< nothing below implies that an Ambisonics microphone is the best
solution for all sound capture cases! >
On 10/21/20 6:37 AM, Augustine Leudar wrote:
Hi Steve,
An interesting proposal. You'll excuse perhaps my misunderstanding here,
but my understanding was that ambisonics A format consists of a omni plus
x,y,z to give 4 signals that are then converted to b format.
First order "A format" is just the four signals coming from the four
capsules. They are matrixed and filtered to generate the B format
signals, which are the equivalent of an omni and three figure of eight
at right angles to each other. Ideally the B format polar patterns are
frequency independent. In reality this is not the case (for any
capsule), but a properly calibrated Ambisonics microphone is very good
in that respect, with polar patterns consistent up to very high
frequencies (as Fons points out elsewhere in the thread).
I built my
own ambisonics mic and decoder once a long time ago so perhaps things have
changed, I think I put Left + & - right = x Front + & - Back = y Top + & -
bottom = z All summed * 0.257 = w into my decoder.
Hmmm, just curious, did you also add high frequency non-coincidence
compensation filters?
A simple add/substract matrix (maybe that is not all you did? - you also
say later in the thread you eq'ed the capsules) will work fine for
frequencies where the array can be considered small relative to the
wavelength in air. The effects of the non-coincidence of the capsules
starts to show up at 3Khz or so in my case, exactly where depends on the
radius of the array, and needs to be compensated somehow. This is
usually done by filtering the B format signals.
(see attached graph of unfiltered and filtered response at 0 degrees
incidence for my TinySpHEAR mics - these are very old plots but useful
anyway as they show the difference those filters make)
Best,
-- Fernando
. Anyway I digress -
this basically will give you a sounds position in 3D space , or at least on
the surface of a sphere proximity not being so well recreated. So if you
had say, 8 ambisonics mics pointed in 8 different directions ,in order to
record that small 8th of a sphere (or in this case cube) in the direction
they are pointed in (which is what the ORTF does) - how would this work if
the decoder if a format gives spherical coordinates rather than an 8th of a
sphere/cube (hope this makes sense), and with the capsules pointing in all
directions... ?
...
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sphear_raw_0.png
Type: image/png
Size: 33125 bytes
Desc: not available
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20201022/17fb6597/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sphear_filtered_0.png
Type: image/png
Size: 26715 bytes
Desc: not available
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20201022/17fb6597/attachment-0001.png>
_______________________________________________
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.
