Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: "Real Stereo" (Read 11476 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

"Real Stereo"

Hi. I recently came across this. The article claims that two-channel stereo, if implemented properly can capture spatial information. Is this article technically correct? If that is so, aren't we being conned into buying multi-channel audio systems? Thanks.


"Real Stereo"

Reply #2
_very_ interesting. Thanks.

- Lyx
I am arrogant and I can afford it because I deliver.

"Real Stereo"

Reply #3
technically I think it is possible, because we only have 2 ears.....anyone has 6 ??? 

2 channels (headphones) should be able to emulate 6 channel sound?

I doubt that 2 loud speaker is good enough to produce "real" stereo....you can place loud speakers in front of you, or the back of you! the effect is different!

brain is the one which inteprete sound, we just need to trick the brain to belief!

"Real Stereo"

Reply #4
Quote
technically I think it is possible, because we only have 2 ears.....anyone has 6 ??? 

brain is the one which inteprete sound, we just need to trick the brain to belief!
[a href="index.php?act=findpost&pid=276872"][{POST_SNAPBACK}][/a]



No, that articles relates to ambiphonic surround, etc which is said to theoretically have better height and sound localization then traditional surround sound setup.

Quote
I doubt that 2 loud speaker is good enough to produce "real" stereo....you can place loud speakers in front of you, or the back of you! the effect is different!


ha. No, it's more complicated then that.

Quote
2 channels (headphones) should be able to emulate 6 channel sound?


It might be possible via HRTF's. A lot of research is being poured into the field of 3D audio using HRTF. Everyone HRTF are different though and they would need to be adjusted for different people. I haven't experimented with them yet the results I hear from people are inconclusive though.
budding I.T professional

"Real Stereo"

Reply #5
The outer ear plays an important role in changing the phase of sounds a a function of the direction it comes from.

source
Quote
The reason is that sound that comes from behind us sounds different than the sound the comes from our front.  The outer ear creates an obstacle so that sounds coming from the back sound different than sound coming from the front.  Having two ears, our brain is able to localize where sound comes from.

Without going into detail how the process works, the brain is able to tell where sounds emanate from by comparing sound that is out of phase with sound that is in phase.  Basically, the brain is able to process input from two ears based on whether the sound arrives in phase, out of phase and if it comes from the front and rear and how much delay is there from the sound reaching one ear canal versus another.

The idea behind stereo is that if we can create true three dimensional sound with only two ears, two speakers should be able to do the same thing.  There are all sorts of reasons why this isn't possible (although it is possible to do this using headphones and something called biphonic recording).

"Real Stereo"

Reply #6
Quote
The outer ear plays an important role in changing the phase of sounds a a function of the direction it comes from.

source
Quote
The reason is that sound that comes from behind us sounds different than the sound the comes from our front.  The outer ear creates an obstacle so that sounds coming from the back sound different than sound coming from the front.  Having two ears, our brain is able to localize where sound comes from.

Without going into detail how the process works, the brain is able to tell where sounds emanate from by comparing sound that is out of phase with sound that is in phase.  Basically, the brain is able to process input from two ears based on whether the sound arrives in phase, out of phase and if it comes from the front and rear and how much delay is there from the sound reaching one ear canal versus another.

The idea behind stereo is that if we can create true three dimensional sound with only two ears, two speakers should be able to do the same thing.  There are all sorts of reasons why this isn't possible (although it is possible to do this using headphones and something called biphonic recording).

[a href="index.php?act=findpost&pid=277083"][{POST_SNAPBACK}][/a]



Quote
The outer ear plays an important role in changing the phase of sounds a a function of the direction it comes from.


Right Pinna deflection, comb filtering, etc hence (Head Related Transfer Function). This topic comes up a lot in these forums hence "spatial filters".  You need binaural recordings though and set of HRTF impulses reponses with headphones ;-D and yes changes in phase by the ear are interpereted as changes in azimuth. The original poster was ferrering to open-air speaker setup though I was just trying to answer that other question to the best of my ability. What he was referring to was in reguard to stereo crosstalk among speaker's we got sidetracked a little here.
budding I.T professional

"Real Stereo"

Reply #7
All users who are inclined to accept the premise of the original should first read the old article by Fletcher and Snow on stereophony. It's chapter 12 or 13 of the reprinted Fletcher papers, I think I recall.
-----
J. D. (jj) Johnston

"Real Stereo"

Reply #8
As far as i look in to this
it seems to me that its true coz when the sound u hear is stereo or 5.1 or 7.1
the output from all the channels will not be the same.
If u take stereo the left and right o/p's are not same , its the essense of stereo
coding.
u have MS stereo coding where at high bit rates the 2 channels of audio will be coded into  1 channel coz our ear's cant differentiate the sounds at high freq's.
so effectively we r takin the advantage of the human auditory system.

"Real Stereo"

Reply #9
Thanks for pointing out the Ambiophonics article, HotshotGG. Brings a certain irony to using a crossfeed plugin with headphones.

"Real Stereo"

Reply #10
Quote
technically I think it is possible, because we only have 2 ears.....anyone has 6 ??? 

2 channels (headphones) should be able to emulate 6 channel sound?

I doubt that 2 loud speaker is good enough to produce "real" stereo....you can place loud speakers in front of you, or the back of you! the effect is different!

brain is the one which inteprete sound, we just need to trick the brain to belief!
[a href="index.php?act=findpost&pid=276872"][{POST_SNAPBACK}][/a]


You have two eye balls and yet, you can see in three dimensions.

Radetz.

"Real Stereo"

Reply #11
Quote
Quote
technically I think it is possible, because we only have 2 ears.....anyone has 6 ??? 

2 channels (headphones) should be able to emulate 6 channel sound?

I doubt that 2 loud speaker is good enough to produce "real" stereo....you can place loud speakers in front of you, or the back of you! the effect is different!

brain is the one which inteprete sound, we just need to trick the brain to belief!
[a href="index.php?act=findpost&pid=276872"][{POST_SNAPBACK}][/a]


You have two eye balls and yet, you can see in three dimensions.

Radetz.
[a href="index.php?act=findpost&pid=281313"][{POST_SNAPBACK}][/a]


It's an interpolation thing (3d vision from two eyes) isn't it? Like I can 'see' things in my peripheral vision but really a lot of it is my brain telling me these things *should* be there. The question is, does this trickery matter?

"Real Stereo"

Reply #12
It's not really the same as your eyes. Your eyes can find 3d positioning fairly easy by parallax. That's the phenomena where if you hold out your arm and stick up a finger, when you close one eye and then switch it will "jump" relative to the background. That is actually really easy to compute, everything obeys really simple rules of highschool geometry.

To sense 3d sound patterns, your brain interprets two big things: delay / phase and frequency changes. Delay is time between hearing in one ear and then the other. If there is a lot of delay, the sound is obviously coming from the extreme right or left, according to which ear heard the sound first. If there is no delay, it must be coming from the direct front or back-- but which? So then the brain looks at frequency. The shape of the ears affects the frequency of the sound, amplifying some and muting others. Since the ears face in one direction, the changes will be different depending on where the sound is coming from.

The second big thing the brain does is compare the sounds to the set of sounds in our memory to compare and contrast them. That can help in the 3d analysis by providing a reference to a sound or sound pattern that is already known. Sound is much more dependent on learned information than vision is. For example, we all can hear and unconsciously interpret the dopler effect, though I'm pretty sure that speeding cars were not part of human evolution.

"Real Stereo"

Reply #13
Quote
technically I think it is possible, because we only have 2 ears.....anyone has 6 ???

Your 2 ears can move. According to a talk I've heard a couple of times, your brain is able to use the fact your ears move to "sample" (I think that's probably a very rough version of the word) the soundfield around you in more than one place. This would imply that while we have 2 ears, they sample the soundfield at more than 2 positions, and therefore you can't say that two speakers are enough.
Quote
2 channels (headphones) should be able to emulate 6 channel sound?

I would imagine that if you captured your own head related functions, you could make such a device. I'm not sure it would work if you used somebody else's head related functions, though.  All heads are not the same, neither are all pinnae.
Quote
I doubt that 2 loud speaker is good enough to produce "real" stereo....you can place loud speakers in front of you, or the back of you! the effect is different!

And then there's the issue of your ears comparing the playback room reverb with the direct sound from the speakers, and being able to extract front/back information from that, even if the "back" information is properly frequency-shaped.
Quote
brain is the one which inteprete sound, we just need to trick the brain to belief!
[a href="index.php?act=findpost&pid=276872"][{POST_SNAPBACK}][/a]


But the brain's ability to interpret sound is very hard to describe in any mathematical sense, or so I'm told.
-----
J. D. (jj) Johnston

"Real Stereo"

Reply #14
given:
Quote
This would imply that while we have 2 ears, they sample the soundfield at more than 2 positions, and therefore you can't say that two speakers are enough.


and:
Quote
All heads are not the same, neither are all pinnae.


i'm not sure what to think.  generally, senses are quite adaptive.  give someone glasses that invert the image, and in a short time they'll be seeing upside-down.  take the glasses off and suddenly they're upside-down again.

so i got this idea just a second ago:  what about those headphones that sense your head's position, combined with a generic HRTF that your brain could get used to in a relatively short time?  this way as you move your head around, your ears get used to their "new shape", rather like how when we wear a helmet or wide-brimmed hat, we're still quite able to localise sounds.

dunno, just thinking out loud, but it seems to me we can adapt to a HRTF that doesn't match our physical features if there's enough information for our brain to work on.

"Real Stereo"

Reply #15
A question:
Can  stereo image be achieved from just ONE sound source?
To explain:
I recently had a problem with my speaker 's cable. So a night i was listening to music, i thought to set foobar to downmix channels to mono, as the sounds from the right channel were completely lost, as i had no right speaker. To my suprise, when i donwmixed to mono, the music was sounded really,really bad and  plain, real mono in other words. Well, after the first shock, i noticed that i had enabled  on option of my lame philips fw-m55(thats were the output of the foobar-pc goes) called "incredible surround". In some way this thing mixes the two channels and  the music sounds stereo.

So can be that true, or i am speaking for different thing?

"Real Stereo"

Reply #16
Quote
Well, after the first shock, i noticed that i had enabled on option of my lame philips fw-m55(thats were the output of the foobar-pc goes) called "incredible surround". In some way this thing mixes the two channels and the music sounds stereo.

So can be that true, or i am speaking for another thing?


I don't really understand what you are getting at here.  It might just use some fancy matrixing techniques to derive a stereo signal from a mono source (much like plug-in in FB2K.  A lot of these features are just "virtual surround" techniques that just do basic DSP things. They aren't really on par with HRTF, etc. "Real Stereo" in this discussion has more to do with crosstalk, etc. 
budding I.T professional

"Real Stereo"

Reply #17
no,no,no. There is no mono source to be made  stereo. I know there are such algorithms out there...
I m speaking of giving stereo(two channels) and listening stereo from just one speaker.

"Real Stereo"

Reply #18
Quote
dunno, just thinking out loud, but it seems to me we can adapt to a HRTF that doesn't match our physical features if there's enough information for our brain to work on.
[a href="index.php?act=findpost&pid=288009"][{POST_SNAPBACK}][/a]


The people from Lake DSP (who did a lot of work trying to commercialise HRTF rendered audio) used to say that the quality of the HRTFs was irrelevant - you could measure the HRTFs of a sphere (i.e. no ears, no hair etc) but what mattered without head tracking was accurate simulation of the reverb of the room the source was supposed to be in. Without that, even with accurate HRTFs, it would just sound like the source was in your head.

However, the _most_ important thing was accurate head tracking - that made the whole listening experience seem very real.

btw, there's a very old paper (1968 IIRC) paper by someone called Batteu (sp?) where he shows that people can get used to other people's HRTFs - he didn't use any electronics - just stuck moulds of other ears on the side of people's heads - people learned to localise sound with them, showing that the auditory system was adaptive. However, it's not clear if people learned to accept the sound they heard through other outer ears as "natural" or not.

Cheers,
David.

"Real Stereo"

Reply #19
Quote
Hi. I recently came across this. The article claims that two-channel stereo, if implemented properly can capture spatial information. Is this article technically correct?[{POST_SNAPBACK}][/a]


No, it's complete junk. It's obvious they know so little of Blumlein's work that they don't know that he pointed out the limitations of "Blumlein stereo" (by which they probably mean crossed pair or MS mic recording) in his own patents! He discussed how manipulating the phase of the signals could push the perceived source slightly outside the width of the two speakers, but that the illusions soon broke down. You can't reliably put sounds outside the stereo speakers with this basic Blumlein stereo technique. This is the limitation, acknowledged by Blumlein, that this article apparently claims does not exist.

The theory says stereo should work for the front 90 degrees, excluding the effect of HRTFs (which tell your ears that the sounds are to the side, not in front!). In practice, we typically use 60 degree spaced speakers to ensure a good central image despite HRTFs and poor equipment.


However, you might get all kinds of sounds perceived to be outside the stereo speakers due to transaural-like processing. More likely (in the linked article) they've heard MS recordings - which, if you don't prevent it, will cause all sounds from behind the microphone to be captured just as well as sounds in front of the microphone, but out of phase. The "problem" is, when these sounds are reproduced over loudspeakers, your brain might interpret these out of phase signals as coming from somewhere outside the speakers, or it might not - it doesn't work properly, or reliably, or predictably, or accurately - it's just the case that lots of out-of-phase information will trick the brain (sometimes!) into hearing something outside the two stereo speakers. So some people believe that Blumlein stereo can do surround - it can't!


However, if you extend MS mic technique ("Blumlein stereo") into 3 dimensions, you get Ambisonics - which really is the surround-sound equivalent of Blumlein stereo, and is nothing like 5.1 (5.1 is not a great step forward for music IMO).

[a href="http://www.ambisonic.net/]http://www.ambisonic.net/[/url]

In ambisonics, you capture the sound at one point in space using an omni directional microphone (the M of an MS mic, if you like, though it's called "W" in ambisonics), and then Front-Back (the 'X' component), Left-Right (the 'Y' component), and Up-Down (the 'Z' component). It's very neat, and a surprising number of audiophile recordings use this, though the results are decoded to stereo or 5.1 for release.

This is a good place to start reading about Ambisonics:
http://www.ambisonic.net/ambi_AM91.html

Cheers,
David.

"Real Stereo"

Reply #20
i imagine the 3d tracking is the tricky part (and expensive one)?

but this should be simple to do from a users perspective, one would just have to look at the direction of the screen and press 'this is my default position' button, i imagine that tracking the rotation only would do the trick since we can count on human adaptivity?
PANIC: CPU 1: Cache Error (unrecoverable - dcache data) Eframe = 0x90000000208cf3b8
NOTICE - cpu 0 didn't dump TLB, may be hung

"Real Stereo"

Reply #21
Jim Johnston (JJ) developed at his final stage at AT&T labs a multichannel soundfield reconstruction system that, according to reports, worked really well, and included lossy compression. However I don't know what happened with it. Its name is Perceptual Soundfield Reconstruction:

http://www.att.com/attlabs/docs/psr.pdf
http://www.research.att.com/history/psr.html
http://www.onhifi.com/features/20010615.htm
http://stereophile.com/news/11035/

"Real Stereo"

Reply #22
Been a long time, but I happened on this when looking for a network reference for PSR.

It's dead, basically. AT&T owns the patents, but is completely uninterested in pushing it.
-----
J. D. (jj) Johnston

"Real Stereo"

Reply #23
That's probably because it doesn't really work. Recording such a soundfield is cumbersome, and the sweet spot for reproduction is usually really small, which makes it only possible for one person in that room to have it sound right.

More speakers make the sweet spot bigger. It makes recording much less constrained. That's why surround is usually done with more than two speakers.
Music: sounds arranged such that they construct feelings.


 
SimplePortal 1.0.0 RC1 © 2008-2021