Jörn Nettingsmeier wrote:
On 12/07/2015 02:55 AM, Stefan Schreiber wrote:
Your plots refer to an empty sphere, don't they?
no, fons' plots are about a rigid sphere with omni mikes in the surface.
Take for example the 8kHz response, which is the most directive one in
the set. It's something like 3.5 dB
down at 90 degrees. Apart from the narrow peak at the
back, that's subcardioid.
See above. We currently don't know the exact frequency responses...
the point is that you are dealing with omni microphones, and the only
directivity that you can work with other than matrixing several
capsules is the acoustic shadow of the sphere.
But IF the microphone is based on stereophonic principles you wouldn't
talk of "matrixing several capsules" - which refers to some form of a
spherical array microphone, and lots of math to derive some acoustical
image. (Just to keep our discussion open and neutral. You know what I
mean... :-) )
And? This is exactly the case here. (IRT cross in two dimensions and the
8-channel hedgehog in the described cuboid form are isotropic. All 12
microphone stereo angles are 90º. A cube is a Platonic solid, "as
everybody knows".)
stefan, i think you are confused about the fact that the hedgehog or
any other open microphone arrays employ directional microphones. the
OZO cannot, because it is a rigid sphere with no way for a second
acoustic path to each capsule.
? I wrote that there is some kind of cardioid response at the capsule
places. The exact response in the Ozo case would have to be measured
anyway. (We will measure Fon's rigid sphere response - already some kind
of subcardioid - combined with some additional shield effect caused by
the electronics inside the Ozo sphere.)
The response is recorded by omnidirectional capsules. What matters is
the system response, so you have some (integral) cardioid response.
The 'hedgehog' doesn't produce anything like that, nor
is it meant to.
It does and is meant to!
this is not the case. the eight-supercards hedgehog is not meant to be
rotated ever. it has a 1:1 mapping from mics to speakers. rotating it
will give massive shifts both in source width and timbre.
Reads well, but wait: The desired optimal binaural rotation would
simulate what happens if you rotate your head in the sound field of 8
REAL speakers. (You simply try to copy this.)
Your timbre and width issues exist - but even in the loudspeaker case.
(And in a different way, even in the case of a real sound field.)
I believe your ear/brain will correct for a lot of these effects, from a
perceptual point of view. Think of timbre changes caused by "pinnae
movements", for example. "Should exist", but who cares... 8-)
the design goal of Theile's arrays is to minimize crosstalk and ensure
well-defined decorrelation between speaker channels, not isotropy. it
is intended to produce ambience only.
Even if the hedgehog design is "classified" as ambience mike, it offers
at least some directivity. (Which is limited by the 10dB separation
issue, mentioned in chapter 6.1 of the cited Theile/Wittek paper.)
Some suggestions:
- The hedgehog design doesn't seem to be Dr. Theile's, at least I never
have read this. (The "IRT Cross" seems to be, for example.)
-
design goal ... is to minimize crosstalk and ensure well-defined
decorrelation between speaker channels, not isotropy. it is intended
to produce ambience only.
I disagree here, because I don't see any X-talk minimization or
decorrelation measures < at all > in this design. (With exception of the
use of < supercardioid > capsules.) The rest is a function of cube
layout and optimized dimensions, not of certain minimization measures.
Now I am turning tables. IF the microphone built into the Ozo camera is
not based on stereophonic principles, why is this supposed "spherical
array microphone" constructed in such an odd way?
- You have only 8 capsules, whereas the competition uses 32 or even 64...
a) http://www.mhacoustics.com/products
b) http://visisonics.com/
- And IF Nokia has developped some incredibly advanced matrixing
operations, why didn't they mount capsules on top of the sphere or in
the sphere - like "everybody else does"?
- IF the Ozo microphone is supposed to capture ("just") B format: why is
the design more complicated than necessary, and why don't they export B
format?
Now you have to prove your point... :-D
Best,
Stefan
_______________________________________________
Sursound mailing list
[email protected]
https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe here, edit
account or options, view archives and so on.
