Sampo Syreeni wrote:
On 2017-01-09, Stefan Schreiber wrote:
The critique I'd have for such panning laws is that they don't
really respect the ambisonic/Gerzon theory, especially at the low
frequencies.
Stereophonic panning laws are based on Blumlein's stereo theory,
which in Wittek's opinion is pretty close to sound fields anyway.
Correct, but only in the high order, dense limit. In the low order,
sparse array case, which especially four speaker POA deals with, you
can do better. That's why POA decoders don't go in-phase but max
energy even at HF, and especially why we have shelf filters which cut
the decode down to velocity coherence at the low end.
But Blumlein developped his "physical cue" stereo just for 2 speakers. ;-)
Forget about any "high order" re-interpretations. It's all about ITD,
ILD and say summing location.
It is pretty obvious that Ambisonics theory is influenced by Blumlein
stereo...
Now, 3-channel stereo (+ extensions) seems to be some early version of
WFS...
http://www.hauptmikrofon.de/HW/Wittek_thesis_201207.pdf
p. 28:
In America Steinberg, Snow and Fletcher (Steinberg and Snow, 1934)
from Bell Laboratories
explored the 'acoustic curtain', see Figure 3-3.
Snow described their ideas in
this way: "The myriad loudspeakers of the screen, acting as point
sources of sound identical
with the sound heard by the microphones, would project a true copy of
the original sound into
the listening area. The observer would then employ ordinary binaural
listening, and his ears
would be stimulated by sounds identical to those he would have heard
coming from the original
sound source." (Snow, 1953)
(= WFS!)
These scientists quickly noticed that, due to technical constraints,
it would not be feasible to
put their ideas into practice. As a compromise, they limited the
practical system to three channels,
accepting that the original aim of recreating the real sound field
would no longer be ful
filled. The three-channel stereophony produced in this way was
therefore not created as a
result of a mathematical analysis of the sound field, but rather as an
engineering compromise.
Its directional effect is based on perceptual phenomena such as the
precedence effect and
level and time difference stereophony.
In contrast, Blumlein (1933) aimed at a proportional reproduction of
the directional image of
the recorded scene by recreating the original physical auditory cues.
Snow (1953) pointed out, regarding the basic difference between the
n-channel acoustic curtain
and 3-channel stereophony: "This arrangement [3-channel stereophony,
see Figure 3-3]
does indeed give good auditory perspective, but what has not been
generally appreciated is
that conditions are now so different from the impractical <infinite
screen> setup that a different
hearing mechanism is used by the brain."
Researchers who observe contradictions in the generally accepted
summing localisation theory
quote this statement by Snow.
Let's say the creators of stereo (including binaural stereo, developped
even earlier in the 19th century) thought all in terms of physical
acoustics and psychoacoustics...
Obviously all of those decoding principles converge to holophony in
the high order, dense array limit, so that Wittek is correct in that case.
Well, look to the title of Wittek's dissertation...
Switching guru mode on:
< Every reasonable stereophonic system will become some copy of
real-world acoustics >, as long as you can use infinite ressources to
prove your point! (Do you agree?!)
Example:
HOA DirAC... :-)
However I'd argue that the whole point of POA is to optimally deal
with the low order, sparse array case, where each of the decoding
principles are pretty far from convergence, and in very different ways
-- intensity panning pretty much corresponding to an in-phase decode,
which we already know is *not* always optimal.
1st order Ambisonics is "just" the 2nd most simple implementation of a
sound field. Should we not start interpretation in this way, not with
some "sparse array" we should use somehow??
The "sparse" decoder array is an implementation issue. (Order 0 was
already obsolete when Ambisonics was developped. But quadraphonic
systems already existed. "Some believe that Ambisonics is a kind of
improved quadraphonic system"... 8-) )
Anyway: I can't quite follow! You do an in-phase decode if you want to
enlarge some (usable) listening area. It is not some optimal decoding
strategy for POA, but so what?
You could also say that your argument is some "proof" that intensity
panning/VBAP has some bigger sweet spot than POA. (Maybe true, but this
is not really the motivation behind some in-phase area decoder.)
I would not treat stereophonic systems as some "deformed" implementation
of Ambisonics. They are not...
Secondly, "theoretically perfect" systems are mostly "not really"
perfect in reality:
http://www.linkwitzlab.com/Recording/record-play-map.htm
The array has equal sensitivity front and back but low pickup from the
sides. I doubt that the Blumlein array is optimal for mapping an
auditory scene for stereo loudspeaker playback.
So, does the Blumlein microphone pick up too much reverberation or not?
And do we have some side problems?
This might all depend on the situation.
Best,
Stefan
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20170110/ff6c263a/attachment.html>
_______________________________________________
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe here, edit
account or options, view archives and so on.
