Can we trust our ears?

[Blumlein 88]

That is fine for results of things like distortion or frequency response aberrations. It doesn't really make sense for perceptual encodins like mp3.

Maybe, maybe not, but if you stick with the ABX comparator I don't see a problem.

I suppose my comment about not hearing a null file was incomplete. If you cannot hear the nulled result at normal listening volume I'd say you can't hear a difference. It will sometimes be true even when you hear something of the nulled result you still would be unable to audibly separate the two files. So one case rules out hearing a difference, and the other is inconclusive unless you move onto the ABX attempts.

 

[pkane]

That is fine for results of things like distortion or frequency response aberrations. It doesn't really make sense for perceptual encodins like mp3.

I'm not sure why. An ABX test is an ABX test. DeltaWave lets you run an ABX test, as well as a blind preference test between two files. It doesn't matter how those files are encoded. What it will do for you is make sure that the levels are matched as perfectly as possible and will align the start of both files to less than a tiny fraction of a sample, so you can compare apples to apples.

 

[Twitch54]

which all goes to prove we listen with our brain, our ears are merely the conduit ..............

 

[pkane]

which all goes to prove we listen with our brain, our ears are merely the conduit ..............

That is clearly the way it is. Getting a bit more philosophical, everything we call reality is experienced only through our brain

 

[Julf]

Maybe, maybe not, but if you stick with the ABX comparator I don't see a problem.

Inded. ABX is no problem, null test is.

 

[Julf]

I'm not sure why. An ABX test is an ABX test. DeltaWave lets you run an ABX test, as well as a blind preference test between two files. It doesn't matter how those files are encoded. What it will do for you is make sure that the levels are matched as perfectly as possible and will align the start of both files to less than a tiny fraction of a sample, so you can compare apples to apples.

Yes, the ABX test is valid even for mp3. The null test isn't - that was my point.

 

[anmpr1]

At the Stereophile site, reviewer Jim Austin argues against ABX testing because, in his words, "testing takes time".

Subjectivist audiophiles have long maintained that long-term listening is necessary to assess the quality and character of an audio component. Scientific testing methodologies such as ABX, which require quick and conscious evaluation of a change in the sound, have long struck many of us as insufficient, seeming to miss much that affects our enjoyment of music.

Could someone over @ Stereophile clue Jim in that an ABX comparator test allows one to listen as long as they like before switching? And then, after switching, one can take as much time as they want in order to determine whether X is A or B? Interesting how these 'subjectivists' can describe every little detail in the sound of amps/preamps/CD players, with utmost precision, but when asked to simply tell us which is which, they get all confused and wind up guessing.

Citing some or another 'study' he writes: Based on a limited perceptual bandwidth and 8 hours of dedicated listening per day, getting to know a room and equipment in any detail would take at least a week, but assuming years would be safer...

So let's break this film down. It takes a minimum of a week of listening to reliably tell these 'sonic' differences, but a year is better? Is he shilling for long term loaners? And if that's the case, why can their 'reviewers' immediately tell the difference among components? I mean, one guy's wife was able to tell differences in gear from the kitchen, when the reviewer swapped out components in the living room. I think they let that guy go, though. Maybe he wanted his wife to get paid for her contributions, and the editor said no. LOL

How anyone can take these folks seriously is something serious people seriously can't.

 

[Kal Rubinson]

Could someone over @ Stereophile clue Jim in that an ABX comparator test allows one to listen as long as they like before switching? And then, after switching, one can take as much time as they want in order to determine whether X is A or B?

Why don't you do it? It is not necessary to join or register in order to comment.

 

[0bs3rv3r]

Why don't you do it? It is not necessary to join or register in order to comment.

No, don't do it. I just stopped frequenting just such a forum (different one) because of the furor such a comment would cause. Nothing sensible will come of it.

 

[audiophile]

At the Stereophile site, reviewer Jim Austin argues against ABX testing because, in his words, "testing takes time".

Jim Austin seems to be more of an objectivist to me. He prefers transparent sounding DACs and once famously double-blind tested dCS Bartok DAC (yes, DBT, in Stereophile with levels matched !) The manufacturer wasn't happy about it ))

Citing some or another 'study'

His article summarizes two AES papers by Genelec researchers:

Time for Slow Listening
http://www.aes.org/e-lib/browse.cfm?elib=20547

On Human Perceptual Bandwidth and Slow Listening
http://www.aes.org/e-lib/browse.cfm?elib=19621

These papers basically claim that short ABX tests are inaccurate since humans can analyze maximum 40-50 bits of audio data per second.

 

[Blumlein 88]

Jim Austin seems to be more of an objectivist to me. He prefers transparent sounding DACs and once famously double-blind tested dCS Bartok DAC (yes, DBT, in Stereophile with levels matched !) The manufacturer wasn't happy about it ))


His article summarizes two AES papers by Genelec researchers:

Time for Slow Listening
http://www.aes.org/e-lib/browse.cfm?elib=20547

On Human Perceptual Bandwidth and Slow Listening
http://www.aes.org/e-lib/browse.cfm?elib=19621

These papers basically claim that short ABX tests are inaccurate since humans can analyze maximum 40-50 bits of audio data per second.

The Stereophile article leaves out everything you want to know. Hints at much, tells us little. The only blogs on the subject by the authors from Genelec tell little, but indicate in a professional environment speakers shouldn't impose restrictions you have to take time to learn. They mostly are about poor directivity in speakers in line with the thinking from Harman being correct. Hardly the idea hinted at by Stereophile. Maybe someone with access can tell us what is actually in the papers.

There already is a thread to the Hi rez article in stereophile and the poorly done listening test it is about.

 

[Blumlein 88]

Okay here is a 30 minute video of one of the presentations on Human perceptual bandwidth by Thomas Lund. I've not watched it yet.

 

[audiophile]

Maybe someone with access can tell us what is actually in the papers.

Here's the most recent paper, published in September:

 

[Blumlein 88]

What a silly paper. Where is the beef? Lots of very loose conjecture reminding me of a solution in search for a problem.

For what it is worth, when I've done careful blind listening with software, as I near thresholds of perception, there is a difference I am no longer conscious of hearing. I'd say I am guessing, and yet I will select correctly with high reliability. So I don't see the issues in this little write up as contradictory to blind listening with quick switching and short intervals. We already know trained listeners, and training on particular features enhance the ability of people to score at a higher level of discrimination. Maybe many hours acclimating to a playback environment enhance it just a little more ( I wouldn't be surprised by that). But there is nothing here to indicate other methodologies would result in better results. The lone exception would obviously be a listener test of listener fatigue and that has lots of thorny issues to be any good already.

Oh, and the Stereophile interpretation was a large over-reach. Exceptionally reaching for what isn't in the paper.

 

[RayDunzl]

At the Audio Show last year, my only chance taken to hear different modern speakers... We pretty much ignored electronics.

Most sounded about the same, casually listening as the opportunities presented themselves..

There were few standouts on either side of the scale.

Can't say anything sounded bad, just, maybe didn't meet expectations.

Two months to wait till my next chance.

 

[anmpr1]

Why don't you do it? It is not necessary to join or register in order to comment.

It was a rhetorical declamation. Like the folks at Stereophile want to hear it from me. But thanks for the good laugh, so early in the AM, Kal. You've restored my faith in humanity. PS: How did you know I was talking about you? And Merry Christmas, by the way. That's coming up, you know.

Yet I think I will learn from your possibly implied point, and refrain from this kind of thing anymore. I mean, how autistic does one need to be to keep going around in circles with this stuff, and with these people? Probably time to move on.

A final word. The 'article' or opinion piece is just another grasping (or is it gasping?) at whatever low hanging fruit can be plucked in order to justify a spurious methodology... really it is a lack of methodology. At this late stage of the game it's so transparent and very tiresome. The only reason it is even worth commenting upon is because just when you think it can't get more ridiculous, it does. Sort of like that old Hunter Thompson quip: When the going gets weird, the weird turn pro.

John Curl said it best. It was in some stuff reposted here from another forum, or something. I paraphrase, of course, but John admitted that, "Yeah, if you match the levels and hide the brands, no one can tell my expensive preamp from the Pioneer receiver. But if you want my special kind of electronic magic, you gotz to give me your money. And a lot of it. I promise that in return you'll get all the magic you can ever want." At least until the next upgrade...

 

[audiophile]

What a silly paper. Where is the beef?

Not sure what you mean? The authors wrote their paper referring to decades of research in the field of information theory which explains how we hear things and what’s happening in the brain when we process what we hear at the max. speed of 40bps. They just applied all those findings to subjective listening tests and made their conclusions. That’s how I understood it.

 

[Blumlein 88]

Uncontrolled listening. I stole that from Don. The polite version for lack of methodology.

 

[Julf]

Not sure what you mean? The authors wrote their paper referring to decades of research in the field of information theory which explains how we hear things and what’s happening in the brain when we process what we hear at the max. speed of 40bps. They just applied all those findings to subjective listening tests and made their conclusions. That’s how I understood it.

The problem is the application of the findings and the conclusions.

 

[audiophile]

The problem is the application of the findings and the conclusions.

Can somebody please explain what is the problem?