Can You Trust Your Ears? By Tom Nousaine

[Cosmik]

Which means you probably can’t go terribly wrong with hi-res files...

Or RBCD. And it's cheaper on all counts.

 

[Arnold Krueger]

You are right; there’s no single empirical research piece out there that concludes once and for all that higher than CD quality is distinguished or better. Not long ago a meta analysis published by AES (I guess everyone read that one? Else, I can post link) stated that higher is distinguishable from lower resolution. However, the researchers needed the «magic» of large samples (meta) to conclude thus.

Are you familiar with the fallacy of false equivalency?

Are you familiar with the fallacy of overemphasizing secondary issues to the point of missing the point of a primary one?

The false equivalency is arguing the there is no evidence that says that CD format sounds better. The actual argument all along, the argument that has given rise to dozens of studies and papers since 1973 , is the argument that so-called high-resolution recordings make no audible difference. The most conservative factual outcome of that argument on the pro-hi-rez side is that right rez makes very little difference.

The overemphasized secondary issue is the real and not uncommonly manifest fact that sometimes high rez sounds worse because of things like IM of ultrasonic sounds into audible distortion manifested below 20 KHz.

 

[svart-hvitt]

Are you familiar with the fallacy of false equivalency?

Are you familiar with the fallacy of overemphasizing secondary issues to the point of missing the point of a primary one?

The false equivalency is arguing the there is no evidence that says that CD format sounds better. The actual argument all along, the argument that has given rise to dozens of studies and papers since 1973 , is the argument that so-called high-resolution recordings make no audible difference. The most conservative factual outcome of that argument on the pro-hi-rez side is that right rez makes very little difference.

The overemphasized secondary issue is the real and not uncommonly manifest fact that sometimes high rez sounds worse because of things like IM of ultrasonic sounds into audible distortion manifested below 20 KHz.

Well, I’m in the business of understanding data (not important) for over 20 years now (important).

It would be more helpful if you pointed out where I make errors in my short text. I often make errors and try and learn from them, even if only grammatical and choice of words errors.

Please note that my concluding remark was intentionally written to mean almost everything, depending on how the reader chose to understand the words.

 

[amirm]

I don't think there has been any attempts to show that high-res produces worse results than CD. If we wanted to show such, we would intentionally pick content that would show such on basis of IM or other theories like that. Most of us have high-res content so why not try to and find such cases and post blind results of high-res making things worse? That would be more constructive than forcing a theory to be fact on others who are more cautious in that regard.

 

[Fitzcaraldo215]

You are right; there’s no single empirical research piece out there that concludes once and for all that higher than CD quality is distinguished or better. Not long ago a meta analysis published by AES (I guess everyone read that one? Else, I can post link) stated that higher is distinguishable from lower resolution. However, the researchers needed the «magic» of large samples (meta) to conclude thus.

And the fact that higher is distinguishable from (after hundreds of listeners made thousands of listening tests) lower doesn’t mean that higher is perceived as better; because lower can, by many, be perceived as better.

Which means you probably can’t go terribly wrong with hi-res files...

I don't quite agree with your summary of the Joshua Reiss meta analysis paper:

http://www.aes.org/tmpFiles/elib/20171103/18296.pdf

In 7 of the 17 separate and independent studies he reviewed and summarized where test subjects were trained in advance (Fig. 2 in the paper), hirez was distinguished on between 56.9 and 74.7% of the trials. Agreed, this does not answer the question of a quality preference, but the distinction was identified to a level of statistical significance in DBTs. So, at least some tests confirmed that. But, they also did not identify why a difference was heard, and it is possible that adverse artifacts from ultrasonics - IMD, etc. - in hirez were fully or partly responsible, though I personally consider that unlikely.

I keep saying that I do not think hirez is a slam dunk, night/day difference maker. Statistics has to be used to provide a reasonable case scientifically, and some studies just did not identify much ability to discriminate. Advance training in DBT testing and the development of a perception for distinctive sonic cues by the test subjects seems to be a persistent major factor. I expect that is true for most DBTs.

I also for myself agree that you cannot go terribly wrong with hirez. I like it.

 

[Jakob1863]

Frank, we have rules.. you want to argue bring evidence that’s pertinent beyond ‘what you reckon’ less you be restricted from yet another thread.

Don´t ask for too much. If you restrict any participation to those who could supply clear evidence for every thought they post, this will imo become a lonely place.
Some speculation and maybe even provoking thoughts should be allowed.

Especially in this case where it depends so much on the internal processes of different individuals. JJ emphasizes on the proper cues for depths that are needed and all the difficulties that are tied to two channel stereophony but otoh it nevertheless seems to work but depends - as was already known around the 1930s (see the linked auditoryperspective collection of papers) - on the experience of listeners.
And imo there is a connection to ASA, as that seems to be easier if the information is coming from a clear defined location in the space (even if that clearly defined location is just an illusion). If the illusion of depth of image works better, due to individual differences, it is not so hard to imagine that the sense of realism is greatly enhanced.

I don´t know if you have experienced the same, but sometimes it is really interesting to feel yourself struggling with diverging informations from your senses as the hearing sense tells you that there must be something to look at (i.e. between the loudspeakers and in different layers of depth) while you don´t see anything.

 

[Jakob1863]

The article i weakly remembered was:
Nichiguchi et al.; Perceptual discrimination of very high frequency components in wide frequency range musical sound, Applied Acoustics 70 (2009) 921–934

 

[Kal Rubinson]

The article i weakly remembered was:
Nichiguchi et al.; Perceptual discrimination of very high frequency components in wide frequency range musical sound, Applied Acoustics 70 (2009) 921–934

Thanks. I just downloaded it and it looks very interesting.

 

[amirm]

Indeed. Unfortunately it is a bust with regards to evidence of non-linear distortion in the tweeter appearing in audible band and they conclude thusly:

 

[svart-hvitt]

I don't think there has been any attempts to show that high-res produces worse results than CD. If we wanted to show such, we would intentionally pick content that would show such on basis of IM or other theories like that. Most of us have high-res content so why not try to and find such cases and post blind results of high-res making things worse? That would be more constructive than forcing a theory to be fact on others who are more cautious in that regard.

Remember, the master is often hi-res. The master is highest-fi what concerns source material options, and must therefore be better than lower-res in hi-fi terms.

 

[fas42]

I don´t know if you have experienced the same, but sometimes it is really interesting to feel yourself struggling with diverging informations from your senses as the hearing sense tells you that there must be something to look at (i.e. between the loudspeakers and in different layers of depth) while you don´t see anything.

Interesting that you find that this is a "struggle" - I very specifically look for this aural behaviour; that everything in the soundfield exists in a virtual space; any awareness that the speakers are the source of the sound means that the system is below par. If a curtain was placed between you and real performers on a stage, say, would you be troubled because you couldn't actually sight the particular instrument being played? In my case, no; and listening to playback "without anything there" is equivalent.

 

[svart-hvitt]

Interesting that you find that this is a "struggle" - I very specifically look for this aural behaviour; that everything in the soundfield exists in a virtual space; any awareness that the speakers are the source of the sound means that the system is below par. If a curtain was placed between you and real performers on a stage, say, would you be troubled because you couldn't actually sight the particular instrument being played? In my case, no; and listening to playback "without anything there" is equivalent.

Are you talking about performers with or without amplification-speakers?

 

[fas42]

Are you talking about performers with or without amplification-speakers?

Both. A rock recording will typically have electric guitars, and they will placed in a certain lateral position, at some point close behind the line of the speakers; the vocals will be in a completely different acoustic, and will be spaced back depending upon what echo was added - this part of the mix will seem the most disconnected from the rest of the sound elements, usually; the drum kit comes across as the most "natural" normally, and occupies yet another 'virtual space' behind the speakers, at some point laterally.

Tracks by the boogie rock group Status Quo are good examples for demonstrating how this type of soundstaging and imaging can be manifested.

 

[NorthSky]

The article i weakly remembered was:
Nichiguchi et al.; Perceptual discrimination of very high frequency components in wide frequency range musical sound, Applied Acoustics 70 (2009) 921–934

From here? http://www.sciencedirect.com/science/article/pii/S0003682X09000036

▪ http://www.enjoythemusic.com/magazi...olution_Audio_Perceptual_Evaluation/page4.htm

▪ http://www.academia.edu/31085315/A_Meta-Analysis_of_High_Resolution_Audio_Perceptual_Evaluation

▪ https://www.revolvy.com/main/index.php?s=Hypersonic effect

 

[Analog Scott]

Interesting that you find that this is a "struggle" - I very specifically look for this aural behaviour; that everything in the soundfield exists in a virtual space; any awareness that the speakers are the source of the sound means that the system is below par. If a curtain was placed between you and real performers on a stage, say, would you be troubled because you couldn't actually sight the particular instrument being played? In my case, no; and listening to playback "without anything there" is equivalent.

Yes you would. IMO one of the biggest misconceptions in the world of audio is about the actual imaging of live acoustic music vs. stereo/multichannel playback. The imaging at a live concert is as good as it is because we can see the musicians. Visual cues dominate aural cues when it comes to identifying location. A great example of that is when we watch any movie. Our brains tell us the sound of the actors' voices is coming directly from their mouths but in fact it is always coming from a speaker that is not positioned exactly behind the image of the actor. IMO stereo imaging is far more specific than it is at actual concerts and most listeners think the opposite is true. With stereo we don't have the benefit of visual cues and the more precise imaging of the stereo playback creates an illusion of greater fidelity to the live performance to conpensate for the lack of visual cues that we always have at live concerts. Also a very very big issue that is often overlooked is that we quite literally listen to live music and stereo playback differently. With live music we have the luxury of moving our heads to better focus our hearing on something specific in the performance. Do that with stereo and the imaging falls apart because stereo imaging is an aural illusion that relies on the position of the listeners head. Much like watching 3D movies. You can't tilt your head to the side or the 3D stops working. In real life you can look in any direction you want at any angle you want and you still see 3D. The ability to move our heads also allows us to better locate positioning by ear alone. But when we do this we are completely changing what is coming to our ears.

Rest assured, if you were to attend a classical concert of an orhcestra in a nice reverberant hall and were taken in blindfolded and sat mid hall and could not move your head from facing directly forward as if you were listening to a stereo you would have a great deal of trouble accurately locating the positions of the soloists by ear alone. Not so much with most stereo recordings of orchestras recorded in revererant halls.

 

[fas42]

Yes you would. IMO one of the biggest misconceptions in the world of audio is about the actual imaging of live acoustic music vs. stereo/multichannel playback. The imaging at a live concert is as good as it is because we can see the musicians. Visual cues dominate aural cues when it comes to identifying location. A great example of that is when we watch any movie. Our brains tell us the sound of the actors' voices is coming directly from their mouths but in fact it is always coming from a speaker that is not positioned exactly behind the image of the actor. IMO stereo imaging is far more specific than it is at actual concerts and most listeners think the opposite is true. With stereo we don't have the benefit of visual cues and the more precise imaging of the stereo playback creates an illusion of greater fidelity to the live performance to conpensate for the lack of visual cues that we always have at live concerts. Also a very very big issue that is often overlooked is that we quite literally listen to live music and stereo playback differently. With live music we have the luxury of moving our heads to better focus our hearing on something specific in the performance. Do that with stereo and the imaging falls apart because stereo imaging is an aural illusion that relies on the position of the listeners head. Much like watching 3D movies. You can't tilt your head to the side or the 3D stops working. In real life you can look in any direction you want at any angle you want and you still see 3D. The ability to move our heads also allows us to better locate positioning by ear alone. But when we do this we are completely changing what is coming to our ears.

Rest assured, if you were to attend a classical concert of an orhcestra in a nice reverberant hall and were taken in blindfolded and sat mid hall and could not move your head from facing directly forward as if you were listening to a stereo you would have a great deal of trouble accurately locating the positions of the soloists by ear alone. Not so much with most stereo recordings of orchestras recorded in revererant halls.

Interesting points you've raised ...

Yes, at a live orchestral concert the imaging is not precise - far from it in fact. Here the "vibe" and energy, impact of the performance are the gains - and I'm interested in generating precisely that type of dynamic with audio playback, for those sorts of recordings.

With conventional stero playback the imaging may "fall apart" when the listener's head moves - but that's precisely what changes when the higher quality replay I talk of is achieved. The extra level of integrity of the presentation of the fine detail in the recording allows the mind to descramble the cues, and the illusion is maintained. Irrespective of where my head moves. The 3D nature of the aural scene is always consistent, no matter how I decide to listen to the sound - and the reason appears to be because the ear/brain has decided, completely unconsciously, to 'accept' the mirage - I can't consciously decide to reject the soundfield structure; it's completely robust.

Recordings contain, completely unintentionally by the sound engineers, a layering of virtual sound spaces of the various sound components that were mixed in. Unless they did the recording in a purist way - a simple stereo pair of mics with every musician in the room. The layering in the aural planes works just like the visual, say, in Photoshop creation of montages - you can "see" a particular layer by casually choosing to focus on it - call it, the cocktail party effect - you only hear the person you want to, by mentally switching your area of interest.

 

[tomelex]

Yes you would. IMO one of the biggest misconceptions in the world of audio is about the actual imaging of live acoustic music vs. stereo/multichannel playback. The imaging at a live concert is as good as it is because we can see the musicians. Visual cues dominate aural cues when it comes to identifying location. A great example of that is when we watch any movie. Our brains tell us the sound of the actors' voices is coming directly from their mouths but in fact it is always coming from a speaker that is not positioned exactly behind the image of the actor. IMO stereo imaging is far more specific than it is at actual concerts and most listeners think the opposite is true. With stereo we don't have the benefit of visual cues and the more precise imaging of the stereo playback creates an illusion of greater fidelity to the live performance to conpensate for the lack of visual cues that we always have at live concerts. Also a very very big issue that is often overlooked is that we quite literally listen to live music and stereo playback differently. With live music we have the luxury of moving our heads to better focus our hearing on something specific in the performance. Do that with stereo and the imaging falls apart because stereo imaging is an aural illusion that relies on the position of the listeners head. Much like watching 3D movies. You can't tilt your head to the side or the 3D stops working. In real life you can look in any direction you want at any angle you want and you still see 3D. The ability to move our heads also allows us to better locate positioning by ear alone. But when we do this we are completely changing what is coming to our ears.

Rest assured, if you were to attend a classical concert of an orhcestra in a nice reverberant hall and were taken in blindfolded and sat mid hall and could not move your head from facing directly forward as if you were listening to a stereo you would have a great deal of trouble accurately locating the positions of the soloists by ear alone. Not so much with most stereo recordings of orchestras recorded in revererant halls.

Yep, good points. Folks forget that stereo was an invention for movie theaters to allow sounds to be spread across the screen to add a degree of realism to the scene. Our eyes do overrule the ears processing of sound as you say. And while the methods to create left to right and a bit of depth to recording is well known, some folks process these illusionary attempts differently that's for sure. The two point source speakers launch the timing differences and the ears interpret them, (and are fooled by them to a more or lesser degree) depending on folks ear/brain interfaces. For me, stereo is not or never has been very convincing but apparently for others it is like being there. Some folks claim it is "real" to them, even though theoretically the small amount of information available from those two sources of sound is in no way enough for you to think you are "there". Binaural recording on the other hand, that is pretty darn convincing if done well. Stereo attempt to bring the event to you, binaural attempts to bring you to the event. Big difference. Too bad stereo came along, we had binaural to start with, just like we veered into tubes when we actually had solid state to start with. Oh well.

 

[fas42]

Tom just reminded me to mention that true mono recordings can also throw up a big sound space - identical information from left and right channels still gives layering, and the ability to 'estimate' the distance back. Which indicates, probably even more strongly, that the mind is reacting to the acoustic cues to provide the subjective experience, rather than differences between left and right. I find that there is no major differences between stereo and mono playback when well done - quite often I've had to check the CD case, to confirm that the recording is in fact mono.

 

[tomelex]

Loudness levels of instruments in mono can give an interpretation of depth, as in stereo.

 

[Cosmik]

I am one of those people for whom stereo sounds 'real'. I think that stereo over speakers does give you a sense of 'being there', and gives you some latitude to move your head around and still maintain the illusion. But there may be some conditions attached to that: the speakers may have to have certain neutral radiation characteristics, and the room to lie within a certain range of acoustic characteristics. Speakers that beam or focus the sound strongly in some way; speakers that have markedly different dispersion vertically from horizontally and things like that? I don't have enough experience of them to know whether they maintain the illusion - even though they provide (at face value, simplistically) superior measurements at the listening position.

I am convinced of one thing: imaging may be exaggerated compared to a real live music situation (not if you're the conductor?), but it is the natural fall-out from a system that is doing what it is supposed to. If a system's imaging is weak or weird it means it is not working properly in some way; it means your hearing is having to work harder to understand what it is picking up.