[gregorio]

Quote:

Originally Posted by mike1127 /img/forum/go_quote.gif I'm saying that some differences may only be detectable when listening to music as music. It is purely an assumption that all differences are detectable as sound.

It seems to me self evident that our perception of any art form depends on our emotional state, and human emotional states are hardly rational or reliable. However, the very basis of the musical art form is sound, whatever is in the music is in the sound, anything else is a pure invention of the brain percieving the sound. So it makes absolute logical sense to judge the quality of audio components purely by their sound. How the music affects you is for you as an individual, how well an audio system recreates sound is for everyone. Hence why scientific measurements and DBTs are important.

G

 

[mike1127]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif Always ?, it is reported that Mozart could reproduce exactly a piece of music on one hearing, but even if true, so what ? how does that help devise a listening test that is more accurate than current listening tests and how would you prove it to be more so ?. You surely are not going to suggest sighted listening tests ?

No I'm not going to suggest sighted listening tests. But it seems to me you are saying I'm not allowed to point out the difficulties in testing (testing for all kinds of differences) without also providing a working solution. Am I also not allowed to point out the difficulty with cold fusion without providing a working cold fusion reactor?

I have not had a lot of time to devote to testing, and my local friends are not interested in helping me go through any kind of elaborate or lengthy protocol. I'm guessing that other people, especially those who do it for a living, actually have very good ideas, but I'm sure most of them have better things to do with their time than post on this forum.

Quote:

So , does this make them right, some audiophiles say Silver sounds different from Copper, do we have ANY verifiable evidence for this assertion ?. Even highly respected and supposedly knowledgeable people say daft things.

I'm talking about how people describe their experience of listening to music when they introspect on it. Some people spend many hours introspecting on how they perceive things---and many things in addition to music. I have trained with several teachers in a particular kind of body awareness method that is built on an understanding of perception. I have a total of 500 hours of training, plus thousands of hours of practice. Many musicians have explored their perception to this kind of depth. So I can say with confidence that a large segment of the population has a similar perceptual mechanism to mine. In that mechanism, listening to music as music reveals different details than listening as sound.

I would be curious what you have learned from introspecting on your own listening process.



Quote:

Okay, it is a not at all helpful statement.

Sure it is. Anything which helps us understand how to conduct listening tests is helpful, even if that thing delivers some bad news about the current state-of-the-art in testing.


Quote:

Quote: Any controlled test rigs the game I cannot discuss this properly unless you give me some concrete examples, from the field of audio preferably but without such examples I can answer this in 3 ways 1) Of course, this is blindingly obvious , so what ? 2) No, it is of course possible to rig any test i.e to set it up in such a way as to deliberately bias the outcomes but that is not the purpose of a controlled test. 3) Would you rather go back to the uncontrolled sighted listening tests ?

The test protocol influences how subjects use their attention. For example a quick switching test encourages subjects to listen to sound as sound, and to focus into minute details. Now, if we imagine a long-term listening test, the situation gets very complex. If we want to control it, we would have to ask the subjects to use their attention in a particular way. They would lose freedom. It's really very simple.



Quote:

Speculating about the internal workings of a subject is unhelpful and unnecessary since you can only infer some of them and not with any real degree of accuracy.

It seems like you are saying "Because it's hard to understand something we should ignore it," which is clearly a logical fallacy.


Quote:

I seldom have anything nice to say about Skinner but sometimes a black box approach is entirely appropriate, I do not discount that something internal is going on but in the context of blind listening tests it is just not interesting. Your subjects can either detect a difference or they cannot, let them use whatever strategy they prefer.

It's very interesting to people who introspect on their own process. It's also not true "they can detect a difference or they cannot". There are many degrees between, since the real question is under what conditions they can detect a difference.

 

[mike1127]

Quote:

Originally Posted by gregorio /img/forum/go_quote.gif It seems to me self evident that our perception of any art form depends on our emotional state, and human emotional states are hardly rational or reliable. However, the very basis of the musical art form is sound, whatever is in the music is in the sound, anything else is a pure invention of the brain percieving the sound. So it makes absolute logical sense to judge the quality of audio components purely by their sound. How the music affects you is for you as an individual, how well an audio system recreates sound is for everyone. Hence why scientific measurements and DBTs are important. G

We're not really talking about sound itself, but talking about how the brain translates sound into consciousness. As you are no doubt aware, there are many layers of processing between the main auditory nerve and the cerebral cortex. I am not a biologist, but some of my teachers are, and the way it is explained to me is that the brain is highly connected vertically and horizontally. That is, there are a lot of layers which connect from higher to lower and vice versa, and there are lots of adjacent systems which are connected, like the auditory cortex and motor cortex.

The idea that we would become conscious of every detail of the sound---that every impulse in the auditory nerve would reach consciousness---simply by turning our attention "to the sound" (and away from the music) is really absurd.

As for the unreliability of emotional states, as I said in the previous post, just because something is hard to control or understand doesn't mean you can safely ignore it.

Mike

 

[nick_charles]

Quote:

Originally Posted by mike1127 /img/forum/go_quote.gif No I'm not going to suggest sighted listening tests. But it seems to me you are saying I'm not allowed to point out the difficulties in testing (testing for all kinds of differences) without also providing a working solution. Am I also not allowed to point out the difficulty with cold fusion without providing a working cold fusion reactor?

Cold fusion does not work , blind tests do.


Quote:

I have a total of 500 hours of training, plus thousands of hours of practice. Many musicians have explored their perception to this kind of depth. So I can say with confidence that a large segment of the population has a similar perceptual mechanism to mine. In that mechanism, listening to music as music reveals different details than listening as sound.

Since perception is partly internal, no, you cannot, all you can report is what others have reported, how do you define large segment ?

Quote:

I would be curious what you have learned from introspecting on your own listening process.

I do not introspect on my listening process, I listen. When I listen for pleasure I listen, when I listen for differences I listen, perhaps differently.


Quote:

The test protocol influences how subjects use their attention. For example a quick switching test encourages subjects to listen to sound as sound, and to focus into minute details. Now, if we imagine a long-term listening test, the situation gets very complex. If we want to control it, we would have to ask the subjects to use their attention in a particular way. They would lose freedom. It's really very simple.

You are still working from the assertion that you have to control listeners' listening strategies.


Quote:

It seems like you are saying "Because it's hard to understand something we should ignore it," which is clearly a logical fallacy.

Not at all, I work with hard to understand things all the time, I have spent the last 12 months analyzing difficult qualitative and quantitive data in great depth, I review academic papers (and publish a few myself) that dig into difficult things in often painful detail, but that is where it was necessary to do so, but the point is that in this case , audio listening, it is simply not necessary to know these things.


Quote:

It's very interesting to people who introspect on their own process. It's also not true "they can detect a difference or they cannot". There are many degrees between, since the real question is under what conditions they can detect a difference.

These conditions are generally stated, we have covered this already. Does this introspection in any way help anybody to detect differences, if so what evidence do you have for this ?

 

[gregorio]

Quote:

Originally Posted by mike1127 /img/forum/go_quote.gif We're not really talking about sound itself, but talking about how the brain translates sound into consciousness.

So if I understand your argument, you are saying that changing to an audiophile cable (for example) could change an audiophile's emotional state or expectation and this would then translate into the audiophile hearing a difference between the cables in the music. What is the difference between this assertion and what many here have called placebo effect? Could I get the same response from the audiophile if I told them the cable was hugely expensive when in fact it wasn't? If this is true, then the cable and any other audio component is irrelevant, only the expectation or emotional state of the audiophile is important. Surely therefore it would be far more economical for an audiophile to buy a very cheap sound system and then to have a course of hypnotherapy so that they have a very high expectation of the cheap system. Wouldn't the end result be identical, only many thousands of dollars cheaper?

Quote:

Originally Posted by mike1127 /img/forum/go_quote.gif The idea that we would become conscious of every detail of the sound---that every impulse in the auditory nerve would reach consciousness---simply by turning our attention "to the sound" (and away from the music) is really absurd.

It takes many years of training but yes, virtually every detail of the sound can become conscious. In fact, a number of jobs within the professional audio industry depend on this ability: Recording Engineers, Producers, Mastering Engineers, Re-recording Engineers, to name a few.

G

 

[JadeEast]

Quote:

Originally Posted by mike1127 /img/forum/go_quote.gif It's not a proposition. An example of a proposition is this: If two systems A and B sound different in any way, under any conditions, then they sound different while listening to "sound as sound." This proposition is assumed to be true, with no evidence, by many DBT advocates. I think the burden of proving it lies with those who assume it.

Seems to me that your reframing my question. I simply was wondering if you would have any way to show or demonstrate the idea beyond a thought experiment.

 

[mike1127]

Hey folks,

Part of me would like to continue this, but I have been through all these arguments before... it is starting to resemble

I realize some of you have valid questions (EDIT: ALL of you have valid question, sorry) that deserve an answer, but I have to pick and choose where I put my energies.

So happy listening!

 

[mike1127]

Quote:

Originally Posted by gregorio /img/forum/go_quote.gif So if I understand your argument, you are saying that changing to an audiophile cable (for example) could change an audiophile's emotional state or expectation and this would then translate into the audiophile hearing a difference between the cables in the music. ... It takes many years of training but yes, virtually every detail of the sound can become conscious. In fact, a number of jobs within the professional audio industry depend on this ability: Recording Engineers, Producers, Mastering Engineers, Re-recording Engineers, to name a few. G

Okay I'm going to resurrect this topic. Maybe another example will explain this more clearly.

I have heard several reports of this sort of happening:

Let's say an orchestra is involved in a recording project. Many decisions need to be made in recording an orchestra, of course: choice of mikes, placement of mikes, choice of mixing and editing equipment, choice of monitoring speakers, etc.

Now let's say among these choices, the recording engineer has arrived at two possible configurations: (A) and (B). The goal now is to determine which configuration sounds most like "the real thing", and the engineer wants to consult the conductor.

Both the engineer and the conductor stand in the hall and listen to the orchestra play, while a recording is made. Then they enter the control room and listen to the recording, and judge its resemblance to the "real thing."

It does happen sometimes that the conductor and engineer disagree which configuration is most lifelike. This would be a possible conversation:

1. Engineer says: In (A), the bass was much tighter.
2. Conductor says: But in (B), the pitch of the timpani was more discernible, and the musical effect was truer to the expression I wanted out of the musicians.
3. Engineer says: In (A) the highs were more extended.
4. Conductor says: But in (B), the piccolo and flute were more differentiated in tone color. In (A) they seemed almost like the same instrument just at different pitch ranges.

... and so on. The engineer and conductor are listening to different aspects of the sound. Their impressions both arise from the sound itself... from the impulses in their auditory nerves. But they each have their own way of translating those impulses into a conscious experience. They have each spent thousands of hours focusing on particular aspects of sound---but different aspects.

I'm not a biologist but I have spent 500 hours training in a sensory awareness technique taught by teachers with expert knowledge of the nervous system, and from what I learned, this kind of disagreement is perfectly consistent with how the brain works.

The impulses in the auditory nerve go through a lot of processing before they reach consciousness. We are not directly aware of them, nor any of our senses.

 

[mike1127]

Quote:

Originally Posted by JadeEast /img/forum/go_quote.gif Seems to me that your reframing my question. I simply was wondering if you would have any way to show or demonstrate the idea beyond a thought experiment.

But I need to make sure we are on the same page. My main point is that there is a common assumption. There is no evidence for this assumption---it arises from a kind of illusion. That's my point.

Again the assumption is: if (A) and (B) sound different under any conditions or mode of listening, then those differences can be made conscious in one particular mode of listening, which is listening to "sound as sound." In the first post I explained why this kind of assumption can arise from an illusion.

Okay, now to be more helpful about your question. Is this assumption false, and how would we prove that? To me, it's plainly false based on my sensory awareness work. I have spent 500 hours in a professional training course that uses sensory awareness as a healing modality, I have spent hundreds of hours in meditation retreats, thousands of hours meditating, and thousands of hours practicing music. Now, I am not an authority. All I can do is offer my experience. I believe that I have learned to be aware at a deeper level how I form my conscious impression of music, and that my observations are broadly applicable. (I can infer they are broadly applicable from my training in modalities with teachers who work with many, many people, and have distilled principles from that.)

I also believe that we can infer something about this idea from an understanding of the brain. Here I am not an expert, but I can relate the general principles I've learned. The brain is incredibly cross-wired and vertically wired, so many parts of our brain can participate in creating the experience of music. One interesting idea is that the motor cortex participates. That's obvious when you dance along to music, but it may also participate with micro movements when you think you are still. (Experiments have shown that if a person does not move but imagines movement, you can register small nerve impulses in their muscles.)

It also seems obvious to me that depending on how you use your attention to movement, you can encourage or suppress the participation of your motor cortex.

So to me, the million-dollar question is: are there details of the sound which are expressed much more in the reaction of the motor cortex than they are in "sound as sound"? If so, encouraging or suppressing your motor cortex would directly affect what details you can perceive.

Now, a common answer to all of this is: "you seem to be speculating." But my point is that those who think that listening to " sound as sound," or that any single mode, is sufficient, are also speculating. In fact worse than speculating: they are operating under an assumption and don't even care to make it explicit.

Furthermore it is my direct experience that some differences come to my attention only through musical or rhythmic qualities, not through the "sound". So that is not speculation at all. Where I'm "speculating" is about the explanatory mechanism.

 

[mike1127]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif Cold fusion does not work , blind tests do.

You missed my point. It was about a logical fallacy. You have implied that it's an invalid (or at least "unhelpful") argument to criticize a particular blind testing protocol without providing a better alternative.

What exactly do you mean by "blind tests work"?

Quote:

I do not introspect on my listening process, I listen. When I listen for pleasure I listen, when I listen for differences I listen, perhaps differently.

It seems to me that if we want to do the best job understanding audio and music perception, we have to do some introspection. Ultimately I think that knowledge we gain from introspection feeds into our experimental designs.


Quote:

You are still working from the assertion that you have to control listeners' listening strategies.

I think we do if we want to have the best chance of understanding what is going on.

Quote:

Not at all, I work with hard to understand things all the time, I have spent the last 12 months analyzing difficult qualitative and quantitive data in great depth, I review academic papers (and publish a few myself) that dig into difficult things in often painful detail, but that is where it was necessary to do so, but the point is that in this case , audio listening, it is simply not necessary to know these things.

What evidence do you have it is "not necessary to know these things"?


Quote:

These conditions are generally stated, we have covered this already. Does this introspection in any way help anybody to detect differences, if so what evidence do you have for this ?

Does lack of introspection have no effect on ability to detect differences? If so what evidence do you have for that?

My point is that none of us have much evidence about relating perception and internal listening strategies to the results of testing.

 

[Pio2001]

Hello Mike, interesting thoughts.

Quote:

Originally Posted by mike1127 /img/forum/go_quote.gif Music is more like a basketball game... too rich to take in everything at once, and there is no way of "laying it all out in front of you."

I agree with this.
One day, in Hydrogenaudio forums, someone pretended to hear easily the difference between Musepack standard and wav. He provided a sample that he had ABXed.
I could not hear any difference. Normally, I would have concluded that they sound identical, but since the guy had succeeded, I listened over and over, until I finally realized that at a given time, one side of the background reverberation of the distorsion of a note of one of the guitars sounded different !
Once this difference identified, the ABX was easy.

I've got a pair of samples at home that I'll post here, that could be very interesting in this discussion. There is a difference between them that should be difficult to find, but that is however into the "shoes / bare feet" category that you spoke about in your other post : I could distinguish which sample I'm listening to, even after a long period and without reference. I just have to remember which is named A, and which is named B.
No musical trick in them, one is just the Vorbis encoding of the other.

Quote:

Originally Posted by mike1127 /img/forum/go_quote.gif It's my opinion that short snippets actually block access to information; namely, musical information.

There are two different cases here.

In many ABX, someone claims that a difference exists, and tries to ABX it. Or you think that you're hearing a difference, and wants to check blinded.
Your statement does not apply in this case, because the listener already knows the difference. The test just has to take place in the same conditions as the ones in which the difference was heard to begin with.

But sometimes, the listeners don't know the difference in advance. For example in public mp3 ABC/HR tests. You usually take part in this test saying "let's see if the difference is audible". Here, your statement, be it true or false, apply.
That's why it is very important in these kind of tests to look for the most possible discriminating conditions. The result entierly depends on the listening conditions. If this kind of test fails, it is not possible to tell if it was because the difference was not there to begin with, or if the listeners failed to find a musical sample, or listening conditions that allowed to reveal it, unless this point was thouroughly investigated before the test, and virtually all possibilities taken into account.

In the first kind of test, in case of failure, a whole range of posibilities is avoided : choice of the sample, listening system... Only some remain : listening fatigue, repetitions...

Quote:

Originally Posted by mike1127 /img/forum/go_quote.gif I will close with an observation of what I believe is a common fallacy. It's this: "Anything you can be conscious of while listening to sound as music, will also come to consciousness while listening to sound as sound."

I am quite in agreement with you, except that in practice, the situation where the difference is musical and not sonic is very rare.
I can think of two distorsions that may fall in this category :

The wow caused by badly centered vinyl records. This causes a slow oscillation of the pitch. On fixed-pitch instruments, like piano, this can appear as a sonic difference. A piano never outputs pitch-varying notes.
But on instruments whose pitch can be modulated, like human voice, the problem can appear as purely musical. The badly centered record is then equivalent to a bad singer.

I think also about dynamics compresion. This is also a kind of sonic alteration that can fall into the musical category, when applied with a given time-delay. It can supress some musical expression, like accentuated attacks, or short-term volume modulation performed by the musician.

 

[Pio2001]

I have no idea how these samples are going to sound for you. I'm curious to know if you can ABX them easily or not.

I can. I just got 16/16 between them. The Vorbis one have been encoded at -q3, and is 110 kbps only.

Original
Vorbis

If you can ABX them, what category would you put them in ? Samples with an overall different feeling, or samples with a detail that has changed ?

 

[mike1127]

Quote:

Originally Posted by Pio2001 /img/forum/go_quote.gif I have no idea how these samples are going to sound for you. I'm curious to know if you can ABX them easily or not. I can. I just got 16/16 between them. The Vorbis one have been encoded at -q3, and is 110 kbps only. Original Vorbis If you can ABX them, what category would you put them in ? Samples with an overall different feeling, or samples with a detail that has changed ?

I listened to these sample. I hate the music here. I guess this tells me something---regarding differences between components, I mainly care about the differences that change the musical feeling with respect to music I love and deeply enjoy. The sound in these samples is dominated by something that may be cymbals or may be electronically generated. There is so much unintegrated sibilance (that is, sibilant energy that doesn't seem attached to a source with body) and the resolution is so low it's impossible to tell.

I have not listened blind. I would have to download the ABX program I think. I'm not sure what software you used to conduct the blind test. I listened to the Vorbis first. About halfway through a very thin piano can be heard through all the hash as the hash recedes somewhat.

Then I listened to the flac. What came to my attention was that the piano had more body and actually could be heard through the hash starting 1/4 of the way through instead of 1/2 of the way.

Regarding your question about whether this is a change in musical feeling or a detail: well, I don't get a lot of musical feeling from this example, so the question is not really relevant.

Maybe that's not right. I did hear the piano doing some interesting rhythmic things and the piano was much clearer, with more body, in the flac. (Non-blind impression.)

I guess it's hard to categorize this example. I described these categories--"music" or "detail"---but I don't always agree with myself.

Some things are not easily categorizeable.

In case you are interested what music is most revealing to me as I conduct tests of cd players, cables, and amps, here is an example. It's a 256 kbps mp3 (the resolution is not the point, though---just giving an example).

My kind of music

I listen for things like:

1. What is the feeling of the rhythm? Driving? Relaxed? Does it have plenty of life and energy without feeling exaggerated? (Linn products seem to exaggerate rhythmic energy.)
2. Tone color. Does the clarinet sound like a clarinet and the oboe like an oboe? Is their tone color clearly differentiated?
3. Microdynamics. Is the subtle swelling and ebbing of the melody clear and easy to focus on?
4. Transparency/separation. If I pick one instrument----say the piano---can it stay clear in my attention even when other instruments enter and do their things? When all the instruments come in together, can they still be heard separately if I wish to do so? Can I tell that the clarinet and oboe articulate the entrance slightly differently?
5. Beauty. No way to describe it. You know it when you hear it.

 

[Pio2001]

Thanks for trying.
This doesn't lead to any conclusion.
I think that I'm going to submit the example to a french forum and see if what I'm thinking about happens. It would prove that sometimes, music is very complex and easily audible differences can be missed even during analytical listening.
Actually, the trick is that the Vorbis sample features a problem that is at the same time easily audible (I think), but not easy to spot. You didn't spot it, by the way.

When I ABX two samples, I use the Foobar2000 player, version 0.9 at least. It features an optional ABX module. Wavoman would like it because unlike usual ABX modules, it has got an Y button that lets the listener listen to the opposite of X (if X is A, the Y button plays B). This way, it is possible to design the listening as you want (AB, ABA, ABX, XX/XY etc).
When I ABX more than two samples, (blind test with several sources and a reference), I use the ABC/HR software. I then simultaneously perform ABX tests between the reference and all other samples, and rank the samples from the closest to the farthest from the reference.

Both softwares allow to set start and stop points inside the samples in order to listen to the part we want.

 

[mike1127]

At the start of this thread, I explained some of the real difficulties in evaluating the musical performance of components. I have made some progress overcoming these difficulties. Read about it here:

http://www.head-fi.org/forums/f21/ho...risons-431057/