On Mon, Nov 04, 2013 at 09:23:37PM +0000, [email protected] wrote:

> How does the calibration process work?  Is it a question of matching
> (trying to find 4 mic capsules with as nearly the same frequency
> response as possible)?  Or is it a process of measuring the frequency
> responses of the 4 mics and then compensating for any differences in
> software after recording?

Both.

The capsules should be selected, the most important parameter to match
being the directivity (polar pattern) as a function of frequency. Small
differences in sensitivity and frequency response are less important as
these can be compensated for by processing.

> Is the ‘mic calibration’ file therefore just a set of octave/third
> octave/one-twentyfourth octave band values between say 30 Hz and 16
> kHz for each individual mic, as measured in an anechoic chamber under
> standard conditions?

What exactly is in the 'mic calibration file' depends on the software
it is supposed to be used with, but it will normally consist of a subset
or all of the following:

* filters to be applied to the mic signals,
* coefficients for the A/B matrix,
* filters to be applied after the matrix,

in some form, or

* a full 4 by 4 matrix of impulse responses, which 
  combines all of the above and more, to be used by
  a convolution processor.

This data is normally computed from impulse response measurements
performed on each mic in semi-anechoic conditions.

If the capsules are very well matched and have very small variations
then simpler calibration measurements can be used, leading to e.g.
just four gain factors wich can be set inside the mic. In that case
the A/B processor is not matched to a single mic but can be used
with any one of the same type. But this requires a very tightly
controlled manufacturing process for the capsules and the complete
microphone which means they will be expensive.

Note that A-format signals are combined (in the matrix) to form
B-format, therefore the phase response is as important as the
amplitude one. A set of traditional octave or fractional octave
band filter gains would be useless.

Ciao,

-- 
FA

A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)

_______________________________________________
Sursound mailing list
[email protected]
https://mail.music.vt.edu/mailman/listinfo/sursound

Reply via email to