[WindowsX]

I think this kind of thing happen to every audiophile. There must be a thing or two that you can't perceive the improvement or even difference with some changes. But one day you can notice the changes. How can we explain this in scientific terms?

 

[VNandor]

Ah, that's a question I've wondered about a lot, I acutally wanted to create a thread for that.
People throw terms around like "audibly transparent" a lot and I have no precise idea of what that actually means.
 
For example  let's say at first I don't hear differences between low bitrate lossy codecs and high bitrate ones. Then, someone points me at the right direction and suddenly the difference becomes quite apparent, easy to discern.
 
Now my question is do the two things count as audibly different for me from the beginning (as I ultimately have the ability to tell them apart), or they only become audibly different after I know what to listen for? And is there a definition for audibly transparent/different in general? I mean theoretically two things can be audibly different maybe we just don't know what to listen for but that, in practice means that they are audibly transparent.
 
I hope I don't derail your thread too much, answering my questions pretty much requires to answer yours for first.

 

[reginalb]

  I think this kind of thing happen to every audiophile. There must be a thing or two that you can't perceive the improvement or even difference with some changes. But one day you can notice the changes. How can we explain this in scientific terms?

 
You might be asking this question in the wrong section of the forum. So what is the specific "difference" that you're hearing over time? Difference with a piece of equipment, a song? What is the actual difference that you're hearing? More detail, different frequencies? A headphone that sounds better to you over time?
 

  Ah, that's a question I've wondered about a lot, I acutally wanted to create a thread for that.
People throw terms around like "audibly transparent" a lot and I have no precise idea of what that actually means.
 
For example  let's say at first I don't hear differences between low bitrate lossy codecs and high bitrate ones. Then, someone points me at the right direction and suddenly the difference becomes quite apparent, easy to discern.
 
Now my question is do the two things count as audibly different for me from the beginning (as I ultimately have the ability to tell them apart), or they only become audibly different after I know what to listen for? And is there a definition for audibly transparent/different in general? I mean theoretically two things can be audibly different maybe we just don't know what to listen for but that, in practice means that they are audibly transparent.
 
I hope I don't derail your thread too much, answering my questions pretty much requires to answer yours for first.

 
Audible transparency in an DAC/Amp means that it doesn't color the sound. That is, the frequency response remains neutral, playing back each frequency from the file at the level in the file, and you don't add audible distortion, too much jitter, etc. This is possible to do at a low cost these days, just look at the ODAC/O2 amplifier for an example. Here is a completely transparent combo, that is, distortion and jitter are below the level that is audible to human ears, and the frequency response is flat (it also has a low enough output impedance, high enough power, and variable gain that allows the O2 to drive any headphone in existence with the exception of some electrostatic guys). 
 
Some people talk about subjectively good coloration. "Warm" sound (usually associated with rolled off treble) "Sweet" sound of tube amplifiers (caused - I think - by audible distortion). That means you're listening to your equipment, not the music as it was intended by its artist when it was created. And if you like rolled off treble, you don't need to spend a bunch of money on a DAP that does it, you can just use an EQ to achieve the same effect. 
 
Regarding learning to hear the difference between lower and higher bit-rate encodings, maybe that's a bit more difficult, but I would say that a lossy encode reaches transparency when humans can't detect a difference with their ears. Any bitrate where you can't hear the difference is definitely good enough for you, however. I now have an "Original" library, where everything is in the original format that I ripped to/downloaded in. Which is backed up, and a couple copies of a second library with everything encoded with Opus (256k VBR). I've never even bothered to ABX the Opus vs. anything else, not worth it. I also have a third mini-library where I re-encoded a bunch of my ogg vorbis CD rips to MP3 for my Sony that can't play back the oggs.

 

[sonitus mirus]

  I think this kind of thing happen to every audiophile. There must be a thing or two that you can't perceive the improvement or even difference with some changes. But one day you can notice the changes. How can we explain this in scientific terms?

 
I had a bit of breakthrough in the "Silver Ears" mp3 artifacts test with the now defunct Philips Golden Ear Challenge.  I was having trouble getting past this point on my EQ'd speaker system, yet I was able to pass this same test with my Denon D5000 headphones.  What I discovered was that I was hearing a slightly elevated bass volume level, which was much more prominent with the particular pair of headphones that I used.  Once I figured out what to listen for, I was then able to easily pass this test with ease using just about any transducer.  I suppose, in a sense, that I trained myself how to detect a difference.  From what I read, this part of the test was a bit of a chore for many of us.  If I didn't have the benefit of the bass-heavy headphones, I may not ever have noticed the key component that made it possible for me to hear a difference.
 
I'm not an audiophile, just a music lover.  I'm just looking for noise-free music that sounds great up to very loud volume levels.  I get that from a lot lower quality components and sources than most audiophiles would appreciate, though I'm not convinced that they could actually tell.

 

[jcx]

no one familiar with psychoacoustics and professional audio denies there is learning, and training can make this faster - without references and regularly doing controlled listening however you can easily build a house of cards rather than objective skill
 

Where expertise is acquired in appropriate environments with adequate experience and feedback, it can be highly effective. In particular, when feedback quality is high (frequent, prompt, and diagnostic) and judgments are made in exacting environments (where mistakes are costly), expert knowledge is likely to be accurate. For example, chess players (Chase and Simon 1973 ) , weather forecasters (Murphy and Winkler 1984 ) , athletes (Ericsson et al. 2006 ) , and physicists in textbook problem solving (Larkin et al. 1980 ) all display highly skilled expertise, developed through experience over an extended period in conjunction with consistent and diagnostic feedback. When feedback quality is low, or when mistakes are not costly to those making the estimates, inaccurate beliefs are easily acquired. In such environments, experts are likely to have difficulty separating the influences of skill from those of chance and are likely to form superstitious beliefs (Kardes 2006 ) .

 
 
responsible pro mixing engineers like Bob Katz can likely make much finer distinctions than anyone with less time spent  - but he is absolutely committed to properly blind, controlled listening to verify what he is really hearing
 
many sound/music production pros however don't verify their impressions - you can find many superstitious beliefs floating around pro audio engineer forums - some claiming with unjustified hubris that their beliefs must be right because their product sells well
 
 
I would advise reading some Psychoacoustics and there are a few educators in the music production business - Moulton's articles, blog is one source: http://www.moultonlabs.com/

 

[castleofargh]

 out of a few typical testimonies I can think of, most are unproven(so useless to try and conclude anything), and a good deal come from using the wrong method to look for something. like trying to hear a difference between 2 sounds by using sighted evaluation with long delays between the samples. or looking for taste preferences with an abx test ^_^.
 
I see 4 big groups but I'm sure we could make more sub groups:
1/ recognizing something in the music where the brain uses memorized patterns to identify them. human speech, a guitar, a "realistic" cymbal. all coming from the model we already have stored from previous experiences. that's not really hearing, but of course it can have a drastic impact on the things we will notice. once we have identified/memorized a given pattern, it will be much easier to recognize it again. like how listening to a song many times makes it easier to recognize the song when we hear it.
 
2/ conscious search, like trying to find out if there are 2 or 3 guitars playing. 1/ can help for sure, but the most obvious is that we need to ask ourselves that question before we can hope to answer it. if you take 10 people and ask them to count the guitars, they will obviously do better than if you let them listen and only afterward ask them how many guitars were playing. that's asking ourselves a question and concentrating on that one thing. IMO if the guy notices something on his own, it's not transparent, if he doesn't then it is transparent(for him). if I need 3years of training to recognize a wine from another one, maybe I shouldn't bother about the difference in the first place.
 
 
3/ identifying a change in air pressure. also known as sound. the only situation where we're looking at actual hearing ability IMO. if something changes too slowly or too fast, my ears may fail to notice it(5hz or 25000hz). same with amplitude, if the variation is too small I may also fail to hear it. that's actual hearing and it won't get better after I have spent 3 days listening to the same song again and again. 
 
4/ placebo. given enough time, every memory will fail enough so that the same sound now feels like it was different a long time ago.
 
 
 
 
if we look for something particular in a song(like 2/), then of course more time can help, just like how given enough time you will find waldo in the picture.
but if we look for any sort of audible variation between 2 sources, then the best method seems to point toward short samples and rapid switching instead of longer time. I honestly don't know if having a long time for such a test can help for anything. my educated guess is no, but only because with more time we soon fall into the "memory loses accuracy over time" problem. so depending on what you're looking for, I would say that more time could end up being a handicap as it will increasingly move us toward 4.

 

[sterling1]

This is easy, perceptions change with sense of a physical change. Things sounding different, better, worse, more detailed, etc., result from things like humidity,  speaker deterioration, volume and on and on. The difficult thing is sometimes determining the source of what is making things sound different. It could be a multitude of things. But, here's the bottom line, a lot of experience or practice "listening" is necessary to be sensitive enough to notice something different. We're talking Princess And The Pea sensitivity here.  

 

[reginalb]

   out of a few typical testimonies I can think of, most are unproven(so useless to try and conclude anything), and a good deal come from using the wrong method to look for something. like trying to hear a difference between 2 sounds by using sighted evaluation with long delays between the samples. or looking for taste preferences with an abx test ^_^.
 
I see 4 big groups but I'm sure we could make more sub groups:
1/ recognizing something in the music where the brain uses memorized patterns to identify them. human speech, a guitar, a "realistic" cymbal. all coming from the model we already have stored from previous experiences. that's not really hearing, but of course it can have a drastic impact on the things we will notice. once we have identified/memorized a given pattern, it will be much easier to recognize it again. like how listening to a song many times makes it easier to recognize the song when we hear it.
 
2/ conscious search, like trying to find out if there are 2 or 3 guitars playing. 1/ can help for sure, but the most obvious is that we need to ask ourselves that question before we can hope to answer it. if you take 10 people and ask them to count the guitars, they will obviously do better than if you let them listen and only afterward ask them how many guitars were playing. that's asking ourselves a question and concentrating on that one thing. IMO if the guy notices something on his own, it's not transparent, if he doesn't then it is transparent(for him). if I need 3years of training to recognize a wine from another one, maybe I shouldn't bother about the difference in the first place.
 
 
3/ identifying a change in air pressure. also known as sound. the only situation where we're looking at actual hearing ability IMO. if something changes too slowly or too fast, my ears may fail to notice it(5hz or 25000hz). same with amplitude, if the variation is too small I may also fail to hear it. that's actual hearing and it won't get better after I have spent 3 days listening to the same song again and again. 
 
4/ placebo. given enough time, every memory will fail enough so that the same sound now feels like it was different a long time ago.
 
 
 
 
if we look for something particular in a song(like 2/), then of course more time can help, just like how given enough time you will find waldo in the picture.
but if we look for any sort of audible variation between 2 sources, then the best method seems to point toward short samples and rapid switching instead of longer time. I honestly don't know if having a long time for such a test can help for anything. my educated guess is no, but only because with more time we soon fall into the "memory loses accuracy over time" problem. so depending on what you're looking for, I would say that more time could end up being a handicap as it will increasingly move us toward 4.

 
I do think that, with headphones, there is an element of burn-in, but the element that's getting "burned-in" is your brain, not the drivers, as you get used to a sound signature. Headphones, which definitely sound different, can take some time to get used to. And even then, I think for those of us with a few pairs, the "burn in" is that we learn what music they sound best with. My AKG Q701s are a bit on the thin side in the low frequencies, so I don't listen to a ton of hip hop on them. I have played hip hop on them, and it didn't sound great. Now I don't do that, so the average performance of them has improved. 
 
Audiophiles ascribe things that their brains are doing to their ears a lot, I think. 

 

[HiFiChris]

 How do you explain the phenomenon when you first can't hear the difference but you can later?

 
I haven't experienced this yet, so when I thing about it, I would call it "imagination". 

 
When I hear/perceive something right from the start, I usually also do it later.
On the other hand, when I have heard/perceived no difference right from the start, this did not change over time.

 

[Watagump]

Delusion?
Brain burn in?
Flat out wrong?

 

[jcx]

there is learning, you can train - and you should have objective scoring, immediate feedback to prevent building imaginary correlations
 
the Harmon Golden Ears testing did just that
 
some things have been positively heard as different in ABX by looking at the statistics even when the participants couldn't articulate the differenece, sometimes even when thay were uncertian they were hearing a difference at all - forced choice can be good even if it "causes stress"
 
you can learn frequency bands, boosts/cuts, focus on attacks, decay, timbre...
 
 putting (objective, officially agreed) names to difference sonic features seems help expanding the distinctions that can be made so expanding your vocabulary and ear discrimination in objective dimensions is possible, should be encouraged
 
 
and you can fool yourself, some objectively false memes become imbedded in subcultures
 
but that doesn't mean there aren't objectively real dimensions to hearing or that they can't require learning, training, possibly some only acessable to "master" level thousands of hours of critical engagement

 

[castleofargh]

  there is learning, you can train - and you should have objective scoring, immediate feedback to prevent building imaginary correlations
 
the Harmon Golden Ears testing did just that
 
some things have been positively heard as different in ABX by looking at the statistics even when the participants couldn't articulate the differenece, sometimes even when thay were uncertian they were hearing a difference at all - forced choice can be good even if it "causes stress"
 
you can learn frequency bands, boosts/cuts, focus on attacks, decay, timbre...
 
 putting (objective, officially agreed) names to difference sonic features seems help expanding the distinctions that can be made so expanding your vocabulary and ear discrimination in objective dimensions is possible, should be encouraged
 
 
and you can fool yourself, some objectively false memes become imbedded in subcultures
 
but that doesn't mean there aren't objectively real dimensions to hearing or that they can't require learning, training, possibly some only acessable to "master" level thousands of hours of critical engagement

I have no doubt that we can learn to notice more. but it's not so much hearing and more about brain plasticity. just like playing online FPS can help you see "better" in the long run. the vision itself doesn't improve, but the brain will get better at interpreting micro contrasts or small movements. that doesn't come by itself, it's by actively trying to see the bad guys that our brain will develop a routine for it. just sitting in my chair listening to music while reading headfi will not develop that kind of improvement. it's an active effort based on something I can actually notice from the get go, it just becomes easier with training.
 
but in audio how do we factor this without involving the "new toy effect"? the novelty factor will be completely different if I'm the owner or not, if it was expensive, if it's FOTM... we need a psy more than an audio expert for those stuff. and I wonder if what OP talks about isn't just getting out of that new toy period? how do we check for this?

 

[ChrisIsAwesome]

For me it's been a learning experience. Most people never concentrate on music like audiophiles do.
 
Take for example, foreign movies. You can watch foreign movies for hundreds of hours while reading subtitles and learn almost nothing of the language that it's in except for maybe a word or two. Only when you start focusing and studying can you learn to pick up on what they're saying.
 
For me the learning experience has been from owning a tube amp and rolling tubes. After experimenting with comparing tubes/sources/headphones/amps I've learned what to listen for.
 
I can listen for each sound, the delay between them 'responsiveness', how fast the transition is 'separation', how clear each note is 'clarity', how much impact it has and whether there is any distortion. These are things i never knew to listened for because I'd never even taken the time to concentrate and focus on what i was hearing also i had no opportunity to experience multiple difference systems.

 

[catspaw]

In simple terms, in most occasions its not the sound that is changing but our perception of it.
 
A good example are those optical illusions. Humans simply are not good data analyzers.