P3D Re: binaural sound

["gl@xxxxxxxx" <gl@xxxxxxxxxxxxxx>]

There are a number of reasons why binaural recordings are supposed to be
closer to real hearing that just using (say) two microphones to record in
standard stereo.

One of the main reasons is the 'dummy head' that is used.  It effectively
blocks out various frequencies in complicated ways for each microphone, ie.
it acts as a complex filter.  The idea is to simulate the way sound gets
affected by the human head before it gets to the ears.

The problem is of course that everybody's head and ears are different
shapes, densities (probably) etc, so it's an approximation that may or may
not work for people (I guess like inter-ocular distance).  It makes sense
though that when it does work, it's much more like 'being there' than normal
stereo because it takes account of your head and the way you actually hear.

Saying that, I've not listened to many binaural recordings.

However, calling it superior to 5.1 sound isn't right - they don't compare
because they do different things.  The dummy head method effectively gives
you sound the way it sounds when it reaches your ears - the 5.1 method
instead roughly recreates the sounds _origin_, and lets your own head/ears
do the processing.  Of course 5.1 doesn't give us up/down cues yet (would
need more speakers for that), however, it gives excellent horizontal
positioning.

Now then, regarding '3D sound' as used in audio cards.  At it's simplest,
this is basically 5.1 type mixing, ie. the sound is 'panned' (positioned)
somewhere between a number of surround speakers.  Then you have 'virtual' 3D
sound from 2 speakers, where the programmer tells the sound card where in
virtual 3D space the sound is supposed to be coming from, and the sound card
does various processing to simulate a 3D landscape from just two speakers
(this involves things like different amounts of delay to simulate the sound
reaching one ear before the other, various filtering etc).

The problem with virtual 3D sound is that it's usually very unconvincing,
needs a person to sit in an exact 'sweet spot' and requires careful speaker
positioning, and the minute the listener moves/turns his/her head, the
effect falls apart.  It can often sound better with headphones, because that
eliminates all the unknowns, eg. room size, speaker placement, head
placement, sweet spot, etc.

The more sophisticated virtual audio algorithms even claim to be able to
reproduce up/down cues (ie. make it sound a though a sound is coming from
above/below you) - that's something that's never worked for me, and by all
accounts it's pretty unsuccessful for most people.

The best 3D sound comes undoubtedly from a good surround system (ideally
with speakers above and below you, although that doesn't seem to be a
standard configuration yet) + perhaps some sound processing to reduce
cross-talk between speakers (sound cards like the Creative SB Live already
do that), and good mixing/production/programming of course ;)
--
gl


----- Original Message -----
From: "George A. Themelis" <DrT-3d@xxxxxxx>
To: "Multiple recipients of list PHOTO-3D" <photo-3d@xxxxxxxxxxxxxxxxx>
Sent: Monday, December 13, 1999 11:15 AM
Subject: P3D Re: binaural sound


> I received this with permission to post in p3d.
>
> Hi George:
>
> So, you've discovered binaural recordings. I've been
> listening to them for years. There are very few worthwhile
> musical recordings done binaurally.  There are a few dramas
> (e.g., I believe Stephen King's "The Mist" is one).  But
> (apart from demo recordings of little lasting interest--e.g.,
> taking a shower from the point of view of the person under the
> water, or having a bee buzz around one's head), the most effective
> binaural recordings are nature recordings by Gordon Hempton.
>
> I should also tell you that while binaural recordings, with good
> headphones, nicely reproduce a spacious soundfield, the area they
> do least well is directly "in front" of the listener. That is,
> they don't convincingly convey the space in front of the dummy
> head, but they're great at doing the back, top, and sides, and
> various areas in between.
>
> I should also tell you that binaural recordings may work best
> with earphones that fit snugly inside the ear, like the expensive
> Entymotic phones. And there's some discussion among binaural
> tweaksters about whether (or how much) small differences between
> one's own ears and those of the dummy head contribute to distortion
> of the stereo field. Sound familiar?
>
> If you think listmembers might be interested in the above, feel
> free to post it.
>
>     Hope all's well. Happy Holidays.
>          Stephen Braude
>
> email: braude@xxxxxxxx
>
>
>