Audio Blind Testing

[pkane2001]

1 hour ago, semente said:

 

The pair of mics is capturing what someone in the audience would have listened but the speakers (in a room) will not be able to recreate the soundfield of original event as it was listened.

 

See these posts:

 

https://www.computeraudiophile.com/forums/topic/38125-audio-blind-testing/?do=findComment&comment=766034

 

https://www.computeraudiophile.com/forums/topic/38125-audio-blind-testing/?do=findComment&comment=766027

 

https://www.computeraudiophile.com/forums/topic/38125-audio-blind-testing/?do=findComment&comment=766030

 

 

 

Problems with speaker systems I understand (and these are not related to what's captured by mics, btw). They obviously cannot reproduce the same waveforms, phases, etc. at the listening position due to many factors, from room interactions, to cabinet resonances, to sound wave interference, reflections, etc.

 

But I don't care about the sound field in the room if I'm using headphones, do I? Can the two mics capture enough information to be reproduced through a set of hi-fi headphones? These are much less affected by room interactions, reflections, interference, cabinet resonances, etc. Do the mics capture enough information to enable reproduction of the illusion of the original venue when using headphones and if not, what's missing?

 

[semente]

30 minutes ago, pkane2001 said:

 

Problems with speaker systems I understand (and these are not related to what's captured by mics, btw). They obviously cannot reproduce the same waveforms, phases, etc. at the listening position due to many factors, from room interactions, to cabinet resonances, to sound wave interference, reflections, etc.

 

But I don't care about the sound field in the room if I'm using headphones, do I? Can the two mics capture enough information to be reproduced through a set of hi-fi headphones? These are much less affected by room interactions, reflections, interference, cabinet resonances, etc. Do the mics capture enough information to enable reproduction of the illusion of the original venue when using headphones and if not, what's missing?

 

 

I don't listen with headphones but I suppose that a recording using a binaural head would in theory produce a more credible illusion although I haven't tried.

Regular recordings feel to me as if the sound is being generated inside my head instead of surrounding me.

(I'm referring to classical music recordings)

[pkane2001]

6 minutes ago, semente said:

 

I don't listen with headphones but I suppose that a recording using a binaural head would in theory produce a more credible illusion although I haven't tried.

Regular recordings feel to me as if the sound is being generated inside my head instead of surrounding me.

(I'm referring to classical music recordings)

 

Yes, cross-feed helps a bit there, but obviously, it's altering the originally recorded sound by introducing some cross-pollination between the two ears that wasn't in the original recording. I'm curious if the problem with the 'in your head' sound is due to the mastering process that destroys the true phase relationship between left and right channels. I have some recordings that sound wonderfully spacious and 3D through the headphones, with a sense of depth without crossfeed, and then some that simply require crossfeed for me to be able to even listen to them. 

[fas42]

8 hours ago, semente said:

I'll reinstate that current 3D/ambio audio is still nothing like "real" sound.

 

It perceptually increases a sense of space, but it's a mere effect that creates other problems, affecting aspects of music reproduction which I find far more relevant: music is aural (sound) not visual (space).

 

Everyone here is making it far too hard - it's completely ridiculous trying to capture a perfect version of the original soundfield, then regenerating all of that in some other location - it also mean that all recordings to date are useless ...

 

Letting our internal processing plant do the work is the answer - it throws up an illusion which is just as satisfying as the "real thing", and often will be far superior, because the soundfield was picked up from better locations. Only one step has to be taken to get there - improving the end to end processing of the captured sound as it's played back, so that all of the information is damaged as little as possible.

 

Visited an audio friend yesterday who's running with my ideas - and he's steadily working through with 'mangling' everything to eliminate those bottlenecks. Just using an ordinary, palm sized music player for source and DAC, and tiny Tannoy bookshelfs the quality is mighty close to optimum at times, and threw up a massive, enveloping soundfield on some modern, very dense, layered pop music. Overall, far superior to 99% of playback I've heard from other systems over the years.

[Ralf11]

I can make it harder - just capture any version of the original soundfield, and do the corrections by altering the processing in the brain, not by the listener (who will have to work hard to do it internally), but by the experimenter.

[semente]

21 minutes ago, Ralf11 said:

I can make it harder - just capture any version of the original soundfield, and do the corrections by altering the processing in the brain, not by the listener (who will have to work hard to do it internally), but by the experimenter.

 

Are you referring to a pre-frontal lobotomy?

[Ralf11]

only for an Australian

 

 

[STC]

7 hours ago, bhobba said:

There are huge egos involved in it - the winner will reap the rewards - the losers will greatly suffer.  There is nothing really in it for anyone......

 

I agree. People are not interested to experiment for fear of being proven wrong.

[STC]

3 hours ago, fas42 said:

Everyone here is making it far too hard - it's completely ridiculous trying to capture a perfect version of the original soundfield, then regenerating all of that in some other location - it also mean that all recordings to date are useless ...

 

I am also having dillema about what is the objective. The truth is no two concert hall sound alike. Physically it is near impossible. What we should be striving for is the ability to replay them and to sound like a concert hall sound. 

 

3 hours ago, fas42 said:

Letting our internal processing plant do the work is the answer - it throws up an illusion which is just as satisfying as the "real thing", and often will be far superior, because the soundfield was picked up from better locations. Only one step has to be taken to get there - improving the end to end processing of the captured sound as it's played back, so that all of the information is damaged as little as possible.

 

 

If only I could train my internal processing to reach that level, I could settle for the sound from transistor and let my internal processing to fill in the missing link.

[fas42]

5 hours ago, pkane2001 said:

 

Yes, cross-feed helps a bit there, but obviously, it's altering the originally recorded sound by introducing some cross-pollination between the two ears that wasn't in the original recording. I'm curious if the problem with the 'in your head' sound is due to the mastering process that destroys the true phase relationship between left and right channels. I have some recordings that sound wonderfully spacious and 3D through the headphones, with a sense of depth without crossfeed, and then some that simply require crossfeed for me to be able to even listen to them. 

 

The "in your head" sound with headphones is because the acoustic clues are too indistinct - one's ear/brain can't interpret their true meaning - so the sound "pushes inward". The equivalent with speakers is that the sound is trapped in the cabinets, the imaging doesn't lift out and lay beyond them.

 

The solution in both cases is to improve the reproduction chain; the better resolved acoustic data now is clearly understood by the brain - and the soundfield then stretches beyond the transducers.

[fas42]

12 minutes ago, STC said:

If only I could train my internal processing to reach that level, I could settle for the sound from transistor and let my internal processing to fill in the missing link.

 

... actually, everyone does this all the time, while listening to the sounds of the world around them - we are all black belt experts in this regard. It's how we can hear, say, a brass band from the other side of town, and instantly recognise its authenticity ...

 

So, our insides are in good shape - the mechanism pretending to be musical instruments is the weak link, it sends out far too many clues that it's not the real deal - and our brains say, "Gotcha!! ... you're not foolin' me!"

[esldude]

33 minutes ago, fas42 said:

 

The "in your head" sound with headphones is because the acoustic clues are too indistinct - one's ear/brain can't interpret their true meaning - so the sound "pushes inward". The equivalent with speakers is that the sound is trapped in the cabinets, the imaging doesn't lift out and lay beyond them.

 

The solution in both cases is to improve the reproduction chain; the better resolved acoustic data now is clearly understood by the brain - and the soundfield then stretches beyond the transducers.

If you listen in an anechoic chamber without moving your head, using two speakers, the sound is in your head like with headphones. Has nothing to do with deficiencies in the chain.

[STC]

7 minutes ago, esldude said:

If you listen in an anechoic chamber without moving your head, using two speakers, the sound is in your head like with headphones. Has nothing to do with deficiencies in the chain.

 

This cannot be correct unless you are playing a mono sound with two speakers where there is no ITD and ILD.

[STC]

Just to refresh the major difference between speakers and headphones.

 

Quote

With speakers at 60 degrees, the speakers may have perfectly flat response, but when two such speakers are operating the response at the ear is anything but flat.  Central low bass is doubled in level and there are a series of sharp peaks and dips starting at about 1500 Hz.  You may not care about stage width or localization, but your brain does.  So when a microphone records say a 700 microsecond difference for a left side instrument, what you get with 60 degree speakers is first an erroneous  220 microsecond difference cue instead of 700 followed by the correct 700 microsecond delay which is now regarded by the brain as just an early reflection, followed by a bogus 220 microsecond signal that is contradicting the recorded time delay signal.  There are similar problems with the recorded level differences delivered at the two ears when both ears hear both speakers at angle.

 

 

Quote

Now for the pinna and headset problems.  The outer ears are direction finders for frequencies above about 1000 Hz.  With speakers all the pinna directional cues are at the same plus and minus 30 degrees no matter where the stereo image seems to be based on the lower frequencies.  This inconsistency sensitizes the brain so that it knows the field is not real and worse makes it overly sensitive to minor things like resolution, LP ticks and pops, harmonic distortion, etc.  Now with headphones, the pinna function is decimated no matter what kind of phones are used.  The frequency response is now flat and the localization cues are correct, but the inconsistency in sound field is still there, so the brain normally internalizes the field so you have a wide stage between the ears inside your head.  If you hum or hiss to yourself while you bring a finger into each outer ear, you can hear this effect.  You can use things like the Smyth Realizer which uses your speaker set up as a model to thereafter avoid internalization with earphones.

 

And about soundfield.

Quote

Now comes the rear soundfield problem.  The brain expects that sounds in front will produce a set of reflections from the sides and rear that vary with the source position in front.  But with speakers the reflection pattern in the room is always the same directionally no matter where the phantom image is in front.  Again this results in a stage without envelopment and a brain that know something is wrong.  With earphones there are only the recorded reflections and they all come from the same direction so again realism is not possible this simple way.  So the use of rear speakers has nothing to do with 5.1 surround.  They are a necessary part of distortion free 2.0 LP, CD, SACD, Download, reproduction.

 

Read more here. 

 

[fas42]

1 hour ago, STC said:

Just to refresh the major difference between speakers and headphones.

 

 

 

 

And about soundfield.

 

Read more here. 

 

 

Yep. All the standard stuff about how it's supposed to work - and I would to get all of that happening if I chose to deliberately downgrade the playback - but for some strange reason I'm not interested in doing that ...

 

Just to refresh where I'm coming from, 30 years ago I was getting the typical stereo playback that everyone normally does, with zero expectations or desire for more than that ... then, one day, I happened to get enough right, at one moment - and, the illusion switched on ... Bang!! Boolean logic and all that ...

 

Everything since has been a natural extension and evolution from that point - people can jump up and done in frustration, froth at the mouth, etc, etc; claiming such isn't possible - but that moment happened for me, and it can never go away - especially as I get confirmation of this behaviour, over and over again.

[esldude]

1 hour ago, STC said:

 

This cannot be correct unless you are playing a mono sound with two speakers where there is no ITD and ILD.

I think it was in Toole's book or referenced someone else's research.  If the listener is equidistant from speakers in an anechoic chamber and their head is held immobile, they hear imaging mostly inside their head much like with headphones.   If their head is not held immobile apparently very small movements of our head are important.  As the imaging then moves outside their head.  That is one of the advantages of the Smyth Realizer.  Head movement is brought into the sound you hear thru the phones. 

 

I am taking the word of others on this not having the chance to try it out in an anechoic chamber for myself. 

[STC]

4 minutes ago, fas42 said:

Yep. All the standard stuff about how it's supposed to work - and I would to get all of that happening if I chose to deliberately downgrade the playback - but for some strange reason I'm not interested in doing that ...

 

I am not trying to convince you. I was just pointing out the documented reasons for the perception of realism with speakers and headphones. What you can hear is entirely within your domain of realism.

[esldude]

12 hours ago, pkane2001 said:

 

Dennis, so are you saying that for a true 'live' sense and feel of reproduced audio, the listener must be able to move through the sound field and sample this field at slightly different points to get the correct illusion? Similar to holography that looks  2D if you just look at it from a single point, but becomes 3D as you start moving your head around it?

 

If that's the case, would I also get only 2D, not truly 'live' sound impression at a concert hall if I sit with my head perfectly still?

Some info I have read indicates head movement is important for the illusion of space.  So having your head held still or wearing headphones so the soundfield never moves, impacts your ability to hear space and imaging at a distance.  

[STC]

6 minutes ago, esldude said:

I think it was in Toole's book or referenced someone else's research.  If the listener is equidistant from speakers in an anechoic chamber and their head is held immobile, they hear imaging mostly inside their head much like with headphones.   If their head is not held immobile apparently very small movements of our head are important.  As the imaging then moves outside their head.  That is one of the advantages of the Smyth Realizer.  Head movement is brought into the sound you hear thru the phones. 

 

I am taking the word of others on this not having the chance to try it out in an anechoic chamber for myself. 

 

To localize sound you need ILD and ITD. Your example applies to mono signal because when the head is held steady both signals arrive your ears at same level and at same time thus making localization impossible and you can  place the sound anywhere along the line. However, your visual clue also would play a role and it is possible for you to imagine a location,IMO.

[fas42]

Just now, STC said:

 

I am not trying to convince you. I was just pointing out the documented reasons for the perception of realism with speakers and headphones. What you can hear is entirely within your domain of realism.

 

Of course. However, the quality of playback is such that others perceive its "specialness" - whether they hear the precise same illusion as I do. Referring to other thread, where Ralph says,

 

Quote

Right On! Almost.  That is called "They Are Here" reproduction and in theory is only possible for say a single instrument like a guitar preferably coming from a single speaker in a good room.  For a large ensemble like an orchestra this is not possible.  For most music and movies you need "You Are There" home reproduction and this is something that requires a bit more technical savvy than you need for just playing 2.0 files of a solo vocalist and a guitar.  Google Ambiophonics, Wavefield Synthesis, Ambisonics, BACCH, XTC, etc

 

what is achieved is the "You Are There" illusion, for all recordings. Part of the illusion is that it becomes impossible to locate the speakers, no matter where you move - if you can't achieve this, then the rest of the package won't happen, or it will be severely limited, to a small subset of recordings.

[esldude]

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0083068

 

https://www.frontiersin.org/articles/10.3389/fnins.2014.00210/full

 

The first link is more directly toward internalization vs head movement. 

 

 

[STC]

1 minute ago, fas42 said:

what is achieved is the "You Are There" illusion, for all recordings. Part of the illusion is that it becomes impossible to locate the speakers, no matter where you move - if you can't achieve this, then the rest of the package won't happen, or it will be severely limited, to a small subset of recordings.

 

I have achieved this and regularly have visitors putting their ears on the bass traps looking liking flat electrostatics speakers to hear whther sound is coming from there. Even with stereo, I did a pretty good job making the Harbeth disappear especially with Roger Water's ATD. There is nothing you say here is unique. Many audiophiles capable of making the speakers disappear and create floating phantom image with boombox or stereo. 

[fas42]

16 minutes ago, STC said:

 

I have achieved this and regularly have visitors putting their ears on the bass traps looking liking flat electrostatics speakers to hear whther sound is coming from there. Even with stereo, I did a pretty good job making the Harbeth disappear especially with Roger Water's ATD. There is nothing you say here is unique. Many audiophiles capable of making the speakers disappear and create floating phantom image with boombox or stereo. 

 

Of course there is nothing unique about it, it's the natural expression of any stereo system working well enough. But we are not talking about special recordings, we are talking all recordings, including mono releases; putting your ears next to the drivers of conventional box speakers, and not hearing them; and that the "floating phantom image" is the aural universe you are immersed in. Doing this is not a gimmick, a party trick to amuse the guests - it's "how it works", if the system is reproducing with competence.

[fas42]

STC, I note that you say you can always hear "the hole in the middle" with a conventional stero setup - has that been the situation every time; if you were to redo your setup with main speakers in the conventional positions, switching off all other speakers, i.e., pure stereo, with everything else unchanged would that still be the case, now?

[esldude]

Here is another on head movement and internalization of sound. 

 

http://epubs.surrey.ac.uk/712909/1/2011 Building Acoustics.pdf