Hello everybody!
I have TDM8 mic array. All works fine. But signal which I get from arecord command is shifted to 65000 approximately. This is not problem 'cos I correct shift programmatically. And recently I noticed that at high sound volume I get very bad picture.
It is very similar to overflowing in somewhere. And I suppose this overflowing is connected to signal shift. How I can adjust Nano to remove this signal offset? I have found that fsync width can enlarge and make smaller signal offset, but not so small as I wish.
Thank you in advance!
I got what I want:
with fsync width = 15:
amixer -c tegrasndt210ref sset ‘I2S4 fsync width’ 15
getting sound with command:
arecord -D hw:tegrasndt210ref,0 -c 16 -r 16000 -f S16_LE cap.wav
BUT, I have 16 channels instead of 8, 8 channels are nulls:
and I get 8 kHz instead of 16 kHz
How I can get 8 channels and 16 kHz?
Hello!
What is the sample size that is produced by the mic array? The above would seem to suggest that it is 32-bit and not 16-bit. Or maybe 24-bits in 32-bit samples. Hence, why you see data in the lower 16-bits only. If that is the case then you could try …
amixer -c tegrasndt210ref sset ‘I2S4 fsync width’ 31
arecord -D hw:tegrasndt210ref,0 -c 8 -r 16000 -f S32_LE cap.wav
When you say it is not 16kHz, what do you mean? By setting ‘-r 16000’ should make the fsync period 16kHz.
Regards,
Jon
Sorry for delay !)
mic array produces 32 bit per sample probably, probably 'cos I did not see it myself without ocsilloscope.
amixer -c tegrasndt210ref sset ‘I2S4 fsync width’ 31
this doesnt work - channels simply shift, first becomes second and so on
about frequency - question not actual, that was my error
about 24bit in 32 bit sample - iteresting idea, I will try to do something about it
hi again!
data was in high 2 bytes of DWORD, as you said + there was an error in codec
80 db
thank you very much!)
Hi!
Thanks for the update. I assume that all your issues are resolved. Please let us know if there is anything else.
Jon
