loud audio glitch between tracks and on stream stop

[mtnbrit]

Thats an interesting result and might suggest a bug in MPD software volume routine. These kinds of should be reported to the MPD maintainer https://github.com/MusicPlayerDaemon/MPD/issues

Is there a tip for setting mpd's log_level to "verbose"? as Im finding if I edit mpd.conf and restart, its getting set back to "default". I tried editing mpd.conf and doing kill -HUP <pid of mpd> but no extra logging appears in /var/log/mpd/log

[mtnbrit]

I figured it out, setting verbose log to mpd (sudo nano /etc/mpd.conf and change log_level option to verbose), restart mpd (sudo systemctl restart mpd)

there's nothing different in the verbose mpd log between events when it glitches or doesnt glitch.

Ill open an issue on the mpd GitHub and see what happens.

[TheOldPresbyope]

My experience with posting MPD issues is that you won't get a useful answer unless you meet three requirements:

1) you observe the troubleshooting dictum in the MPD documentation: "Do not ask for help unless you have the latest MPD version." MPD 0.21.21 is current; moOde 6.4.2 contains MPD 0.21.16. At least you're close.
2) you structure your issue following the format and including the information you'll see in the github issues log.
3) you demonstrate that this is an issue with MPD and not with ALSA.

If there's any doubt about #3, you'll likely be told to go talk to the ALSA maintainers.


Just my 2-cents worth.
Regards,
Kent

[mtnbrit]

Heres a link to the issue, we'll see what happens.

https://github.com/MusicPlayerDaemon/MPD/issues/813

[Tim Curtis]

If MPD maintainer requests a later version of MPD I can provide you with either a binary or the compile recipe if you want to build it yourself.

[mtnbrit]

The mpd maintainer is asking me to configure mpd to output to a loopback device and capture the audio to see if its reproducible. Any tips how to get mpd to output to a loopback device? 

In researching this, I found this oled thread and post where you may be considering adding the ability to send the output to multiple devices. Any progress on that? Would be a great general purpose feature for multiple applications IMO.

I have added the aloop device and I can create an asound.conf file, but how to make moode tell mpd to use the loopback?

[Tim Curtis]

ALSA loopback provides an exact mirror copy of the PCM audio output from the designated application which in our case is MPD. The capture would show whether or not the audio waveform coming out of MPD contained the glitch. If glitch is present in the waveform then prolly a bug in MPD. If no glitch in the waveform then prolly the audio device has no auto-mute capability when it detects PCM stream stopped.

I have some ALSA loopback recipes but no clue how to capture the PCM stream to a file that can be loaded into a pro waveform editor like Adobe Audition @gbh_uk

[TheOldPresbyope]

Or even the FOSS application Audacity which has many (but not all) of the features of Adobe Audition. 

It runs on Windows 10, macOS, and Linux, according to their website. I've used it on Windows and Linux to show the local high school's theater sound techs how they can create/manipulate sound effects for shows.

Not to denigrate Adobe Audition. Like all the adobe tools, it's feature-full and well-polished.

Regards,
Kent

[mtnbrit]

It turned out to be quite straightforward. I loaded the aloop module and created this /etc/asound.conf from here. Restarted. Then used arecord to grab a wav file from the loopback. 

Code:

arecord -D plughw:Loopback,1 -d 60 -f dat test.wav 

I moved the wav files over my Mac and opened them with Audacity.

The wav files dont have the glitches, which were clearly audible on my DAC while recording.

So I guess it points to the wm8804 device driver as you say Tim. 

Is there a way to tell if the moOde kernel has this version of the driver? The issues that are fixed could well be what I am seeing here.

[Tim Curtis]

The commits for that driver were made on Nov 25, 2018. The kernel 4.19.97 used in moOde 6.4.2 was released on Jan 22, 2020 so unless the commits were reverted at some point they would be in this kernel. You can browse the files for 4.19.97 below.
https://github.com/Hexxeh/rpi-firmware/c...cd05ccbc19