Audiofools turned objectivists or vice-versa?

[audio2design]

These are big and complex questions. My sense:

1) Standard DBT protocols were developed by objectivists, who, let's face it, didn't believe there was a difference to be found. Hence, the standard: unfamiliar system, unfamiliar music, unfamiliar room, short bursts and no volume control. Exact opposite of actual listening. Some shortcomings are captured in this paper:

Unfamiliar music, unfamiliar system, unfamiliar room is not at all a requirement of DBT, nor even is short bursts or even no volume control.

Levels must be matched between the samples, that is simple reality of equal loudness curves, and volume preference, but technically you could randomize the volume, so one is not always the loudest (or quietest), but then you will make differentiation even harder.

Short "bursts" are not remotely a requirement of DBT, but from much much testing, it has been shown that this vastly increases the ability to detect differences.

DBT means just that. Double blind. There is no other requirement.

 

[richard12511]

short bursts and no volume control.

These things are actually not negatives at all. Due to how terribly short our audio memory is, "short bursts" are super helpful for teasing out even the slightest of audible differences. These tests we be much, much harder to pass in normal listening(longer music samples) situations. There are definitely many blind tests that have been passed because of those short term bursts that would have never been passed with long term listening.

If you can't hear a difference with short term bursts, you have 0 chance of hearing it in longer sessions. The opposite is not. There are tiny differences you can hear with short clips that you can't with longer clips.

The no volume control is also not just a positive, but necessary for any meaningful results at all.

 

[richard12511]

OK.

Is there a universally accepted DBT test standard, that includes the clip duration and volume control, among other things, as exists in most mechanical testing in my field (ASTM)?

And I am not in my field, do there well maybe one.

DBT doesn't require short clip duration at all. Many of them will let you listen for as long as you want. People generally stick to short clips though since it's way easier to hear differences that way.

 

[DimitryZ]

Unfamiliar music, unfamiliar system, unfamiliar room is not at all a requirement of DBT, nor even is short bursts or even no volume control.

Levels must be matched between the samples, that is simple reality of equal loudness curves, and volume preference, but technically you could randomize the volume, so one is not always the loudest (or quietest), but then you will make differentiation even harder.

Short "bursts" are not remotely a requirement of DBT, but from much much testing, it has been shown that this vastly increases the ability to detect differences.

DBT means just that. Double blind. There is no other requirement.

The paper I posted suggests that short duration music selection reduces listener acuity.

 

[DimitryZ]

These things are actually not negatives at all. Due to how terribly short our audio memory is, "short bursts" are super helpful for teasing out even the slightest of audible differences. These tests we be much, much harder to pass in normal listening(longer music samples) situations. There are definitely many blind tests that have been passed because of those short term bursts that would have never been passed with long term listening.

If you can't hear a difference with short term bursts, you have 0 chance of hearing it in longer sessions. The opposite is not. There are tiny differences you can hear with short clips that you can't with longer clips.

The no volume control is also not just a positive, but necessary for any meaningful results at all.

Paper I posted suggests otherwise.

 

[SIY]

The paper I posted suggests that short duration music selection reduces listener acuity.

That’s incorrect. You can easily prove it to yourself by creating two identical files with a 0.3-0.5dB level difference. Try rapid ABX, then long sequences with gaps.

 

[DimitryZ]

DBT doesn't require short clip duration at all. Many of them will let you listen for as long as you want. People generally stick to short clips though since it's way easier to hear differences that way.

That sounds good.

But it doesn't seem to agree to an actual standard.

 

[SIY]

That sounds good.

But it doesn't seem to agree to an actual standard.

Procrustes nods.

 

[DimitryZ]

That’s incorrect. You can easily prove it to yourself by creating two identical files with a 0.3-0.5dB level difference. Try rapid ABX, then long sequences with gaps.

Have this paper's findings been challenged or disproved?

 

[SIY]

Have this paper's findings been challenged or disproved?

No idea, link doesn’t work for me.

Do the experiment I suggested.

 

[MaxBuck]

This whole colloquy highlights a potential problem I see with the objectivist hifi community. That is, they know the theory very well, but appear not to have developed the procedural discipline to go beyond "just do a double blind test yourself." That's not a formula for building scientific consensus.

 

[DimitryZ]

Procrustes nods.

This seems like Latin and mine is, well, non-existent.

As a Russian/Jewish emigrant to the US, my working languages are Russian and English.
Grandma spoke Yidish, which came handy as our family passed through Austria in the 1970's, but Latin just wasn't in our vocabulary..

Sorry.

 

[DimitryZ]

No idea, link doesn’t work for me.

Do the experiment I suggested.

I am sorry, but that's a coloquial suggestion.

I am not a recognized test entity. There is no personal or community value here.

https://drive.google.com/file/d/1xh7Tno_QoVpDdQTxVlf9C0o-JObtPRO8/view?usp=drivesdk

 

[SIY]

This whole colloquy highlights a potential problem I see with the objectivist hifi community. That is, they know the theory very well, but appear not to have developed the procedural discipline to go beyond "just do a double blind test yourself." That's not a formula for building scientific consensus.

This depends on specifically what question you’re trying to answer. That determines experimental design. Complaining that something designed to answer question A doesn’t answer question B is silly, and sadly common in the faith-based audio world.

 

[SIY]

I am sorry, but that's a coloquial suggestion.

I am not a recognized test entity. There is no personal or community value here.

The value is for you.

 

[raistlin65]

Thank you again.

This is a very thoughtful recommendation. It is not a standard. Is there an industry group, like ASTM that controls DBTs for audio?

And my casual sense (perhaps an incorrect one) is that it's not widely followed. Everyone seems to do their own thing.

Consider that your discipline needs those extensive standards and protocols to prevent error and provide certification to avoid liability. The audio industry doesn't have the same need to certify DBT.

In fact, a large percentage of the industry probably doesn't want anything to do with DBTs, because that would get in the way of what they've been marketing to people. lol

 

[Phorize]

I wonder what the correlation is between audio “objectivist” and atheist or agnostic.

No data to back this up but I doubt they correlate. Atheism is merely an absence of belief, most people just absorb a culture from birth and have no say in what to believe. In societies where atheism is common, such as the U.K where I live, the audio industry makes a big effort to infer both a ‘magic’ or arcane element to audio design and a relationship between cost and sound quality. A separate but highly symbiotic audiophile sub-culture exists, where scientific ignorami acquire social standing by promoting an entire body of pseudo-knowledge. I’d imagine that this pincer movement leads most people in societies where religious non belief is common, if they think about it at all, to assume that expensive equals better and that audiophile ‘experts’ have special knowledge that enables them to ‘hear’ transparent electronics. I have spent along time practicing martial arts/combat sports and this sort of phenomenon is very common in that field, or at least it was until Youtube gave us the spectacle of Tai Chi masters with decades of training getting pulverised in seconds by people with rudimentary boxing skills. Interestingly there is no amount of evidence that will lift the cloud of delusion with some, luckily the average latte drinking atheistic Aikido practioner will avoid violence purely by being a nice person and frequenting coffee shops rather than downtown night clubs so will be none the wiser. I love humans but we are a funny bunch.

 

[DimitryZ]

The value is for you.

Understood.

 

[DimitryZ]

Consider that your discipline needs those extensive standards and protocols to prevent error and provide certification to avoid liability. The audio industry doesn't have the same need to certify DBT.

In fact, a large percentage of the industry probably doesn't want anything to do with DBTs, because that would get in the way of what they've been marketing to people. lol

There is a fair amount of discussion of the right and wrong way to do DBTs. If there were a couple of widely accepted standards, at least we wouldn't have that to argue about

 

[SIY]

There is a fair amount of discussion of the right and wrong way to do DBTs. If there were a couple of widely accepted standards, at least we wouldn't have that to argue about

I’m not sure that an ASTM test is a reasonable analogy. DBT is a basic control, not a method.