Hi Tim,
Indeed, SNR is poor at low frequencies for this sort of technique as
it relies on the phase difference between the two capsules as produced
by their separation. The separation can't be too large as once it gets
over half a wavelength the phase starts wrapping around but this means
that at
HI, the front-back reversals problem isn't peculiar to ambisonics - it can
sometimes happen in real life (most often, though not always, in a symmetrical
space) and is well known in the laboratory. It's really a subset of the 'cones
of confusion' problem, which Batteau proposed that Pinnae
An open paper on directional mics
http://www.iwaenc.org/proceedings/2001/main/data/buck.pdf
I had a idea that I could use 2 close mounted mics (o) for Upper frequency
ranges and a longer distance between 2 additional elements (x) to have better
signal to noise ratio for the lower frequency
Thanks very much. You are so kind. I will have a try as your suggestion.
Best regards,
Rilin
Chen
Do let us know with what results !
These type of problems are usually a
You can make a velocity microphone by subtracting two closely spaced omnis. I
saw a paper on making ambisonic recordings like this a couple of years ago:
http://www.aes.org/e-lib/browse.cfm?elib=15389
The problems, though, are spatial aliasing at high frequencies and lack of
sensitivity at low
At 11:54 08-01-13, Tim Collins wrote:
You can make a velocity microphone by subtracting two closely spaced omnis.
I seem to recall that Blumlein did this and also found that the s-n
ratio was poor.
David
___
Sursound mailing list
Without more details, it hard to speculate about problems, but I'll note
that subtracting the outputs of two omnis to get a fig-8 response will
result in a frequency response that rises 6dB/octave with a 90 degree phase
shift relative to the sum. Unless that is corrected, these signals are not
Without more details, it hard to speculate about problems, but I'll note
that subtracting the outputs of two omnis to get a fig-8 response will
result in a frequency response that rises 6dB/octave with a 90 degree phase
shift relative to the sum. Unless that is corrected, these signals are not
On Mon, Jan 07, 2013 at 03:12:26PM -0800, Aaron Heller wrote:
Without more details, it hard to speculate about problems, but I'll note
that subtracting the outputs of two omnis to get a fig-8 response will
result in a frequency response that rises 6dB/octave with a 90 degree phase
shift
The mics are omnis
It might be stating the obvious here but how would you get ANY directional
information out of four OMNI mics close together? :-)
Surely to get a fig-8 response you need two CARDIODS back to back (one of
which is polarity reversed obviously).
Mike Felton
On 7 January 2013
Hi,
Seems likely then that the problem is related to the usual in
head problem that many (maybe even most) people people have with
simple, non matched HRTF based systems, whether driven from decoded
Ambisonic signals or from binaurally recorded (dummy head) material.
There's been a lot of
Hello, all
I have designed an array of four microphones to get B-format signal. The
coordinates of mics
are mic1 (-d/2,0), mic2 (d/2,0), mic3 (0, d/2), mic4 (0,-d/2),and d is about
18mm.Every opposite
pair can be composed into a first order differential array and get a pattern of
figure 8. So
Hello, all
[ ... ]
The W component equals to the mean of signals from four mics. Then I
decoded them to get the
feeding signal for four loudspeakers, and used HRTFs to get the left and
right signal for headphone.
1. Furthermore, the experiment is made that a person was walking
On Sun, Jan 06, 2013 at 09:18:43AM +, Chenrilin wrote:
Hello, all
I have designed an array of four microphones to get B-format signal. The
coordinates of mics
are mic1 (-d/2,0), mic2 (d/2,0), mic3 (0, d/2), mic4 (0,-d/2),and d is about
18mm.Every opposite
pair can be composed into a
Do you get the same problem if you playback through four loudspeakers, or
is it only after HRTF manipulation ?
We'll need a bit more information.
* Are the mics omnis, or directional ? If directional, what
are the directions ?
* How exactly to you compute X and Y frm the mic signals ?
15 matches
Mail list logo
