[alsa-devel] [PATCH v2 0/6] ASoC: fsl_ssi: Fixing various channel slips and bad samples insertions

[Caleb Crome]

On Mon, Dec 7, 2015 at 4:41 AM, arnaud.mouiche at invoxia.com
<arnaud.mouiche at invoxia.com> wrote:
>
> Hi all.
>
> Yes, Caleb and I are working on issues when dealing with N channels, N > 2, but everyone can experience those issues with N=2.
> It is true that a Left/Right switching on audio in not tragic for most users, but I think is will be good if this kind of channel switch can be definitely fixed by the patches I propose.
>
> Here is a very basic test to demonstrate a 2 channels classical scenario that everyone can experience on a imx6 SOC.
>
>   #!/bin/sh
>
>   amixer set PCM 100%,0%
>   arecord -D hw:0,0 -c 2 -r 48000 -f S16_LE  /tmp/foo.wav </dev/null &
>   pid=$!
>   speaker-test -t sine -D hw:0,0 -r 48000 -c 2 -l 1
>   kill $pid
>
> what is done:
> - Right channel is muted at the codec level (so if we try to play something on the right channel only, nothing is heard).
> - We start recording (arecord)
> - then we start a tone generation (speaker-test), starting by the left channel during few first seconds.
>
> what is expected:
> - We should hear the tone on left channel right after speaker-test starts.
>
> what is experienced:
> - 50% of the time, we don't here anything, until speaker-test switch to the right meaning Left and Right channels are switched.
>
> Regards,
> Arnaud
>
Since I haven't been testing with channels==2, I never noticed that
the problem was very common even with 2 channels!  Yikes.

I've been delayed in working on your patch set.  I hope to get to it
soon, but other deadlines are looming, so I might have to wait for the
new year to put it together.

Thanks so much for the patches, I'm really looking forward to finally
having a robust MX6 driver.

-Caleb