Make mono to stereo conversion automatic (fixes fmradio)

[linknum23]

Closes #454.

To test a nearby radio station is 91.7, verify that stereo audio is output on each source.

[Lohrer]

If you'd rather just use the auto mono->stereo conversion that the plug plugin uses, we can add a route plugin in between plug and dmix. The current problem is plug thinks it's output is 8 channels, not 2, and I think that is messing with the auto conversions. So the current situation: [1/2]->plug->[8]->dmix. [N]=>N channel audio.

With the following setup we'll get: [1/2]->plug->[2]->route->[8]->dmix

pcm.ch1 {
    type                plug
    slave.pcm {
        type            route
        slave.pcm       "dmixer"
        slave.channels  8
        ttable.0.6      0.64
        ttable.1.7      0.64
    }
    slave.channels      2
}
pcm.ch2 {
    type                plug
    slave.pcm {
        type            route
        slave.pcm       "dmixer"
        slave.channels  8
        ttable.0.0      0.64
        ttable.1.1      0.64
    }
    slave.channels      2
}
pcm.ch3 {
    type                plug
    slave.pcm {
        type            route
        slave.pcm       "dmixer"
        slave.channels  8
        ttable.0.4      0.64
        ttable.1.5      0.64
    }
    slave.channels      2
}

[linknum23]

Oddly this does not work with vlc on ch1 and ch3. Works fine for ch2.

[linknum23]

With the extra layer in play here is what the verbose output -vvv looks like from vlc:

[0204b6d8] alsa audio output debug: using ALSA device: ch1
[0204b6d8] alsa audio output debug:  Plug PCM: Route conversion PCM
  Transformation table:
    0 <- none
    1 <- none
    2 <- none
    3 <- none
    4 <- none
    5 <- none
    6 <- 0*0.64
    7 <- 1*0.64

Without the extra layer:

[0135d6d8] alsa audio output debug: using ALSA device: ch1
[0135d6d8] alsa audio output debug:  Plug PCM: Direct Stream Mixing PCM

[Lohrer]

Not so odd I guess, ch2 uses physical channels 0 and 1 on the 8-channel CM6206, so no remapping is required. What is weird is that the FM radio stream (mono) and Spotify (stereo) both worked so I assumed the rest of the stereo streams would work...

My AmpliPi has been borrowed so I can't test, but I assume when you try ch3 you get:

Plug PCM: Route conversion PCM
  Transformation table:
    0 <- none
    1 <- none
    2 <- none
    3 <- none
    4 <- 0*0.64
    5 <- 1*0.64
    6 <- none
    7 <- none

[Lohrer]

Those messages come from https://github.com/alsa-project/alsa-lib/blob/master/src/pcm/pcm_dmix.c and https://github.com/alsa-project/alsa-lib/blob/master/src/pcm/pcm_route.c and look exactly like what I'd expect. I think if there's an error it might lie elsewhere...

[linknum23]

Another way to do this would be with 2 layers of plug:

pcm.ch1 {
    type                plug
    slave.pcm {
        type            plug
        slave.pcm       "dmixer"
        slave.channels  8
        ttable.0.6      0.64
        ttable.1.7      0.64
    }
    slave.channels      2
}
pcm.ch2 {
    type                plug
    slave.pcm {
        type            plug
        slave.pcm       "dmixer"
        slave.channels  8
        ttable.0.0      0.64
        ttable.1.1      0.64
    }
    slave.channels      2
}
pcm.ch3 {
    type                plug
    slave.pcm {
        type            plug
        slave.pcm       "dmixer"
        slave.channels  8
        ttable.0.4      0.64
        ttable.1.5      0.64
    }
    slave.channels      2
}

[linknum23]

2 layers of plug works and was tested with Pandora, Airplay, Spotify, Internet Radio, and FM Radio

[codecov-commenter]

Codecov Report

Merging #455 (d1c89d9) into develop (17390bb) will increase coverage by 8.09%.
The diff coverage is 67.00%.

📣 This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more

@@             Coverage Diff             @@
##           develop     #455      +/-   ##
===========================================
+ Coverage    56.25%   64.34%   +8.09%     
===========================================
  Files           12       15       +3     
  Lines         2839     4171    +1332     
===========================================
+ Hits          1597     2684    +1087     
- Misses        1242     1487     +245     

Flag	Coverage Δ
unittests	64.34% <67.00%> (+8.09%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
amplipi/asgi.py	0.00% <0.00%> (ø)
amplipi/updater/asgi.py	0.00% <0.00%> (ø)
amplipi/rt.py	26.30% <23.07%> (-0.13%)	⬇️
amplipi/mpris.py	37.62% <37.62%> (ø)
amplipi/streams.py	46.11% <50.32%> (+4.48%)	⬆️
amplipi/utils.py	70.98% <71.87%> (-0.24%)	⬇️
amplipi/ctrl.py	80.99% <74.86%> (-1.44%)	⬇️
amplipi/app.py	76.97% <75.00%> (-8.42%)	⬇️
tests/test_rest.py	94.07% <97.87%> (ø)
amplipi/models.py	95.47% <100.00%> (+0.47%)	⬆️

... and 1 file with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[linknum23]

I think this setup makes the most sense going forward since it handles any weird cases where mono audio will get played. We will need to continue testing this to make sure there aren't any performance penalties with the double plug setup.

[linknum23]

@Lohrer Do you think this makes sense?
Basic testing with 4 various streams playing seems to use about 15-20% for each stream which seems pretty reasonable.

[Lohrer]

Looks good! I just think there's a some comments that might need to reflect the latest changes.