Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Have you ever regretted ABXing? (Read 45285 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Have you ever regretted ABXing?

Reply #50
No, data about totally empty rooms is *not* relevant, at least until we start generally making recordings in totally empty rooms.  I think there's only one piece of music that could be recorded this way. It was written by John Cage, if memory serves.  ;-)

There are also quieter rooms than those. I think one of them is in the Carlsbad Caverns in New Mexico. If data is going to be relevant it has to have a useful level of generality.


Do you have any reason to believe that the sounds made by the musicians are universally sufficient to mask the dither noise? After all, two of the Meyer-Moran subjects were able to detect the 16 bit noise floor with the peak level set to 115 DB. This despite the listeners' self noise (how's that for a new term?) and the possible presence of others in the room (I don't know enough about the experimental details to know if others were present).

Have you ever regretted ABXing?

Reply #51
Electrostatics tend to have very low measured distortion and unusually good impulse response. I don't know whether and to what extent that may have contributed to the test results. I'm also not aware of any experiments that might confirm or refute their reputation for being unusually revealing of distortion or artifacts in source material, or identify factors that might contribute to that reputation. It sounds like an interesting area for research.


There have been enough mediocre and even  *bad* electrostatic speakers pushed onto the market that trying to characterize electrostatics as a class of Spears as having any unique sonic advantages is very questionable. Even the Quad ESL curves you reference have a rather obvious flaw - lack of what most of us would call bass response. Some of that might be due to the measurement environment. I just can't tell from the accompanying text.  The Quad electro stats of that era also have some pretty strong dynamic range limitations.

It is no accident that the most accurate of currently available loudspeaker systems are usually direct radiators with cone and or dome drivers.  Furthermore accuracy is always strongly limited by the room, and the matching of the speaker system to the room.  So raw speaker system measurements are not sufficiently representative,


The guy who did the Quad measurements writes "Don't be fooled by the ESL57's apparent rolloff below 400 Hz; this is an artifact of the short time window and the Quad's location in the room." I found three other measurements, one showing a (lesser) bass rolloff, and two showing a rising bass response:

http://quadesl.nl/img/Stereoplay_June_1978.jpg
http://quadesl.nl/img/esl57_response_graph.jpg
http://quadesl.nl/img/Hifi-Stereophonie.jpg

Which is to say, who knows?

They definitely had serious dynamic range limitations.

Otherwise, yeah, your point is well taken: there are good and bad electrostatics. I suppose I should have qualified my statement, since I referred generically to electrostatics and headphones as classes of transducers that might potentially affect the outcome of a test. Artifact audibility could potentially depend on the specific design of the transducer, and the acoustics of the space in which its located. -- Are people then justified in suggesting that electrostatics, as a class, are unusually revealing of certain artifacts? I have little objective evidence of that, just the digital/analog comparison test to which I referred. However, measurements do suggest that well-designed members of the breed might be less likely than most other speaker types to mask low levels of non-linear distortion, and, perhaps, other forms of distortion, as well. Also, some psychoacoustic phenomena appear to have a lower threshold of audibility on headphones than on speakers, so perhaps the same is true of certain headphones as well.

The point I think is that some audible characteristics may have a lower threshold of audibility on certain transducers, including, it seems, especially bad ones, and that that possibility has to be taken into account in interpreting the significance of a negative result.

 

Have you ever regretted ABXing?

Reply #52
No, data about totally empty rooms is *not* relevant, at least until we start generally making recordings in totally empty rooms.  I think there's only one piece of music that could be recorded this way. It was written by John Cage, if memory serves.  ;-)

There are also quieter rooms than those. I think one of them is in the Carlsbad Caverns in New Mexico. If data is going to be relevant it has to have a useful level of generality.


Do you have any reason to believe that the sounds made by the musicians are universally sufficient to mask the dither noise?


Die to the high levels of generality indicted above ("universally sufficient") the asnwer to the qustion has to be "no".

Quote
After all, two of the Meyer-Moran subjects were able to detect the 16 bit noise floor with the peak level set to 115 DB.


As I read it, the paper describes the 16 bit noise floor as that produced by "conventionally dithered digital audio equipment". I take that to mean spectrally *unshaped* dither. Since the 16 bit (or any other) noise floor can and often is shaped to a very large degree with significant subjectively-perceived benefits, the choice of conditions for characterizing 16 bit encoding used in the paper seems to be unfortunate and perhaps even prejudicial.


Have you ever regretted ABXing?

Reply #53
As I read it, the paper describes the 16 bit noise floor as that produced by "conventionally dithered digital audio equipment". I take that to mean spectrally *unshaped* dither. Since the 16 bit (or any other) noise floor can and often is shaped to a very large degree with significant subjectively-perceived benefits, the choice of conditions for characterizing 16 bit encoding used in the paper seems to be unfortunate and perhaps even prejudicial.


I assumed it was unshaped as well. Fielder concludes that the dynamic range achievable with shaped dither "is comparable to that needed by consumer end use but does not satisfy the strictest requirements of professional use." This qualified endorsement is I assume a consequence of his very interesting measurements of the dynamic range of consumer loudspeakers, since it seems that the quieter home listening rooms don't limit dynamic range.



Have you ever regretted ABXing?

Reply #54
I assumed it was unshaped as well. Fielder concludes that the dynamic range achievable with shaped dither "is comparable to that needed by consumer end use but does not satisfy the strictest requirements of professional use."


Saying that shaped dither "...does not satisfy the strictest requirements of professional use."  is a very interesting statement that in retrospect seems to be very hard to justify.  Of course the paper was from 1994 - 16 years ago, and we didn't know then what we know now, and by a lot! When I predicted in Y2K that DVD-A and SACD would fail in the marketplace, even I was not 100.00% convinced.  Of course now I am. But, its a decade later. So much for my prophet's license! ;-)

Quote
This qualified endorsement is I assume a consequence of his very interesting measurements of the dynamic range of consumer loudspeakers, since it seems that the quieter home listening rooms don't limit dynamic range.


It really comes down to what percentage of consumers that you want to satisfy. I suspect that far less than 0.001% of all listeners actually maintain and routinely use a listening environment that is not well-served by 16 bits done well.  OTOH 0.001% of all listeners is still thousands of people. But thousands of people don't make a market for mainstream products. There's something very democratic about distribution media - everybody feeds at the same trough.

Reality is that much distributed media has squashed dynamic range that probably disappoints 1% or more of all listeners. If you can't get all media to be made to a standard that doesn't disappoint 1%, then the futility of trying to make a business out of pleasing 0.001% is very clear.

Have you ever regretted ABXing?

Reply #55
Saying that shaped dither "...does not satisfy the strictest requirements of professional use."  is a very interesting statement that in retrospect seems to be very hard to justify.  Of course the paper was from 1994 - 16 years ago, and we didn't know then what we know now, and by a lot! When I predicted in Y2K that DVD-A and SACD would fail in the marketplace, even I was not 100.00% convinced.  Of course now I am. But, its a decade later. So much for my prophet's license! ;-)


Yeah, I'm not sure what he had in mind when he wrote that. Of course, professional use requires the combination of multiple channels with cumulative noise increases, extra headroom for recording, and consideration of the possibility of future releases into a more technologically demanding environment. Also, a mixer might decide to increase the relative level of a track, in which case the noise of that track would become more audible. Compression can also increase noise level, whether applied in post production or broadcast. Those are all I think arguments for the use of a robust standard in the studio, one which guarantees that in any reasonable use noise is unlikely to be a problem, and I think in general the industry has followed that practice.

I'm not sure why SACD and DVD-A failed. They certainly weren't going to succeed on the basis of HBR recording, which by all accounts leads at best to a very minor sonic improvement. I was present once at a demonstration conducted by a major electronics manufacturer in which listeners were asked to vote their preference, a multichannel surround version, or a higher sampling rate version. With the exception of an engineer who was involved in the design of high rate converters, everyone voted for the multichannel.

IMO, if high resolution and multichannel audio have a future, and I think they do, it's in the form of downloads; discs are dead, even if they don't quite know it yet.

Quote
It really comes down to what percentage of consumers that you want to satisfy. I suspect that far less than 0.001% of all listeners actually maintain and routinely use a listening environment that is not well-served by 16 bits done well.  OTOH 0.001% of all listeners is still thousands of people. But thousands of people don't make a market for mainstream products. There's something very democratic about distribution media - everybody feeds at the same trough.

Reality is that much distributed media has squashed dynamic range that probably disappoints 1% or more of all listeners. If you can't get all media to be made to a standard that doesn't disappoint 1%, then the futility of trying to make a business out of pleasing 0.001% is very clear.


Interestingly, though, some of the big labels have started releasing high quality vinyl. It's a format, after all, that people can't steal!

My sense of it is that the new downloadable formats offer an opportunity for record companies to offer higher quality versions of their product to paying customers, or to allow a company that specializes in high fidelity recordings to do so. Whether this small market justifies the expense of a remix, I don't know, but I tend to suspect that it does, just as DVD and now Blu Ray sales have apparently been sufficient to finance new film-tape sessions. Certainly, there have been a lot of remixes of varying quality as record companies seek to squeeze new revenue from old material.

Have you ever regretted ABXing?

Reply #56
Interestingly, though, some of the big labels have started releasing high quality vinyl. It's a format, after all, that people can't steal!



They can and do. It works like this: one guy (let's just call him something random like PBTHAL) with a great set-up rips the vinyl to the "audiophile-preferred" 24-bit FLAC. From thence he uploads it to the usual file-sharing sites.

I hope that CDs won't be soon dead. I have a fear you are quite right in your prediction though.
The Loudness War is over. Now it's a hopeless occupation.

Have you ever regretted ABXing?

Reply #57
Of course, professional use requires the combination of multiple channels with cumulative noise increases, extra headroom for recording, and consideration of the possibility of future releases into a more technologically demanding environment. Also, a mixer might decide to increase the relative level of a track, in which case the noise of that track would become more audible. Compression can also increase noise level, whether applied in post production or broadcast. Those are all I think arguments for the use of a robust standard in the studio, one which guarantees that in any reasonable use noise is unlikely to be a problem, and I think in general the industry has followed that practice.


All those things really happen, but they don't detract from the use of perceptually-shaped quantization,  If the noise floor of a track is for example 20 dB below the threshold of imperceptibility, and you amplify it by 10 dB, that noise floor is still 10 dB below the threshold of perception, no matter whether the noise floor was shaped or unshaped.

I have repeatedly encountered this suspicion among older audio professionals that somehow perceptual gains "aren't real".  For example, people who would never ever demonstrate one of their speakers with even the best-made MP3.

Have you ever regretted ABXing?

Reply #58
Interestingly, though, some of the big labels have started releasing high quality vinyl. It's a format, after all, that people can't steal!


I don't think that anybody seriously sees vinyl as a strategic tool in the war on piracy.

We must remember that music media is a world of style and taste, and what and how some piece of music is released is all about what someone thinks someone else thinks is cool. If the guy down the street releases something on a certain niche medium, just make a few phone calls or send a few emails, the last of which places the press release that you just released something in the same format.

Finally, only about 30 years later, a signficant part of the potential niche market for vinyl seems to be realizing that digitizing does not destroy the unique sound of music on vinyl.

Have you ever regretted ABXing?

Reply #59
Interestingly, though, some of the big labels have started releasing high quality vinyl. It's a format, after all, that people can't steal!



They can and do. It works like this: one guy (let's just call him something random like PBTHAL) with a great set-up rips the vinyl to the "audiophile-preferred" 24-bit FLAC. From thence he uploads it to the usual file-sharing sites.

I hope that CDs won't be soon dead. I have a fear you are quite right in your prediction though.


I confess I've never understood that. I mean, what can the vinyl stage here possibly do except add noise and distortion?

Have you ever regretted ABXing?

Reply #60
Interestingly, though, some of the big labels have started releasing high quality vinyl. It's a format, after all, that people can't steal!



They can and do. It works like this: one guy (let's just call him something random like PBTHAL) with a great set-up rips the vinyl to the "audiophile-preferred" 24-bit FLAC. From thence he uploads it to the usual file-sharing sites.

I hope that CDs won't be soon dead. I have a fear you are quite right in your prediction though.


I confess I've never understood that. I mean, what can the vinyl stage here possibly do except add noise and distortion?

Well for what it's worth i have had some vinyl transcribed that i could not get on CD. "The Ship" is an album that has never been on any other format than vinyl and there are also a bunch of Jazz releases (some of the works of Miles Davis, John Coletrane, Thelonious Monk and others in that bunch) that have also never been released on other formats.

Also, i prefer the original mono to stereo tracks when i can get them and sometimes (as for Kind of Blue) you can only get the mono on vinyl. I really dislike the hard panning that was used early on to prove to you that you had 2 different speakers, as if the producer was saying "Look! There's your right one! In case you forgot There's your left one and it has one saxophone in it and one piece from the drum kit while the high-hats are in your right one all by themselves! Cool huh?"  Ugh. The drummer must have had really long arms and legs.

That makes it worth it to me.
Music lover and recovering high end audiophile

Have you ever regretted ABXing?

Reply #61
Interestingly, though, some of the big labels have started releasing high quality vinyl. It's a format, after all, that people can't steal!


I don't think that anybody seriously sees vinyl as a strategic tool in the war on piracy.

We must remember that music media is a world of style and taste, and what and how some piece of music is released is all about what someone thinks someone else thinks is cool. If the guy down the street releases something on a certain niche medium, just make a few phone calls or send a few emails, the last of which places the press release that you just released something in the same format.

Finally, only about 30 years later, a signficant part of the potential niche market for vinyl seems to be realizing that digitizing does not destroy the unique sound of music on vinyl.


The point, which is not mine but a label executive's, was that record companies can make money selling vinyl pressings because they can't be downloaded, not that vinyl disks will somehow eliminate privacy. Surprisingly, , the vinyl niche has been growing, both in number of units shipped and as a percentage of total sales, at a time when CD sales continue to shrink. So it's a profit opportunity for the record companies and they've availed themselves of it, moving back into vinyl even though at 2.1 million units in 2009, vinyl accounts for less than 1% of total sales.

Have you ever regretted ABXing?

Reply #62
Well for what it's worth i have had some vinyl transcribed that i could not get on CD. "The Ship" is an album that has never been on any other format than vinyl and there are also a bunch of Jazz releases (some of the works of Miles Davis, John Coletrane, Thelonious Monk and others in that bunch) that have also never been released on other formats.

Also, i prefer the original mono to stereo tracks when i can get them and sometimes (as for Kind of Blue) you can only get the mono on vinyl. I really dislike the hard panning that was used early on to prove to you that you had 2 different speakers, as if the producer was saying "Look! There's your right one! In case you forgot There's your left one and it has one saxophone in it and one piece from the drum kit while the high-hats are in your right one all by themselves! Cool huh?"  Ugh. The drummer must have had really long arms and legs.

That makes it worth it to me.


Sure, and some people also use vinyl as the source of transfers because it's just what they happen to have. All worthwhile.

In the case of the Beatles albums, I've found at least some of the mono mixes substantially better from a musical perspective, perhaps because they were the ones in which the Beatles themselves were involved -- they apparently weren't interested in the stereo mix. Of course, they're available on CD.

Have you ever regretted ABXing?

Reply #63
Never regretted learning that my hearing isn't that good and Lame @ -V5 is good enough for my instrumental music.  More music on smallish flash drives and players is a-ok by me.
foobar2000, FLAC, and qAAC -V90
It just works people!

Have you ever regretted ABXing?

Reply #64


Quote
I see the Fielder paper as being a justification for HDCD, which we now know failed in the marketplace.


It wasn't written to sell HDCD, Fielder worked for Dolby Labs.

Have you ever regretted ABXing?

Reply #65
There seems to be a critical missing paragraph in the Fielder paper. It might be titled "Noise in recording Environments".

The paper as presented usecontains an unstated assumption that appears to me to be something like: "The dynamic range requirement for musical playback is irrelevant to noise in the space where the music is recorded."


Look again, it is there: Section 4.1, Noise in the Recording Environment.


No, data about totally empty rooms is *not* relevant, at least until we start generally making recordings in totally empty rooms.  I think there's only one piece of music that could be recorded this way. It was written by John Cage, if memory serves.  ;-)


Fielder isn't assuming a totally empty room.  He *is* assuming the best (real, not imaginary) recording venue, circa 1994.

Have you ever regretted ABXing?

Reply #66
Though as the authors of the comparison paper pointed out, high res CD's are frequently produced to a higher standard of quality. This is because the record companies know that they're selling mainly to audiophiles and don't have to accommodate Grandma's Philco with sonic compromises like excessive compression.


"It's available in 24/96! Buy the 16/44.1 for sound quality!"



Oh, then you'd positively LOVE these. >:D


Obscenely cool, I want those in my living room. (Did I mention I live alone?)

Have you ever regretted ABXing?

Reply #67
The weird thing for me was that I really could tell the difference between 24 bit and 16 bit when I did ABX testing....48khz and 96khz however? Not so much.  Got about 40%.

Mp3s vs lossless though is night and day.

Have you ever regretted ABXing?

Reply #68
Got about 40%

40% is the result  one would expect from randomly choosing (read: guessing).

difference between 24 bit and 16 bit
[...]
Mp3s vs lossless though is night and day.

ABX log and samples for these, please.

Have you ever regretted ABXing?

Reply #69
Got about 40%

40% is the result  one would expect from randomly choosing (read: guessing).

difference between 24 bit and 16 bit
[...]
Mp3s vs lossless though is night and day.

ABX log and samples for these, please.



Depends on the number of trials. 40% after a certain number of trials is not what one would expect from random choosing and is evidence of something going on actually. It's weird but a statistically significant margin of wrong answers would point to something other than chance at work.

Have you ever regretted ABXing?

Reply #70
40% after a certain number of trials is not what one would expect from random choosing and is evidence of something going on actually. It's weird but a statistically significant margin of wrong answers would point to something other than chance at work.


40 precent after a certain amount of trials suggests that there is something very wrong with the experiment. The most common source of this kind of problem that we've seen is where a bunch of listeners are communicating with each other during the test, so the effective number of independent trials is far less than initially estimated.

Have you ever regretted ABXing?

Reply #71
Got about 40%
40% is the result  one would expect from randomly choosing (read: guessing).

difference between 24 bit and 16 bit
[...]
Mp3s vs lossless though is night and day.
ABX log and samples for these, please.


Got about 40%
40% is the result  one would expect from randomly choosing (read: guessing).

difference between 24 bit and 16 bit
[...]
Mp3s vs lossless though is night and day.
ABX log and samples for these, please.


Depends on the number of trials. 40% after a certain number of trials is not what one would expect from random choosing and is evidence of something going on actually. It's weird but a statistically significant margin of wrong answers would point to something other than chance at work.


Code: [Select]
foo_abx 1.3.4 report
foobar2000 v1.0.1
2010/11/11 02:15:23

File A: G:\Music\Audioslave\Audioslave\02 Show Me How to Live.m4a
File B: G:\Music\Audioslave\Audioslave\02 Show Me How to Live.wav

02:15:23 : Test started.
02:15:39 : 01/01  50.0%
02:16:05 : 02/02  25.0%
02:16:41 : 03/03  12.5%
02:17:36 : 04/04  6.3%
02:17:55 : 05/05  3.1%
02:18:21 : 06/06  1.6%
02:18:40 : 07/07  0.8%
02:19:12 : 08/08  0.4%
02:19:39 : 09/09  0.2%
02:20:23 : 10/10  0.1%
02:20:39 : 10/11  0.6%
02:20:41 : Test finished.

 ----------
Total: 10/11 (0.6%)

This was 16-bit vs 24 bit.  Same 24 bit source file.

I had a test from long ago with mp3s vs lossless, but I don't think I saved the log.  (results were similar)  So I did another, the old one I did with 128kbs mp3 conversions, so I decided to do this new one with 320s, and I found that I really couldn't tell the difference.  LAME ain't bad at all.  I retract my statement of "night and day".  But I'll have to retry the 128s again.

Code: [Select]
foo_abx 1.3.4 report
foobar2000 v1.0.1
2010/11/26 01:25:20

File A: G:\Music\Other\Joshua Redman\Freedom In The Groove\01 Hide And Seek.m4a
File B: G:\Music\Other\Joshua Redman\Freedom In The Groove\01 Hide And Seek.mp3

01:25:20 : Test started.
01:25:32 : 00/01  100.0%
01:25:38 : 00/02  100.0%
01:25:49 : 00/03  100.0%
01:25:55 : 01/04  93.8%
01:26:41 : 01/05  96.9%
01:27:31 : 02/06  89.1%
01:27:58 : 02/07  93.8%
01:28:46 : 03/08  85.5%
01:29:15 : 04/09  74.6%
01:30:08 : 04/10  82.8%
01:30:34 : 05/11  72.6%
01:30:48 : 06/12  61.3%
01:31:11 : Test finished.

 ----------
Total: 6/12 (61.3%)


But i did save the 48 vs 96 one...I think you misunderstood me, I said I really couldn't hear the difference here.

Code: [Select]
foo_abx 1.3.4 report
foobar2000 v1.0.1
2010/11/11 02:24:26

File A: G:\Music\24\Reiner - Strauss, Also Sprach Zarathustra, 1962 (24bit-96KHz vinyl)\1. R.Strauss; Also Sprach Zarathustra, Op.30 - Part 1.flac
File B: G:\Music\24\Reiner - Strauss, Also Sprach Zarathustra, 1962 (24bit-96KHz vinyl)\1. R.Strauss; Also Sprach Zarathustra, Op.30 - Part 1.wav

02:24:26 : Test started.
02:25:37 : 01/01  50.0%
02:27:39 : 01/02  75.0%
02:28:26 : 02/03  50.0%
02:29:10 : 03/04  31.3%
02:20:11 : 03/05  50.0%
02:21:06 : 04/06  34.4%
02:22:51 : 04/07  50.0%
02:24:02 : 04/08  63.7%
02:25:19 : 05/09  50.0%
02:26:52 : 05/10  62.3%
02:27:41 : 05/11  72.6%
02:28:49 : 05/12  80.6%
02:28:58 : 05/13  86.7%
02:29:48 : Test finished.

 ----------
Total: 5/13 (86.7%)

To sum it up, I was saying that 24 bit vs 16 is noticeable for me, but I have been humbled as far as mp3s go. :/
I seem to be in the minority when it comes to 24bit audio, and I don't know why.  I can definitely hear the difference, particularly in horn parts, drums.  Guitar depending on the distortion, but it really isn't the biggest deal.  I listen to a lot of classical music, and it helps with making the different parts sound separated and clear, this I am sure of.  What do you guys think?

P.S- I didn't have anyone with me for any ABX test. Sorry for the late reply! Turkey Day and all...)

Have you ever regretted ABXing?

Reply #72
Your ABX report saying you can tell the difference between 16-bit and 24-bit is m4a vs. wav.

Please provide 30 second samples or the part that you can most easily distinguish.

Have you ever regretted ABXing?

Reply #73
Your ABX report saying you can tell the difference between 16-bit and 24-bit is m4a vs. wav.

Please provide 30 second samples or the part that you can most easily distinguish.


Yes, please. And tell us how the 16bit file version was created or obtained.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Have you ever regretted ABXing?

Reply #74
Your ABX report saying you can tell the difference between 16-bit and 24-bit is m4a vs. wav.

Please provide 30 second samples or the part that you can most easily distinguish.


Yes, please. And tell us how the 16bit file version was created or obtained.

(continuing the off-topic-ness a bit i guess it all can be split off to another thread if need be)

I'm actually more interested in how the 24-bit file was obtained. I looked around for a dvd-audio/sacd of this track and all i found was a mention on wikipedia:

http://en.wikipedia.org/wiki/Audioslave_(a...on_of_the_album

which states "This test market version of the album is rare. ... The DVD side of the Audioslave DualDisc featured the entire album in higher resolution 20bit 48 kHz sound, as well as some videos."

So where did the 24-bit file come from? The only 24-bit copy i've seen of this is a needle drop online.  If that's the source then we're actually comparing vinyl to CD aren't we?  That's a whole 'nother game isn't it?

Music lover and recovering high end audiophile