Skip to content

Commit

Permalink
aop: Stream microphone samples via AOP
Browse files Browse the repository at this point in the history
Can capture *something* from ADMAC, see povik's commit message from
fc66046. Tested j293 13.5 (known 12.3 ABI breaks are noted, but I don't
think it'll be feasible to support 12.3 given how ugly aop is).

I figured I'd do ALS first just to get the boilerplate down. In the
meantime someone can work on figuring out the output format. The samples
streamed to ADMAC starts off with a sine wave, then it sends what looks
to be 3 LE 32-bit float samples per frame (apple website says j293 has a
3-microphone array). But it's weird: the first two are always identical,
and only the third is by itself. Either I'm getting the PDM calib and/or
misunderstanding the output format, or channel 0/1 is garbage.

000c654c: 008ef84b 008ef84b 00bbc592   K...K.......
000c6558: 00e2ac1d 00e2ac1d 00000000   ............
000c6564: 00000000 00000000 00000000   ............
000c6570: 00878344 00878344 00b613c6   D...D.......
000c657c: 00d181cb 00d181cb 00000000   ............

Also, device trees say up to 7 channels and "bytesPerSample = 2" ??
And we set those exact values in the PDM config too??

    audio-pdm2 {
	channelsEnabled = 7
	channelCount = 3
	channelsSupported = 7
	bytesPerSample = 2

Signed-off-by: Eileen Yoon <[email protected]>
  • Loading branch information
eiln committed Jan 20, 2024
1 parent 7ff0430 commit 4c696e1
Show file tree
Hide file tree
Showing 10 changed files with 614 additions and 468 deletions.
310 changes: 0 additions & 310 deletions proxyclient/experiments/aop.py

This file was deleted.

Loading

0 comments on commit 4c696e1

Please sign in to comment.