Trond, not sure if this will be of any help but I had contacted Izotope
Support about best practices for denoising B-format last year.
(A-format was not an option for us since our Soundfield ST450 MKII mic
outputs B-format directly.)
Here was their answer:
"I spoke with our head DSP Engineer about
Thanks for all the replies to my initial question.
I have done some testing with all three approaches today. My conclusion so far
is that running noise reduction on the B-format signal as two stereo pairs (WX
and YZ) ends up distorting the sound field (spatial artefacts, less pronounced
spatial
I think I would prefer to do the noise reduction on the B format, rather
than the original A format, as it's a bit more robust to small gain/phase
errors. Although it's often useful to do a B-to-A conversion, process then
go back using the ATK, I can't see any advantage in this case.
There's been
Right after hitting "send", I remembered:
I did the noise reduction in a DAW.
First sampled the noise from the WX pair.
Did adjustments to the noise reduction module.
Processed that and bounced WX.
Then removed the WX clip from the DAW track,
opened the YZ file onto the same track
(starting fro
Trond
Been there, but that was many years ago.
Have you tried using the noise reduction modules as plugins?
- Split the channels into 2-channel pairs.
- For example in ProTools or Reaper or similar, insert two instances of
the noise reduction VST into two separate stereo tracks.
- Route and bou
PS. of course amplitudes of the signal should also stay the same
On Mon, Jan 4, 2016 at 1:33 PM, Andres Cabrera
wrote:
> I would guess that as long as the phases of the remaining components in
> the encoded signals are not significantly affected, there should be no
> spatial artifacts introduced
I would guess that as long as the phases of the remaining components in the
encoded signals are not significantly affected, there should be no spatial
artifacts introduced (only the spectral artifacts should be present). Not
sure however how much phases are affected by denoisers.
Cheers,
Andrés
O
--On 04 January 2016 21:17 +0100 Trond Lossius
wrote:
> 1) I treat the original A-format recordings, and afterwards I convert
> the processed files to B-format
I have done this - when I had a situation that one channel only of the
A-format contained the noise. I was not aware of any anomalies a
Hi,
I’ve done some ambisonic recordings of ambience in very quiet environments
using the SoundField SPS200 mic and a SoundDevice 788T recorder. The resulting
recordings are A-format, and I use the SoundField SPS200 plugin to convert the
A-format recordings to B-format. The resulting recordings
