[Gary Eickmeier]

KH wrote:
On 4/26/2013 7:06 PM, Gary Eickmeier wrote:


There is no HRTF or head shadowing in a
recording. It just can't work.

There is in binaural, which you never cease to conflate with stereo in
these discussions. But yes, you're correct; there isn't any HRTF in
stereo recording - and no one has ever even suggested such.

Dick Pierce said in his response on 4/4/2013 at 7:27PM to AE's previous
thread:

"The fact is that the HRTF of the original sound field is
eliminate from the listening chain is precisely the problem."

But it does.

I assume you mean "spatial reproduction" here. Unless are you just
contradicting your previous sentence?

Yes, it means spatial reproduction, and yes, I am
contradicting my previous sentence "It just can't work."


I have tried to put across the major, major concept that
stereophonic is not a head-related system,

And you are clearly incorrect in that concept as presented. There is
no HRTF in the recording, true. But the listener, wherever he may
be, or whatever type of system he is listening to, will perceive the
sound as filtered through his HRTF, as you later state clearly
yourself.

Good. You got that. But then why are you harrassing me about it right now?
In a field-type system, which has nothing to do with the number of ears,
head shadowing, HRTF, or anything about the human hearing mechanism,
everyone who hears the music hears it the same way he hears live music. We
are reproducing the object itself, not signals for any particular being's
ears, so we do NOT have to worry about anything that has to do with how we
hear. I have noted the summing localization which happily lets us get away
with fewer channels, but that is heard in the same way by everyone as well,
and does not need to be encoded into the signals in any way beyond normal
stereo recording techniques.

Mr. Pierce was also very confused about how we can possibly know from the
recording where the instruments are. Hence, this thread relating how we
place those recorded sounds within our rooms and how that settles the
question about the difference between a field-type system and a head-related
system.

the recording of ear signals in
any way.

Can you just drop the "ear signals" nonsense? No one has ever
suggested anything about this, but you perpetually erect this
strawman.

If some still think that all arriving sounds must be "encoded" as to
direction from the microphones then they may not be able to conceive of a
field-type system with the direction decided physically and acoustically.
They may still think that the signals alone, going into the ears, should be
able to tell you the directions of all sounds. This sounds silly, I know,
but that is what I have gleaned from some of the remarks.

There is no HRTF or head shadowing because those concepts have
nothing to do with stereo, just binaural.

On the playback end, they certainly do play a primary role. Please
tell us how they do not? Do they play a role in the construction of
your reproduction model? No. That's the major deficiency, and the
part you seem not to grasp. Whatever "field" you construct, it will
be transformed by the listeners HRTF, and will be interpreted for
spatial clues based on that transformation. Your "model" provides
erroneous spatial clues by redirecting what spacial information is on
the recording in all directions.

Do you see what I mean about your thinking that the information on the
recording needs to be directed strictly toward your ears? No, that is NOT
the way it works. Not recording and reproducing ear signals, recording and
reproducing sound in rooms, not sound in ears or heads.

OK here is a deeper explanation to illustrate the difference. There are
basically two ways to reproduce a sensory experience.We can reproduce the
sensory inputs or we can reproduce the object itself and let your own senses
experience it. The first one would be like binaural, in which we record and
reproduce ear signals by using a dummy head and then headphones. The sensory
experience of when the head was there at the live event, reproduced by
direct sensory input with headphones. The stereophonic system is like the
second method - reproduce the object itself, a sound field in a room, and
let everyone experience that sound with his own senses.

The part about redirecting spatial information on the recording in all
directions is simply a part of reconstructing sound fields that were
recorded so that they can come from similar directions at home


Finally, and the
hardest to understand,

Simply because it is factually inaccurate, and physically impossible.

we can bounce some of the output of the speakers from
the surfaces of our room in order to use the acoustics of the room
to help build a real space around the recorded sound. In a properly
set up system this effect can actually decode,

You already stipulated that it is not "encoded", therefore you cannot
"decode" it.

I told you this isn't easy. No, the directions of all of the spatial content
is not encoded as such, it is reconstructed int he playback room by means of
time delay and positioning of extra speakers or just using reflected sound
from the main speakers.

or paint, the recorded reverberance onto
the appropriate walls of your listening room.

And yet you have no theory for how this is actually performed. You
continually use words like "decode" and terms like "appropriate walls"
in a definition-free manner and expect the readers to somehow "grok"
what you feel they mean to you.

Tell us how you define "appropriate", and tell us how you direct the
appropriate sound to the appropriate wall.

A recording is made of a saxphone on the left side of the orchestra. The
recording contains reverberance from the sax bouncing some of its output off
the left side wall. We play this back on a system with multidirectional
speakers. The left speaker contains the vast majority of the sound of the
saxophone and its reverberance from the left wall of the concert hall. The
right channel has little or none of it. On playback both speakers direct
part of their output toward their nearest side wall, but only the left
speaker has the sound of the sax and its reverberance. Therefore, in the
playback model the sax sound is bounced more from the left wall of the
playback room than the right - which would be the appropriate wall.

Even shorter version: How does stereo know where the sound came from?
Because we PUT it there, in front of us, where it belongs. The
sounds that we hear in the reproduction are real, physical sounds
that exist in real space in front of you, you hear them with your
natural hearing, your own HRTF and head shadowing, a summing
localization permits auditory events anywhere along a line between
the speakers, and the recorded space gives the presentation the
"flavor" of the original space if such is contained in the recording.

Whoa there partner, this "shorter version", scrubbed of undefined
terms and physical impossibilities, is pretty much accurate. And yes,
it works surprisingly well. But once you add in all the reflections
of sounds that should be "there, in front of us, where it belongs",
thinking that somehow that decodes information not present on the
recording, that's where you run off the tracks.

How is the reverberance not contained in the recording?


Obviously you like the way that sounds, great, but it simply is not
doing what you claim it is, and you've provided no plausible
mechanism for how it could.

You want me to keep saying it again and again until - never mind.


All of this was well known by the pioneers at the Bell Labs who were
doing the experiments, but of late it has become confused with
binaural

Maybe you have been confused about it; I see no evidence that anyone
else is so afflicted.

with some
thinking that what is wrong with stereo is we need crosstalk
cancellation, or we need to record the HRTF, or head shadowing,

Again, who? I've not seen any evidence of that here.

Pierce. And many of your statements, as above.

Gary Eickmeier