[edstrelow]

 
This Wikepedia article on localization of sounds gives a fairly decent discussion of how the ear does this. http://en.wikipedia.org/wiki/Sound_localization
 
Most scientific work has concentrated on time and loudness differences between the 2 ears as the basis of the perception of direction.
 
There are a couple of interesting points the author(s?) make such as that loudness differences don't apply at lower frequencies because the ear doesn't cause a shadow at lower frequencies. In effect you don't get a physical loudness difference between the ears, as frequencies drop, for any external sound source or speaker.  This is the reason why some people say that it doesn't matter where you place a subwoofer, your hearing can't localize it as well as higher frequency speakers.
 
However this is not the case with headphones where you can create a loudness difference between the 2 ears because each earphone feeds a separate ear. This would imply that you can have a better localization of bass with headphones than you can with speakers, or even in a live listening situation.
 
You will note no discussion of Head-related-transfer-functions (HRTF's) other than a brief mention that pinna structures could affect directional hearing. This probably is because there has been little experimental  evidence that this is an important cue compared to time and loudness differences between the ears which have been extensively studied in both psychophysics and neuroscience for a century or so. .  This will, like my discussion of  'phantom channels," also probably come as a surprise to some here who seem to assume that HTRF effects are the basis of stereo/ directional hearing.
 
 However,  anyone who has listened to an IEM, which does not involve the pinna at all, and thus creates no HTRF, will realize that a good stereo image can be obtained even just feeding signals straight down the ear canal.  So it is pretty hard to argue that HTRF's are especially important for localization.  But I don't think IEMS give much sense of externalization, sounding aggressively "in-the-head."  My guess is that they are more important for that purpose. 
 
Now most non-IEM  headphones produce a strange HTRF because the drivers sit at the side of the ears.  So at best headphones are telling the brain that the sounds are beside the head, certainly not in front.  This is the only real advantage I can see with speakers, except for a few phones such as the AKG K1000 and Stax Sigma, which place the drivers ahead of the ears and, I think, sound a bit more speaker-like, while preserving the advantages of headphones.
 
 
 

 

[arnaud]

Hello Ed,
 
Thank your for bringing additional explanations.
 
I think you're pushing it a little when suggesting HRTFs are overrated though . From my understanding of jgazal's posts:

• The brain can be pretty good at horizontal localization by working out level and time differences between ears which is the reason why stereo sound reproduction with speakers, or headphones, or iems doesn't do too bad a job.
• The brain is also using reverberation clues to estimate depth / distance so again, stereo recordings can handle that regardless of playback system.
• Where it gets complication is for elevated sources, sources in the plane of the ears, and sources behind you when using front speakers. In these cases, personalized HRTFs are mandatory along with some head-tracking system as the people unconsciously move the head to help localize such events.

 
Quote:

 
This Wikepedia article on localization of sounds gives a fairly decent discussion of how the ear does this. http://en.wikipedia.org/wiki/Sound_localization
 
Most scientific work has concentrated on time and loudness differences between the 2 ears as the basis of the perception of direction.
 
There are a couple of interesting points the author(s?) make such as that loudness differences don't apply at lower frequencies because the ear doesn't cause a shadow at lower frequencies. In effect you don't get a physical loudness difference between the ears, as frequencies drop, for any external sound source or speaker.  This is the reason why some people say that it doesn't matter where you place a subwoofer, your hearing can't localize it as well as higher frequency speakers.
 
However this is not the case with headphones where you can create a loudness difference between the 2 ears because each earphone feeds a separate ear. This would imply that you can have a better localization of bass with headphones than you can with speakers, or even in a live listening situation.
 
You will note no discussion of Head-related-transfer-functions (HRTF's) other than a brief mention that pinna structures could affect directional hearing. This probably is because there has been little experimental  evidence that this is an important cue compared to time and loudness differences between the ears which have been extensively studied in both psychophysics and neuroscience for a century or so. .  This will, like my discussion of  'phantom channels," also probably come as a surprise to some here who seem to assume that HTRF effects are the basis of stereo/ directional hearing.
 
 However,  anyone who has listened to an IEM, which does not involve the pinna at all, and thus creates no HTRF, will realize that a good stereo image can be obtained even just feeding signals straight down the ear canal.  So it is pretty hard to argue that HTRF's are especially important for localization.  But I don't think IEMS give much sense of externalization, sounding aggressively "in-the-head."  My guess is that they are more important for that purpose. 
 
Now most non-IEM  headphones produce a strange HTRF because the drivers sit at the side of the ears.  So at best headphones are telling the brain that the sounds are beside the head, certainly not in front.  This is the only real advantage I can see with speakers, except for a few phones such as the AKG K1000 and Stax Sigma, which place the drivers ahead of the ears and, I think, sound a bit more speaker-like, while preserving the advantages of headphones.
 
 
 

 

 

[jgazal]

Instead of sharing what I think/guess about this subject, I would suggest a personal experiment. 
 
You would buy a digital portable recoder such as Korg MR-2 (1 bit DSD) or Sony PCM-D50 (24 bit, 96 Khz linear PCM).
 
You would also buy binaural microphones such as:
 
a) Sound Professionals MS-TFB-2
 

 
b) Microphone Madness MM-BSM-8
 

 
Then call any friend who plays a saxophone and ask him/her to walk around you while he/she is playing and you recording. You can try several arrangements: microphones on your ears (HRTF), with a stereo T mount/stand (f.i. SP-GNA-10) and jecklin disc between both microphones, with different distances (interaural delays) etc.
 
Play it back with in-ear, earbuds, supra-aural, circumaural headphones etc. Play it also with speakers.
 
You could try also out of ear binaural such as: http://microphonemadness.com/products/mmcroakstylb.htm or http://www.soundprofessionals.com/cgi-bin/gold/item/SP-HMC-1.
 
If you use preamp you may need a battery pack. If your friend play the saxophone loud enough, use line-in level, but then a battery pack is mandatory (see http://www.soundprofessionals.com/cgi-bin/gold/item/SP-SPSB-10).
 
Going to an anechoic chamber to measure your own HRTF seems more difficult. Have a look at this one from NHK: http://www.nhk.or.jp/digital/en/pressrelease/090421/processor.pdf
 
Then report back. I am looking forward to hearing from you.

 

[NamelessPFG]

This topic is most likely what drew me to this place to begin with...except I'm not looking to recreate the sound field of a surround speaker system for music presentation.
 
I'm looking for the ultimate imaging in games. That means I'm dealing with environments that should not be pre-mixed for any arbitrary number of speakers to begin with (but tend to be like that anyway for newer games due to the adoption of APIs like XAudio2 while OpenAL is increasingly ignored). Instead, it should be more like having a mono speaker for every sound source within that virtual 3D space, each of those sound sources having specific X/Y/Z coordinates.
 
Vertical sound cues are definitely part of that. It helps to know if those gunshots you just heard are coming from the floor above or below, and I find it a bit more immersive if I end up in a dogfight with another aircraft, make a roughly head-on pass, and actually hear the opponent pass by above or below...or am hovering in an attack helicopter knowing that an enemy chopper is somewhere in the vicinity that I can't see, and while I do get great audio positioning along the horizontal plane through virtualized 7.1 (yes, it's one of those XAudio2-based games), I get NO vertical cues whatsoever and find myself frantically looking up and down, trying to spot exactly where my target is. Not good in a game that makes such a big deal out of the vertical plane of movement.
 
I'm sure you all have heard discussions on the merits of CMSS-3D Headphone and Dolby Headphone and so forth for gaming, with most of the flaws stemming from the use of generic HRTFs. Then you learn about devices like the Smyth SVS Realiser that use individual, easy-to-tune HRTFs with the help of earbud microphones and existing speaker systems...except it's designed to emulate surround speaker configurations. What if it could be adapted to use the personalized HRTFs for gaming environments that aren't limited to sounds coming from arbitrary speaker directions? (One of CMSS-3D Headphone's advantages is that with DirectSound3D or OpenAL games, it doesn't emulate a virtual 7.1 speaker setup, but instead filters the sounds in the game environment based on their spatial 3D coordinates and their location relative to your position. This still doesn't fix the generic HRTF problem, though.) I'm not so sure if that's even possible using earbud microphones, but I'm sure someone could find a way.

 

[jgazal]

Another binaural microphone: http://www.roland.com/recorder/CS-10EM/#about (Binaural recordings on CS-10EM Experience audio 3D!)

 

[jcx]

http://www.wendycarlos.com/gosurround.html - some options have been explored over time

 

[Maxvla]

Even though binaural may not perfectly handle elevation cues because of individual shapes of heads, it's still an almost perfect technology for recreating sound if you have a good artificial head to record it. If you've ever heard example sound files with sounds moving around your head at different distances, you know that it works brilliantly. At the same time it's not complex at all.
 
In the end, all sounds we hear are combined in our two ear canals, hence two channel audio is in theory all you ever need to perfectly recreate a sound experience. Catch these combined sounds in each ear canal in an artificial head, and then reproduce them at that point in a real person's ear canal and you've recreated the sound experience perfectly or almost perfectly. I guess therefore IEMs are even better suited for this than normal headphones, but normal ones come very close.
 
More recordings should be made with binaural technology (duh). It would be easier to use DSP to reproduce a binaural recording on speakers than to try to make an ordinary recording sound as if it was done binaurally. At least with a binaural recording, you know how the microphones were placed; with ordinary recordings, you do not and can therefore not rebuild the spatial information well.
 

x2, especially the bold part. IEMs are the way to go for immersive sound stage via cues. Large reason I threw off normal headphones altogether.