roll-off off-axis, but instead saw more like true constant directivity to the indicated angles, at which response dropped off quite rapidly and sometimes unevenly.
Our consensus was that the LSR 308 is a true next-generation improvement on the Behringer Truths.
Speaking of pro audio monitors, we finally got around to doing some response versus direction testing on LSR 308s.
We??
Will you post them?
What is most interesting is that we can't confirm most of these:
(http://www.prosoundnetwork.com/portals/0/PRO_12_13_Final-185.jpg)
I'm curious about those also. Even if they are averaged at those angles around the speaker, how does an 8" boxed driver get directional above 150hz?
cheers,
AJ
Speaking of pro audio monitors, we finally got around to doing some response versus direction testing on LSR 308s.
We??
Will you post them?
In due time they should be posted here: GedLee PolarMap Display Program + library of existing tests (http://gedlee.azurewebsites.net/Application%20Files/RunPolarMap.aspx)
What is most interesting is that we can't confirm most of these:
(http://www.prosoundnetwork.com/portals/0/PRO_12_13_Final-185.jpg)
I'm curious about those also. Even if they are averaged at those angles around the speaker, how does an 8" boxed driver get directional above 150hz?
There must be a typo in that question because the answer seems way too obvious. ;-)
That would be the:
JBL LSR308
8" Two-Way Powered
Studio Monitor
at about $250 USD each
http://www.jblpro.com/www/products/recordi...08#.VNzexCyOf7M (http://www.jblpro.com/www/products/recording-broadcast/3-series/lsr308#.VNzexCyOf7M)
That would be the:
JBL LSR308
8" Two-Way Powered
Studio Monitor
at about $250 USD each
http://www.jblpro.com/www/products/recordi...08#.VNzexCyOf7M (http://www.jblpro.com/www/products/recording-broadcast/3-series/lsr308#.VNzexCyOf7M)
Exactly, except that a number of us have picked them off of Amazon for $200 each including shipping.
There must be a typo in that question because the answer seems way too obvious. ;-)
No typo. I see the expected directionality >700hz or so for an 8" driver, but 5db down at 300hz 90 degrees off axis? Driver should be omni in this range.
Gedlee Summa, with 15" dia woofer
(http://www.linkwitzlab.com/images/graphics/summa_FR-s.jpeg)
Guess I'll have to wait for Earls measurements of the LSRs.
cheers,
AJ