How does one get "trained" ears?

[amirm]

Let me add some answers to the original question now:

1. The important thing about training is to start from bottom up. That means whatever distortion you are seeking, you need to hear it in excess and with absolute clarity. Do not start to learn compression artifacts by encoding into AAC at 320 Kbps. Even I can fail such tests. Instead encode something into MP3 at 64 kbps. Do a bunch of comparisons with the original to make sure you can identify what is different.

If you want to hear difference in speaker cable, get a 100 foot spool of cheap speaker cable and test that against your short better one. If that is not enough go to 500 or even 1000 foot. Until you have achieved this step, you cannot proceed to other ones.

In formal listening tests this is called a "control" by the way.

2. After #1, go up one step at a time and see if you can continue to hear the artifacts. If you cannot and the artifacts are reliably there (e.g. in lossy compression) then go back to the beginning again. Cycle back and forth until you can climb the ladder reliably.

3. As I explained to Greg, you must guarantee that the only thing you are using to evaluate these impairments by ear and ear alone. You cannot allow any other senses to intervene or you are fooling yourself to put it bluntly. If you don't believe in this then don't bother going this path.

4. When you go through #1 try a lot of different stimulus (i.e. music). Some are more revealing than others. For speaker and room for example, research shows that content that is full spectrum brings lower threshold of detection. Reason is simple: if there is a frequency response anomaly -- which there always is in this situation -- the content needs to "hit it" or you won't hear the impairment. See this article I wrote for a deeper dive on perceptual aspects of hearing small differences: http://www.audiosciencereview.com/forum/index.php?threads/audibility-of-small-distortions.67/

And for selection of tracks that are more revealing based on research see this thread: http://www.audiosciencereview.com/f...sic-tracks-for-speaker-and-room-eq-testing.6/. If you are looking for music that "moves you," you are in the wrong thread. Again, that is not the task at hand. The task at hand is to hone your hearing. You want to enjoy your music, leave this thread and go do that.

5. Knowledge of the domain can be super useful. It will let you triangulate what you are hearing versus what is supposed to be happening. I think it was Rob who said DIY speaker building helped him hearing artifacts there. As with blind testing, if you are against engineering and science of audio, please don't bother going down this path. The foundation of training listeners is based on what we want to test for. If you don't believe in that, then the value is very limited here. In the Ethan ADC/DAC loop I used this knowledge to zoom into critical segments that were revealing.

6. Domain specific training gets rid of a lot of homework here. Seek those out. They don't always exist and many are proprietary but some are not. For compression artifacts for example search for "MPEG codec killers."

I think that is it for now. I will probably write a longer article on this one day.

 

[853guy]

We don't have to speak in theory here. We have practical proof right here, right now. Ethan post a set of audio files where he washed the same file through A to D and D to A conversion on a cheap sound card a dozen times or so. He then post the files and asked people if they can show they can tell the difference. I did: http://www.whatsbestforum.com/showt...Loopback-files&p=279103&viewfull=1#post279103

"foo_abx 1.3.4 report
foobar2000 v1.3.2
2014/07/18 06:40:07

File A: C:\Users\Amir\Music\Ethan Soundblaster\sb20x_original.wav
File B: C:\Users\Amir\Music\Ethan Soundblaster\sb20x_pass1.wav

06:40:07 : Test started.
06:41:03 : 01/01 50.0%
06:41:16 : 02/02 25.0%
06:41:24 : 03/03 12.5%
06:41:33 : 04/04 6.3%
06:41:53 : 05/05 3.1%
06:42:02 : 06/06 1.6%
06:42:22 : 07/07 0.8%
06:42:34 : 08/08 0.4%
06:42:43 : 09/09 0.2%
06:42:56 : 10/10 0.1%
06:43:08 : 11/11 0.0%
06:43:16 : Test finished."

Vast majority of people cannot. Now, our standard excuse for not taking blind tests is that they stress us, they are artificial, etc. as to explain how they are not able to find such differences. Let's say all of that is true. The fact that my training allowed me to hear those differences blind shows that formal training has that incredible power. It is able to get past our normal excuses and allow us to use our ears and only our ears to hear differences that are there.

Note that I have no formal training in hearing loops through ADC and DAC. What I used however was general techniques that run across all such training. And far, far improve acuity than normal people. Not because I was born with it but because I spent months and years training myself.

In this regard if Mike is going to post here and say that his method is superior to formal training, the blind test is an excellent way to cut through the ego boosting and see where the truth is. He took a blind test and failed. He could not tell his expensive cable from monster when all he could hear is the sound and nothing else. It reasons then that he can benefit from formal training because a test being blind is no barrier to people who are trained.

Okay. Now what?

 

[amirm]

View attachment 30846

Okay. Now what?

Good. Now run the second set in that link with a better sound card. Here are mine again in that post:

"Ah, didn't realize there were a second set of files on that page. Thanks for upping the fidelity and providing the variety. I gave one set of files a try. Here is how I did:

foo_abx 1.3.4 report
foobar2000 v1.3.2
2014/07/29 14:05:31

File A: C:\Users\Amir\Music\Ethan's new generational loss files\focusrite_3a.wav
File B: C:\Users\Amir\Music\Ethan's new generational loss files\focusrite_3c.wav

14:05:31 : Test started.
14:05:50 : 00/01 100.0%
14:06:18 : 01/02 75.0%
14:06:24 : 02/03 50.0%
14:06:31 : 02/04 68.8% -<<< Difference found
14:06:42 : 03/05 50.0%
14:06:53 : 04/06 34.4%
14:07:02 : 05/07 22.7%
14:07:19 : 06/08 14.5%
14:07:35 : 07/09 9.0%
14:08:01 : 08/10 5.5%
14:08:12 : 09/11 3.3%
14:08:31 : 10/12 1.9%
14:08:54 : 11/13 1.1%
14:09:32 : 11/14 2.9% <--- Dog barked.
14:09:52 : 12/15 1.8%
14:10:03 : 13/16 1.1%
14:10:19 : 14/17 0.6%
14:10:53 : 15/18 0.4%
14:11:33 : 16/19 0.2%
14:12:47 : 17/20 0.1%
14:13:18 : 18/21 0.1%
14:13:39 : 19/22 0.0%
14:13:41 : Test finished.

----------
Total: 19/22 (0.0%)"

 

[853guy]

The main problem is that knowledge is an infinitum entity, and we can only grasp a limited quantity of it. In order to be really useful, you have to get a balanced amount of it. A trained listener for high-end purposes will need developing capabilities in several areas of sound reproduction and be able to integrate them critically.

We can learn a lot from the interviews of audio designers - most of the time they explain how they listen and what are their priorities.
I have no problem to accept that our preferences are influenced by what we have read along decades.

Hi Micro,

True. We still only know what we know, and we're only know a little of what we don't know. Like I posted above, I think knowledge on how the brain distinguishes music from sound is invaluable here because without it we continue to use the above methodology that tells us what, exactly?

Again, my strong intuition is that until we can correlate what it is exactly that our brain perceives as music with what we currently understand it to be, the above test is relatively meaningless, except to work out whether I want to buy a SoundBlaster X-Fi sound card or not. Although I did manage to discover I really, really dislike Ethan's skills as a composer.

 

[microstrip]

On the contrary, I agree with the post and the replies you cited. Why should we deny something obvious about human nature? Not everyone has the same talent. I also had to learn to hear some things others were more readily able to perceive.



Why? How does the truth weaken it? Or is the high end about fooling ourselves? In some ways, it seems to be. That is why it's so important to be discerning and to not always run after the latest and allegedly greatest. Keeps you sane.

The question was not on the qualitative aspect of the post, but in the quantitative. Identifying the problems of a few and then generalizing it is not adequate and is not the truth. Very recently we had a thread asking for talented people in the high-end and the list was really large.

BTW, you forgot to quote the part of the post having the answer to your question

(...) Sorry to say, but this kind of arguments only weakens the high-end audiophile position. The high-end should be strong for its respect for diversity and respect for preference, those are just the views of people with extreme positions. Fortunately all WBF members and readers are part of the minor group of competent people.

For the record, I have listened to many excellent systems in homes of friends or acquaintances. Probably not all of them match my preferences, but good and enjoyable systems. But yes, most shops and audio shows have poor sound quality.

 

[853guy]

Good. Now run the second set in that link with a better sound card. Here are mine again in that post:

"Ah, didn't realize there were a second set of files on that page. Thanks for upping the fidelity and providing the variety. I gave one set of files a try. Here is how I did:

foo_abx 1.3.4 report
foobar2000 v1.3.2
2014/07/29 14:05:31

File A: C:\Users\Amir\Music\Ethan's new generational loss files\focusrite_3a.wav
File B: C:\Users\Amir\Music\Ethan's new generational loss files\focusrite_3c.wav

14:05:31 : Test started.
14:05:50 : 00/01 100.0%
14:06:18 : 01/02 75.0%
14:06:24 : 02/03 50.0%
14:06:31 : 02/04 68.8% -<<< Difference found
14:06:42 : 03/05 50.0%
14:06:53 : 04/06 34.4%
14:07:02 : 05/07 22.7%
14:07:19 : 06/08 14.5%
14:07:35 : 07/09 9.0%
14:08:01 : 08/10 5.5%
14:08:12 : 09/11 3.3%
14:08:31 : 10/12 1.9%
14:08:54 : 11/13 1.1%
14:09:32 : 11/14 2.9% <--- Dog barked.
14:09:52 : 12/15 1.8%
14:10:03 : 13/16 1.1%
14:10:19 : 14/17 0.6%
14:10:53 : 15/18 0.4%
14:11:33 : 16/19 0.2%
14:12:47 : 17/20 0.1%
14:13:18 : 18/21 0.1%
14:13:39 : 19/22 0.0%
14:13:41 : Test finished.

----------
Total: 19/22 (0.0%)"

Hi Amir,

But why?

Passing a test such as the above has limited utility value. It's domain specific. It relates solely to whether I can hear differences of a tiny magnitude related to a very, very specific set of circumstances (the degradation of the SoundBlaster X-Fi after one pass). It neither helps me appreciate recorded music when played back via another system more, nor decide on the criteria for how those differences are meaningful outside the parameters of the test itself. It certainly gives me no insight into what our brain's are looking for when it comes to distinguishing music from sound.

So, y'know, I could take test after test, but again, what's the utility value in doing so? A job at Microsoft making codecs?

 

[Gregadd]

Amirm I see no reason to rebut some very valid points. Especially since we have done so before.
I see no reason why audiophiles should not engage in blind test when properly designed and conducted. I have my own problems with such tests. The role of memory and lack of famualrity with the "dut" and the source material. But then nothing is perfect. I do however stand by my belief that no one should accept an ABX challenge or any solo blind test. I discussed this previously.

Just for your edification last I checked there were 2 states that offered alternatives to taking a bar exam as the basis for admission.
I concede professionals require formal training. In addition they better follow accepteted protocol less they be sued by a lawyer or face a disciplinary proceeding by the state licensing board.
OTOH there really is no requirement for someone trying to hack into a computer from his parents basement. Indeed the term "hacker" implies a lack of training.
There is an obvious difference between a wine conniseur and a brewer.
A wine taster may be excellent but not have a clue about brewing.
My point is you keep attacking wine tasters for the lack brewing knowledge.

 

[Al M.]

The question was not on the qualitative aspect of the post, but in the quantitative. Identifying the problems of a few and then generalizing it is not adequate and is not the truth. Very recently we had a thread asking for talented people in the high-end and the list was really large.

Of course. The list of Nobel Prize winners in the natural sciences is also really large. But the list of less talented scientists is much, much larger.

BTW, you forgot to quote the part of the post having the answer to your question

I did not forget. I did not quote that part of your post because I did not think it provided a complete answer. The partial answer that it provides is valid, but it's not the complete picture.

 

[853guy]

Of course. The list of Nobel Prize winners in the natural sciences is also really large. But the list of less talented scientists is much, much larger.

Hi Al M,

Wouldn't it be more true to say that history is only concerned with remembering the winners?

I'd like to suggest that history deems one the "winner" not based on talent per se, but in fact by numerous other factors not limited to politics, socio-cultural alignment, vested interests, the talent pool against whom they were 'competing', and especially, the bias of the voting membership. It's entirely possible the "talent" is not "less" than the one who "wins", and in fact may indeed be equivalent or even superior, but for the reasons mentioned in the previous sentence were not considered "worthy". Talent very rarely comes into it.

 

[amirm]

Again, my strong intuition is that until we can correlate what it is exactly that our brain perceives as music with what we currently understand it to be, the above test is relatively meaningless, except to work out whether I want to buy a SoundBlaster X-Fi sound card or not. Although I did manage to discover I really, really dislike Ethan's skills as a composer.

The test proved things way beyond its borders. You are absolutely right that Ethan's music in that test was terrible. No alignment with anything "live" either. Yet it provided highly purposeful for finding the differences between those clips. You and I demonstrated critical listening ability to hear those differences while not enjoying the music one bit. Enjoyable music may have actually hurt our results if it were not as revealing of what was impaired there.

Thanks for running the test by the way.

 

[Folsom]

In this thread we are discussing what it means to become a trained listener. Mike interjected saying formal training is not necessary and that years of being an audiophile is sufficient. It is *core* to the whole training endeavor to use a 100% reliable feedback loop as to whether the assignments and ultimately the training itself has worked. You can't practice Law in US without passing the Bar exam. Same is true here.

Now imagine if I gave the same test blind. You would hear them both, get confused which is which and vote randomly which using statistics we can then tell that you were unable to identify them as being different. That feedback would then tell you that the files were the same even though during the trial you may have thought there were difference. Immediately you get a sense of what it means to hear the same thing but have the impression of it sounding different.

Take a situation where there were audible differences and two other listeners could hear them blind but you cannot. That immediately tells you that you are not there in your journey of hearing artifacts that are clearly and audibly there. If I ran the same test sighted, all three of you could say they sound different corrupting the training.

In this regard if Mike is going to post here and say that his method is superior to formal training, the blind test is an excellent way to cut through the ego boosting and see where the truth is. He took a blind test and failed. He could not tell his expensive cable from monster when all he could hear is the sound and nothing else. It reasons then that he can benefit from formal training because a test being blind is no barrier to people who are trained.

I'm not sure Mike was comparing being an audiophile to trained listening in the way you are, so much as saying he thinks it's more useful to an audiophile to have audiophile experience; not audiophile experience making you good at DBTs. But for me I don't believe anyone is going to know too much without changing very small components within the appliances. The associative properties need to align more strongly with repeatable changes that can be made, where as changing preamps is swapping a million variables all at once.

And I actually disagree with current forms of doing DBT's because for audio they've proven mostly useless. Consider most would subscribe to the fact that training to hear the differences defeats the purpose of the DBT.

I don't personally believe training defeats the purpose, because people don't think much about hearing the difference in audio so why would one expect them to be able to do it so casually? Adding a power conditioner can change audio reproduction in ways that people can't even imagine until they hear it, so why but the stakes on no training?

Here's the thing, I'm more so under the impression in order to make audio judgements we as humans may need associative properties. Whether we're hearing the difference or not may have some dependency on one's ability to associate it. The obvious way not to do it is to tell them what is what, but to use representative forms. I thought some street intersection light sized bulbs would be good, with changable colors. So you could do things like have 1 color be different and 2 different colors be the same. So the person is still blind to what is what, but knows the rules of the game and has strong properties to associate with.

 

[Al M.]

Hi Al M,

Wouldn't it be more true to say that history is only concerned with remembering the winners?

I'd like to suggest that history deems one the "winner" not based on talent per se, but in fact by numerous other factors not limited to politics, socio-cultural alignment, vested interests, the talent pool against whom they were 'competing', and especially, the bias of the voting membership. It's entirely possible the "talent" is not "less" than the one who "wins", and in fact may indeed be equivalent or even superior, but for the reasons mentioned in the previous sentence were not considered "worthy". Talent very rarely comes into it.

Sure, there are many scientists who probably are as talented as the Nobel Prize winners *), but even if that pool were 10 x larger than the actual winners (quite possible) it would not change anything about the fact that, while there may be many most highly talented scientists, there is still a much larger pool of comparatively less talented scientists (I am one of the latter, by the way, even though I think I am still quite good at my job). So it is somewhat besides the point I was making.

I think obfuscating the issue for the sake of "peace and harmony" (oh, we are all good!) diverts us from the naked truth, which Morricab first brought to the forefront in this thread in post #44.

_________

*) and yes, talent does come into it. Nobel Prize winners are talented. Let's not obfuscate the issue, please.

 

[microstrip]

Of course. The list of Nobel Prize winners in the natural sciences is also really large. But the list of less talented scientists is much, much larger.

I did not forget. I did not quote that part of your post because I did not think it provided a complete answer. The partial answer that it provides is valid, but it's not the complete picture.

Wow, this is what I would call an incorrect use of statistics and a misleading argument, leading us away of the main subject. Nobel prizes are limited to one per year, they are a not a free variable, no way related to the number of scientists, and scientists are not forcefully talented, but can be very competent in their job. It seems to me you have not read with care post #44.

 

[microstrip]

(...) I think obfuscating the issue for the sake of "peace and harmony" (oh, we are all good!) diverts us from the naked truth, which Morricab first brought to the forefront in this thread in post #44.

Sorry, no one is telling "they are all good". Morricab was the one telling "most are mediocre at best in their professional category", something I profoundly disagree.

But OK, we different views on merit in professions and high-end. Fortunately it seems we agree that the high-end products I have paid for were designed by talented people.

 

[NorthStar]

Post #44 ? http://www.whatsbestforum.com/showt...ined-quot-ears&p=433051&viewfull=1#post433051

"I feel that most, if not all of the posts, are talking around the elephant in the room. Namely, that elephant is TALENT. Just like any human endeavor there are those who can and those who try and fail to one degree or another. Think about all the children who put in endless hours per day playing, training, dreaming of being at the top. However, only a handful actually achieve the subtle control, touch and strategic ability to be in the top 10 or even 20 in the world. Plenty of wannabes, only a few with the exquisite ability and training to actually get there.

It is the same in ALL human endearvors; science, engineering, cooking, brewing, wine making, wine tasting, art critique and, yes, hifi critique. There are those who can hear what is right and wrong with a sound, relative to the real thing to one degree or another and those who cannot...regardless of experience. Is it related to intelligence? To some degree, but also a focus and an ability to make connections that others overlook. A great chef might not be an overall genius but he just KNOWS which foods and spices go together...training helps to sharpen the inate skills but only to the limit the ability allows. Lots of practice also helps and can compensate over someone who has great inate talent for hearing and applying but doesn't use it much or doesn't care about hifi so much.

What are the implications for the audiophile community and the WBF in particular? It means that money doesn't matter, experience doesn't matter...much... and exposure to live music doesn't matter UNLESS, you can somehow connect what is hear live to what is heard from playback and find a way to bridge the two. 99% of the systems I have heard convince me that the owner of that system has no clue how to do this and it doesn't matter if he has been an audiophile 2 weeks or 30+ years. I have met people who simply get it from the get go and those who are floating around like a feather in the wind...with no compass whatsoever to guide them.

The more clever of those who are wannabes find someone who is the real deal and emulate them. Maybe they eventually come to realize what made the system of the truly gifted and now their own sound so good but usually they do not and are happy that others compliment them on the system.

So, just like the fact that most doctors (or engineers, or lawyers or scientists or whatever) are mediocre at their profession at best...and incompetent in a surprisingly large number of cases, the same is true of the audiophile in building a truly realisitc sounding system regardless of money or experience. Without the insight to understand what does what, the experience and money (exception: The money can BUY the talent) can't get there. I can train a monkey forever to write a poem and the best I will ever get is gobbeldygook on the screen.

Of course the real problem is demonstrating this convincingly to the outside world that one has talent in audio listening. Other endeavors have metrics that make it possible to do some sort of evaluation...wine tasting does not and you see a lot of variance in opinions...

Anyway, I guess the inability of most people is what keeps a lot of companies in business so it is not bad as it makes jobs." - morricab

 

[Al M.]

Sorry, no one is telling "they are all good". Morricab was the one telling "most are mediocre at best in their professional category", something I profoundly disagree.

I don't. I would not have formulated it in such a harsh way, but the statement that most in their own profession are at a rather average level compared to the very best is pretty uncontroversial, I would think.

 

[Al M.]

But OK, we different views on merit in professions and high-end. Fortunately it seems we agree that the high-end products I have paid for were designed by talented people.

I hope the same holds true for the high-end products I have paid for as well.

 

[microstrip]

I don't. I would not have formulated it in such a harsh way, but the statement that most in their own profession are at a rather average level compared to the very best is pretty uncontroversial, I would think.

I was not addressing your own convictions, BTW quite different from those found in #44, and will refrain from entering such debate.

 

[Al M.]

I was not addressing your own convictions, BTW quite different from those found in #44,

Ok, we probably can agree that Morricab's hyperbolic statement of "99% of the systems I have heard convince me that the owner of that system has no clue how to do this and it doesn't matter if he has been an audiophile 2 weeks or 30+ years", is wrong.

His broader point that some are more talented than others is not.

 

[Muser]

Bobvin:

There is no school to develop "trained ears." There is, however, a test for it. If you have to ask about the test . . . then you're . . . not ready for it. Just kidding.

Seek external validation of your own hearing or capacity for discernment and you'll always come up short to others in some way.

If you enjoy your purchases . . . maybe that's passing "the test."

I hope you can enjoy whatever test you assign yourself.

Larry