Open arpena opened 3 weeks ago
Looking at the code it seems that the stream processor is consuming stderr from librespot
to get track name and duration.
I made some tests and messages from librespot
zeroconf negotiation are still being output to stderr, no stdout were the audio should be written. I couldn't find any extra message on stdout.
I remember that without zeroconf, messages from librespot
username/password auth was written on snapserver
logs. Sadly, I cannot reproduce that anymore. So somehow the stream processor is being confused by the delayed authentication that with zeroconf occurs only when the device is selected on the Spotify app.
Darn, I reverted back to the librespot://
source and cannot reproduce the noise. I'll stick to this configuration to test if I get the noise again.
Describe the bug Since Spotify disabled the use of username/password for librespot, I tried to configure a
librespot://
source without username/password parameters to use zeroconf. It works on the Spotify part, but reproduce as noise on Snapcast.Relevant part of
snapserver.conf
:Everything works fine If instead I configure a new fifo on Snapcast server, and run an external librespot using pipe backend.
Steps to Reproduce
librespot://
source without username/password parametersEnvironment details
Attach logfile if applicable
Interestingly, the output from librespot is a lot more verbose when I run it outside of snapcast:
So the issue seems that some INFO messages from
librespot
are consumed as sound data bysnapserver
, also something is changing the format of log messages forlibrespot
UPDATE 16/Aug/2024 Cannot reproduce this anymore. Will stick to the
librespot://
source configuration to see if this happens again