DBT Is Flawed Because Bob Stuart Says So, Split from Topic ID #11442 Options

[Woodinville]

How do you explain

Robert Harley -- Mon, 07/14/2008 - 10:28
My point is that the better-sounding piece of equipment more easily allows you to forget the equipment and focus on the music. Electronic artifacts are a constant reminder that we're listening to a re-creation, rather than to music itself.

Greater musical enjoyment comes not from listening in analytical mode for specific sonic attributes, but through lower awareness of the electro-mechanical system between listener and performer. The listener might not be aware of why he feels a closer connection to the music with better equipment. One doesn't have to be consciously aware of a reduction in artifacts to appreciate the deeper musical involvement the reduction in those artifacts engenders.

Maybe this is how.

That's so far off I don't know where to start. You can listen any way you want in a blind test, you just don't know what you're listening to.

So sorry, that's baloney.

[knutinh]

"lots of other 'clever' people who also happen to reject orthodox scientific testing say the same thing, so I must be right!"

Add Barry Diament of Soundkeeper Recordings to that long list of 'clever' people. He publicly stated that, in his own opinion, the audible difference between properly recorded 2x (96 kHz or 88.2 kHz) and properly recorded 4x (192 kHz or 176.4 kHz) samplerate is even bigger than the audible difference between 1x (48 kHz or 44.1 kHz) and 2x. He also wrote on ComputerAudiophile forum that, if he could invite you to his recording studio, it would take him two seconds to prove the audible difference between 16/44.1 and 24/192 to you.

Why is it that you have no scientific sources? Would you not think that the people whose career is built upon investigating physics and perception and documenting their findings in such a way that it passes peer-review would be better suited to support your claims than people who are good at twisting equalizer knobs to make pleasing audio recordings?

What if we were discussing the health consequences of smoking, would you use Philip Morris as a source that it is actually healthy to smoke? (they got to know, they have been chruning out gazillions of sigarettes, right?).

-k

This post has been edited by knutinh: Aug 14 2012, 06:47

[probedb]

Add Barry Diament of Soundkeeper Recordings to that long list of 'clever' people. He publicly stated that, in his own opinion, the audible difference between properly recorded 2x (96 kHz or 88.2 kHz) and properly recorded 4x (192 kHz or 176.4 kHz) samplerate is even bigger than the audible difference between 1x (48 kHz or 44.1 kHz) and 2x. He also wrote on ComputerAudiophile forum that, if he could invite you to his recording studio, it would take him two seconds to prove the audible difference between 16/44.1 and 24/192 to you.

You did the the bit in bold didn't you? That doesn't make it true, it's one persons opinion.

I'm really no expert at all but all I see from you is blah, blah, blah, Bob Stuart, blah, blah, Bob Stuart.

I'm not entirely sure why people continue to argue, it's like trying to stop a priest believing in a god, not gonna happen unless you have Dougal from Father Ted on your side

[greynol]

Barry Diament? Really?!?

No...




...really?!?!?

This is the guy who first made an impression on this forum for stating how a CD-R copy sounds better than the original source disc?

Like Stuart, Diament has a vested interest in having people believe that up is down.

What a wonderful example to use in committing the logical fallacy of argument from authority.

[Arnold B. Kruege...]

He's obviously 100% oblivious to the following well-known rather highly relevant article:

Audibility of a CD-Standard A/DA/A Loop Inserted into High-Resolution Audio Playback
Authors: Meyer, E. Brad; Moran, David R.
Affiliation: Boston Audio Society, Lincoln, MA, USA
JAES Volume 55 Issue 9 pp. 775-779; September 2007

Hey, it had only been out for a year and a half and plastered all over the web the whole time... ;-)

It was later proven that the Meyer & Moran tests were not using truly HiRes material, but PCM converted to DSD.

I believe that the conversion in question was well-documented in the article.

However, the idea that converting DSD to PCM necessarily causes a significant loss of resolution is a false claim. In fact, 24/192 has far more resolution than SACD.

Your apparent ignorance of Shannon's relevant theorems is noted.

Bob Stuart was one of the test subjects BTW.

And???

The audible difference between 16/44.1 and proper HiRes is actually quite enormous.

True as long as you do sighted evaluations.

Your major problem at HA is that many of us have done bias-controlled evaluations of various kinds (not just ABX) and know that at the very best, the differences are very subtle and in general they are vanishing.

Let's be clear - it is now known that the Meyers and Moran tests might be flawed because M&M took at face value, vendor claims that SACDs and DVD-As are high resolution recordings.

We now know that detailed technical analysis puts the actual percentage of all such recordings (not just the ones M&M used) closer to 50% high resolution. IOW the vendors didn't tell the truth, and it came around and bit them when M&M believed them.

You see there was a sort of shadow DBT that had already happened - all of these CD (or worse) resolution recordings were sold on DACDs and DVD-As as being high rez, but they weren't. For quite a while no golden eared reviewers or audiobphiles blew the whistle. In the end, most if not all people who noticed the loss of resolution did so by means of technical tests, not listening tests!

But many of us have been doing similar tests with recordings known to be high resolution, for over a decade. No joy!

Again, you are revealing what you don't actually have any personal experience with which is lots!

[Arnold B. Kruege...]

I'll try to be more clear. My point was if step (2) reveals a certain detail after step (1) did not reveal it, it's perfectly possible that we can switch back to system A and always hear this detail even though the same detail could not be heard during step (1).

Of course! That's why ABX has allowed people to repeat step(1) and step(2) in a completely transparent and open way, whenever they want to including right in the midst of a trial.

Obviously your belaboring this particular point shows that you have zero hands-on experience with ABX and have been listening to people who either lack any experience with ABX at all, or they completely missed many important points about ABX when they did try it.

What you need to understand is that many well-known golden ears seem to get hysterical at even the mention of ABX. I can name names based on personal experiences. They literally have to be carried away in some cases! So, they may evaluate ABX while in a hysterical state, and what they report after is not complete or reliable.

[Arnold B. Kruege...]

Robert Harley -- Tue, 07/15/2008 - 10:22

First, my assertion that audible differences exist between 44.1kHz/16-bit, 96kHz/24-bit, and DSD is based on my own listening experiences.

Since Harley has been pretty consistently negative about blind tests for at least 2 decades, its safe to assume that none of the below is based on blind tests.

This includes hearing a live microphone feed and comparing it to standard resolution PCM, high-resolution PCM, and DSD.

Been there, done that and with numerous other people.

Have you performed the same test and reached a different conclusion?

Obviously not the same test, but many like them.

Or are you content to base your belief that they all sound identical on the results of a published test?

Not at all.

(A “test” devised and conducted, by the way, by two individuals with a long history of attempting to discredit audiophiles.)

I believe that Harley is referring to Vanderkooy and Lipshitz. I know those guys personally and he appears to be libelling them. First off, they are actually audiophiles themselves so discrediting audiophiles would be like intellectual suicide. Secondly, their targets are mostly some of the more visible bozos and charlatans in high end audio, not high end audio in total.

Are you suggesting that I should reject my own direct experience and conclude that I was simply deluded?

This is yet another excluded-middle argument because the kind of error that Harley is talking about is more like being fooled by an illusion than some kind of mental defect. Let's be clear here - everybody is affected by sighted and other bias. The mental defect is denying it.

Is everyone who hears a difference between standard- and high-resolution digital audio similarly deluded—people like Meridian’s Bob Stuart, Keith Johnson, dCS founder Mikey Storey, Peter McGrath, and other credentialed and respected leaders in the field who have decades of academic research and hands-on experience with the subject?

None of the hands-on experimenters mentioned are exactly what you'd call advocates of reliable listening tests. ;-) Most if not all of them have their own dogs in the fight.

This post has been edited by Arnold B. Krueger: Aug 14 2012, 14:19

[Arnold B. Kruege...]

How do you explain

Robert Harley -- Mon, 07/14/2008 - 10:28
My point is that the better-sounding piece of equipment more easily allows you to forget the equipment and focus on the music. Electronic artifacts are a constant reminder that we're listening to a re-creation, rather than to music itself.

Greater musical enjoyment comes not from listening in analytical mode for specific sonic attributes, but through lower awareness of the electro-mechanical system between listener and performer. The listener might not be aware of why he feels a closer connection to the music with better equipment. One doesn't have to be consciously aware of a reduction in artifacts to appreciate the deeper musical involvement the reduction in those artifacts engenders.

Maybe this is how.

The obvious fallacy is the idea that blind test advocates use only cheap inferior equipment. Visible shades of class warfare.

It's just another piece of evidence that much of high end audio is strongly ego-driven, not nearly as sound quality driven as many would like to pretend.

This post has been edited by Arnold B. Krueger: Aug 14 2012, 14:18

[greynol]

Let's be clear - it is now known that the Meyers and Moran tests might be flawed because M&M took at face value, vendor claims that SACDs and DVD-As are high resolution recordings.

We now know that detailed technical analysis puts the actual percentage of all such recordings (not just the ones M&M used) closer to 50% high resolution. IOW the vendors didn't tell the truth, and it came around and bit them when M&M believed them.

You see there was a sort of shadow DBT that had already happened - all of these CD (or worse) resolution recordings were sold on DACDs and DVD-As as being high rez, but they weren't. For quite a while no golden eared reviewers or audiobphiles blew the whistle. In the end, most if not all people who noticed the loss of resolution did so by means of technical tests, not listening tests!

Now this would have made for good material in the parent discussion.

[2Bdecided]

If a sound is masked (undetectable), it doesn't work this way. The memory doesn't bring it back once its masked again.

Once again, our opinions differ. I suggest you watch that Audio Myths Workshop vid I linked to earlier, more in particular the part where Poppy Crum shows how it does work.

Do you hear an 's'? I think I hear an 's'.

"lots of other 'clever' people who also happen to reject orthodox scientific testing say the same thing, so I must be right!"

Add Barry Diament of Soundkeeper Recordings to that long list of 'clever' people. He publicly stated that, in his own opinion, the audible difference between properly recorded 2x (96 kHz or 88.2 kHz) and properly recorded 4x (192 kHz or 176.4 kHz) samplerate is even bigger than the audible difference between 1x (48 kHz or 44.1 kHz) and 2x. He also wrote on ComputerAudiophile forum that, if he could invite you to his recording studio, it would take him two seconds to prove the audible difference between 16/44.1 and 24/192 to you.

How does a "quite enormous" difference disappear completely when someone says the words "which one are you listening to then?"

Judging by my own listening experience, the difference doesn't disappear even one bit.

So given this apparently enormous indestructible difference, how come we've had HiRes for 13 years, and no successful ABX tests against CD?

There are published industry and broadcaster DBTs of HD video, used to determine at which viewing distance the improvement becomes visible. How come the audio industry couldn't manage it for HiRes audio?

If a sound is masked (undetectable), it doesn't work this way. The memory doesn't bring it back once its masked again.

Once again, our opinions differ. I suggest you watch that Audio Myths Workshop vid I linked to earlier, more in particular the part where Poppy Crum shows how it does work.

Do you hear an 's'? I think I hear an 's'.

given that I understand multimodal perception and auditory scene analysis, I understand how these tricks work.

To suggest this disproves masking thresholds is damn close to suggesting that optical illusions show how picture coding will never work. Amazingly, optical illusions cope with JPEG encoding just fine

Cheers,
David.

[Stone Free]

I'll try to be more clear. My point was if step (2) reveals a certain detail after step (1) did not reveal it, it's perfectly possible that we can switch back to system A and always hear this detail even though the same detail could not be heard during step (1). The explanation why this is perfectly possible is because step (2) did not only reveal this detail, but at the same time also it caused our brain to stored important information about this detail, into the memory of our brain. The fact this information is still accessible to our brain after we finished step (2) is because we do not have the power to erase it from our memory. Hence, we do not have the power to eliminate the bias that can result from this information, regardless of whether there was any practicing involved.

I can think of a similar situation where details that could not be detected before a second step, that became obvious after it.

I would genuinely like to understand how come this is possible, the situation is as follows:

I'm watching a TV programme and a character says something, and I just can't understand what they've said, so I rewind my PVR and re-listen to them speak, nope, can't hear it, so I turn up the volume real loud play it several times, and no, I can't understand what they are saying.

However I rewind it once more and turn on the subtitles, so now I know what they actually said, and somehow from that point on if I rewind and replay it without subtitles, I can no longer re-create the situation of being unable to hear what was said.

What is the scientific reason behind this?

[db1989]

What is the scientific reason behind this?

Er, I don’t think there’s anything really scientific about it. You now know something that you did not know before. How could you ever recreate a time when you did not know it?

[lvqcl]

[Ron Jones]

Robert Harley -- Mon, 07/14/2008 - 10:28
One doesn't have to be consciously aware of a reduction in artifacts to appreciate the deeper musical involvement the reduction in those artifacts engenders.

Appreciation denotes a level of consciousness. Unless he's referring to some sort of subconscious level of appreciation (which wouldn't be worthwhile to the listener in any way), this statement doesn't make sense.

This statement also suggests that a reduction in artifacts will lead to "deeper musical involvement", which isn't necessarily the case.

[greynol]

With the previous examples about subtitles and the dalmation in mind, someone is going to come away thinking that listening to hi-res material will reveal details that didn't exist on the standard-res format and these details will forever be ghosted when listening to the standard-res format. Because of this phenomenon ABX can never work.

Of course ABX has been shown to work just not with the enormous differences of hi-res over standard-res. Every passed ABX test must be the exception!

The key difference is that the actor really did say what the subtitles said and the dalmation really was in that picture. IOW there never was any difference to begin with. The equivalent is more like succesfully training yourself to hear a particular coding artifact. Once you know what to listen for you can spot it. Unortunately for those now trained, the original and the coded now sound different, rather than the same as certain people would like for you to believe.

Now how come I don't hear people say that there's an enormous difference between hi-res and standard-res but once you listen to the hi-res version you no longer hear this difference and the standard-res version sounds every bit as good?

This post has been edited by greynol: Aug 14 2012, 19:34

[Arnold B. Kruege...]

I'll try to be more clear. My point was if step (2) reveals a certain detail after step (1) did not reveal it, it's perfectly possible that we can switch back to system A and always hear this detail even though the same detail could not be heard during step (1). The explanation why this is perfectly possible is because step (2) did not only reveal this detail, but at the same time also it caused our brain to stored important information about this detail, into the memory of our brain. The fact this information is still accessible to our brain after we finished step (2) is because we do not have the power to erase it from our memory. Hence, we do not have the power to eliminate the bias that can result from this information, regardless of whether there was any practicing involved.

I can think of a similar situation where details that could not be detected before a second step, that became obvious after it.

I would genuinely like to understand how come this is possible, the situation is as follows:

I'm watching a TV programme and a character says something, and I just can't understand what they've said, so I rewind my PVR and re-listen to them speak, nope, can't hear it, so I turn up the volume real loud play it several times, and no, I can't understand what they are saying.

However I rewind it once more and turn on the subtitles, so now I know what they actually said, and somehow from that point on if I rewind and replay it without subtitles, I can no longer re-create the situation of being unable to hear what was said.

What is the scientific reason behind this?

There are at least two different effects.

(1) Once you learn that you are supposed to hear something, you will hear it even if it isn't there. I believe this is the essence of the McGurk Effect. McGurk Effect

(2) Long term learning to decode audio does take place. For example since I've been watching BCC America quite a bit, I can more easily and accurately decode English with a British accent.

[Arnold B. Kruege...]

Now how come I don't hear people say that there's an enormous difference between hi-res and standard-res but once you listen to the hi-res version you no longer hear this difference and the standard-res version sounds every bit as good?

I don't know how this relates but in general the enormous difference goes away when you start doing blind tests, any number of apparently successful sighted evaluations preceding or intermingled notwithstanding.

[mzil]

I'm watching a TV programme and a character says something, and I just can't understand what they've said,...

However I rewind it once more and turn on the subtitles, so now I know what they actually said, and somehow from that point on if I rewind and replay it without subtitles, I can no longer re-create the situation of being unable to hear what was said.

What is the scientific reason behind this?

It's pareodolia. Here's a hilarious version of it that works the way our TV and closed captioning works, ie humans tendency towards suggestibility:
Oh Four Tuna.

It works both if the suggested words are true OR fictiional.

This post has been edited by mzil: Aug 14 2012, 21:18

[greynol]

I don't know how this relates but in general the enormous difference goes away when you start doing blind tests, any number of apparently successful sighted evaluations preceding or intermingled notwithstanding.

Apparently this difference may never come back thanks to being exposed to the hi-res version in which case double-blind or sighted tests will no longer reveal any differences. The upside to this is you only need to audition (insert gear or media of your choice here) and can go back to whatever (insert gear or media of your choice here) for it to provide the same stunningly wonderful experience! Of course if this phenomenon can be reset then the excuse that double-blind tests can't ever indicate a successful outcome falls flat on its face.

[2Bdecided]

OT, but here's the best example I've seen of only being able to hear something when you're told what it is (posted recently to another thread)...
http://www.youtube.com/watch?v=muCPjK4nGY4

greynol makes a far more useful and serious point. These criticisms, if true, prevent listening comparisons full stop (period). And yet people compare, and report enormous differences. It's insane.

Arny too repeats the truth that those who have actually tried double blind testing have a far better insight into the limits of human perception than those who haven't.

Cheers,
David.

P.S. I love some of Barry Diament's recordings. They sound pretty great as -V2 mp3s. I don't think the new albums are as good as the first one (and that's aside from the music!) - he managed a perfect acoustic mix on the first album (Lift) which is the way I'd like to hear most recordings. He hasn't managed to match that perfection on the others (IMO). Different musicians and different halls, I know, but it suggests to me that the perfect microphone placement on Lift may have been partly luck, or at least something that's not so easily recreated. And mp3 vs CD vs DSD is irrelevant in this case compared with placing the microphone 1 foot too far back.

[greynol]

greynol makes a far more useful and serious point.

Thank goodness for logical conclusions and preceding posts by other members. I wonder if that's why earlier questions on the subject were being evaded.

And mp3 vs CD vs DSD is irrelevant in this case compared with placing the microphone 1 foot too far back.

We forgot about the superior CD-R copy.

Seriously though, I've been happy with the remastering work he did in the '80s and perhaps other work, but I don't typically get too caught up with that type of stuff unless the engineer was grossly incompetent beyond annoying stylistic trends. What's unfortunate is that while the man obviously has gifted ears, it would seem he has limited his potential by not embracing unbiased critical listening exercises (to Arny's point).

[Hotsoup]

Arny too repeats the truth that those who have actually tried double blind testing have a far better insight into the limits of human perception than those who haven't.

Not to continue veering off topic, but all I needed was about 45 minutes of calibrating my subwoofer until realizing it wasn't even turned on. Any golden ear I thought I had was gone after that. ABX's have done the same.

[greynol]

I have a feeling audioclaudio decided to stop digging his hole, so I think letting the discussion progress elsewhere is fine except that anything worthwhile to future readers may go overlooked when the topic title comes up in a search.

Having this thread meander into the weeds does not concern me anywhere near as much as what these posts did to their parent topic that is linked in our TOS.

[Stone Free]

What is the scientific reason behind this?

Er, I don’t think there’s anything really scientific about it. You now know something that you did not know before. How could you ever recreate a time when you did not know it?

Its not that you now know what was said, but that you now hear what was said, even though you couldn't before.

The McGurk effect seems to be about vision affecting the perception of sound, whereas this is knowledge affecting perception. I don't think it is Pareidolia either as that seems to be about the brains attempt to order/classify things and producing incorrect links.

Without experiencing it I would have thought that it should still sound indistinct as it did before. I wonder if someone told me the following video would contain the word before I heard it, whether I would hear it correctly with that foreknowledge.

[2Bdecided]

I've been thinking more about the Bob Stuart statement in the original context: when listening to more "revealing" audio gear, you often hear things that you didn't before - but then you'll notice them on less "revealing" audio gear. This is suggested as a reason that ABX cannot work - because you can't "unhear" things.

Laying aside the question of whether it's the gear or your state of mind that causes the revelation ("this gear is really good - I'll listen very carefully to see what it reveals" ), I have to say that, every time it's happened to me (usually with different headphones), I can still very clearly hear the differences between the headphones, even though the newly revealed feature is now audible through both.

Contrast this with the situation where something really is removed (IME a barely audible jangling of keys in a live recording, removed by a 16kHz low pass filter), and the keys were audibly missing from the LPF rendition, in a blind test, even though I knew they should be there, and they were the only audible clue to the LPF (there being little HF content apart from those keys). No amount of imagination could bring back something that was missing, and there was no other clue that something was missing.


Again, those who do ABX tests often find the opposite to what Bob's anecdote implies - ABX testing, while removing completely imagined differences, often lets you demonstrate that you could hear differences that were so subtle that you doubted them yourself. I have done several tests where I suspected I was guessing, but passed the test (without cherry picking).

Myself, I still wonder if HiRes vs CD will be successfully ABXed one day. I mean, I'm sure we can create a rigged test (bad converters, extreme content) where they can, but I wonder if something fairer will ever be passed. I don't discount the possibility, though given how hard it appears to be, it can't be that big a difference. It can't be "Life vs Death", as described by Stereophile.

Cheers,
David.