Repository navigation
[Bug][TGL]Pop noise on I2S max98373 speaker when doing pause/resume test #2255
Description
Describe the bug
Slight pop noise can be observed when do pause during playing sine wav via speaker
To Reproduce
- aplay -Dhw:0,0 -f s16_le -c 2 -r 48000 sine_48k_16bit_10s_stereo_l5_r2.wav -vv -i
- press SPACE to pause resume
Reproduction Rate
3/3
Expected behavior
no pop noise after resuming playback
Impact
slight pop noise
More info:
Check captured audio via DSM
dmesg.txt
sof_trace.txt
capture.zip
Environment
FW: on https://github.com/thesofproject/sof/tree/tgl-007-drop-stable.
TPLG: same with FW branch, sof-tgl-max98373-rt5682.tplg
Kernel:https://github.com/thesofproject/linux/commits/topic/sof-dev (commit: 1b2f260)
- PR#2208 + PR#2240 for DP fix
- PR#2184 for wov S0i2
+PR#2238 for max98373-sdw jack switch
Activity
[-][Bug][TGL]Pop noise on I2S max98337 speaker when doing pause/resume test [/-][+][Bug][TGL]Pop noise on I2S max98373 speaker when doing pause/resume test [/+]RanderWang commentedon Aug 27, 2020
update status: I can reproduce this issue. no pop noise is generated with only stream of zero data. And I also check it with SSP loop back, no garbage data is captured. I don't think it is caused by FW for above evidence
RanderWang commentedon Aug 27, 2020
if we test start-stop pattern, stream of zero data also produce garbage data at beginning & end. The difference of start-stop pattern and pause-release pattern is that SSP clk is stopped in start-stop case.
RanderWang commentedon Aug 31, 2020
I enabled SSP loop back in this case and disable DSM feedback input. The loop back data showed nothing wrong. So the data sent by SSP was correct. It is caused by 98373 only. The pop noise is very slight and user can't notice it in daily usage.