Public Beta Test: Audio Without Compromise - Refining the Future of Volumio on Bookworm

[hifiswede]

Hello, and thanks for replying.

1. the problem is there even with connected dac at boot.
2. I using the native client whatever that is. I also have a golang script that detects server, volumio does not show. link below.
3. cant.
4. not important for me

I installed ubuntu server with latest git versions of upmdcli and mpd, and everything works there.

https://dumpinen.com/gm-xJq9pEtl

[benatreptar]

Hello World !

Fresh install on SD 4.022 → works fine
RPi 4 + DigiOne Signature + Official Touch Screen V2
WiFi and BT disabled as setting (not usefull in my case)

Plugins :

• Touch Display : OK
• SqueezeLite MC : OK
• Radio Paradise : OK

Playing music FLAC from NAS up to 24 bit @192kHz OK

Boot Logs if usefull for community :
http://logs.volumio.org/volumio/IBgJaXi.html

Thanks again and Big Up for all dev teams

Regards

Ben

[itsmaxdd]

Is OTA supposed to work? if not how should i proceed with upgrades?

EDIT:
Resolved changing channel in /dev to Test

[B505]

Successful upgrade via wired LAN from 4.21 to 4.22
(Pi4b 2GB, 32GB SD)

[SimonE]

OTA from 4.021 4.022
Pi4B, SD boot, Topping D10S USB DAC
Log straight after update: http://logs.volumio.org/volumio/Dc2mXrM.html

Tested and working:
Tidal and Tidal Connect playback
Playing local files
Web Radio

Bluetooth minor issues:

• connected via my Android phone and played a track on my phone. No sound
• forgot device and re-paired: still no sound
• played track from my Volumio library - ok and then tried Bluetooth again, this time it worked and continues to work, also after switching sources

Log after testing Bluetooth: http://logs.volumio.org/volumio/PvtyJcQ.html

[rkorell]

Good morning!
Not that sure if I can/should post it here or create a new thread/post …
Today in the morning my Volumio refuses to play any song.
I’m on beta 4.021, Pi5.
Volume can be adjusted, skip song works partly, playing any sound doesn’t work.
All findings in my one and only use-case: Playing a queue in Spotify-Plugin.

Any Idea?
Logfile is here.

Thanks for any hint.

If wrong posted here - please feel free to move to another location.

Warm regards,
Ralf

[nerd]

Hey @SimonE,

Thanks for testing the OTA from 4.021 to 4.022 on your Pi4B with the Topping D10S DAC. Good to hear that TIDAL, local files, and web radio all worked as expected.

Regarding the Bluetooth issue: from the second log (PvtyJcQ.html), the connection is successfully established with your Android phone, and BlueZ reports a proper A2DP sink route:

info: CoreBluetooth::Device connected: F8:8C:21:4F:02:F6
info: bluealsa-aplay started with args: [... --profile-a2dp ... F8:8C:21:4F:02:F6]

However, prior to that, in the first test, the connection is accepted but the audio path is silent, and there’s no indication that bluealsa-aplay launched immediately. This suggests the first A2DP stream wasn’t fully initialized, or the connection was idle.

Once you played a track from Volumio, the audio system woke up, and subsequent Bluetooth audio started to work reliably.

This is a known edge case: if A2DP connection is established before bluealsa-aplay is invoked or before ALSA state is properly initialized, the stream may silently drop.

I’ll flag this as a minor regression in the Bluetooth plugin path: the first A2DP stream after pairing may not auto-route unless another audio source primes the ALSA stack.

Thanks for confirming that once it works, it stays working even across source switches. That narrows the issue to initial activation timing, not persistent routing failure.

Kind Regards,

[nerd]

Hey @rkorell,

Thanks for reporting this. The issue you’re seeing - Spotify queue not playing under 4.021 while other UI functions like volume and track skip still respond - appears to be a plugin-specific failure, likely within the go-librespot backend used by the new 2023 Spotify plugin.

Why This Is a Plugin Issue

• The playback path for Spotify in Volumio is isolated from core MPD playback, meaning it uses its own stream endpoint and queue manager via go-librespot.

• Volume and skip are still handled via Volumio’s general command router, which is why they continue to respond.

• A known class of failures in this plugin can result in:

  • go-librespot crashing or failing to bind audio backend
  • Transport controls working, but no actual audio output
  • No log messages visible unless the daemon logs are inspected directly

Recommended Next Step

To make sure this is addressed by the plugin team, please post your findings in:

• New 2023 Spotify Plugin - Community Thread

Or, if you want to go straight to structured bug reporting:

• GitHub: volumio-plugins-sources-bookworm (Spotify)

Optional Cross-Reference

You can cross-reference this issue in the Bookworm Beta platform thread here, just for visibility:

• Public Beta - Bookworm Compatibility Thread

Let me know if you want help pulling out specific log lines related to the plugin crash - but this should now be routed through the correct channels for resolution.

FYI: @devgianlu, @balbuze

Kind Regards,

[rkorell]

@nerd

To make sure this is addressed by the plugin team, please post your findings in:

• New 2023 Spotify Plugin - Community Thread

Dear @nerd ,
just have done exactly this.
Thanks for advice!

Warm regards,
Ralf

[michel8166]

first flash, v4.022, all okay.
raspberry pi 5 Model B Rev 1.0, (4 gb)
dac: DDDac MK2 4 deck and fifoPiQ7ll.
driver: generic i2s DAC.

installed plugins:
radio paradise, v2.0.16
squeezelite MC v2.0.0.
youtube2, v2.3.3.
system information, v3.010.

only failed to open alsa when I press play from Squeezelite MC to Radio Paradise.
So I have to press stop first and then play again to restart the stream.

system log: http://logs.volumio.org/volumio/rsnYmPs.html

[Axel]

Hi all,
I’m not sure if this has already been discussed, but I’ve noticed that my Raspberry PI5 (Model B Rev 1.0) gets significantly hotter than it does with version 3.832 (85 vs 67°C). If it helps or is necessary, I’m happy to provide detailed system information and log reports. I have compared with V4.021

[Axel]

btw, where to find the changelog?

[nerd]

Hey @Axel,

Are you looking for something else than the beginning of this thread?

Kind Regards,

[nerd]

Hey @michel8166,

Thanks for the detailed report and the log. The symptom matches a straight device-arbitration issue: MPD and Squeezelite are trying to open the same ALSA device at the same time. In your log MPD fails with:

• Failed to open ALSA device “volumio”: Device or resource busy
• problems opening audio device while playing https://stream.radioparadise.com/flacm
  And the process list shows Squeezelite running when this happens:
• /opt/squeezelite/squeezelite -o squeezelite -C 1
  This is why pressing Stop first and then Play works: it releases the handle and MPD can grab it.

What to change

1. Make Squeezelite let go more reliably

• In the Squeezelite MC settings, set Close output when idle to 1 or 2 seconds and enable options to close output on Stop and on Pause if available. Your instance already runs with -C 1, but the explicit close-on-stop option prevents brief races when switching sources fast. Then switch away by powering off the Squeezelite player in LMS or pressing Stop before starting Radio Paradise.

2. Avoid the shared alias

• Your /etc/asound.conf points the default and the name volumio to the I2S DAC. Squeezelite is opening the same target via its -o squeezelite setting, which maps to the same ALSA path in this config. Change Squeezelite MC Output device to hw:sndrpihifiberry,front:CARD=sndrpihifiberry,DEV=0 or sysdefault:sndrpihifiberry so it does not use the volumio alias. This reduces chances of both stacks chasing the same named device at the same instant.

3. One engine at a time

• If you mainly use Radio Paradise through Volumio, set Squeezelite MC to not autostart at boot, or temporarily disable the plugin when you are not using LMS. That guarantees MPD owns the DAC and removes the race entirely.

4. Ignore the mixer warning

• The HifiBerry DAC has no PCM mixer control, so the MPD log line “no such mixer control: PCM” is expected and harmless. It is unrelated to the busy error.

If the busy popup still appears after trying the above, please:

• Reproduce once more and share a fresh log link.
• Also copy the current Squeezelite MC settings, especially Output device and any close-on-stop options.

This is not a core Volumio bug but a plugin arbitration issue: the plugins are not coordinating ALSA device hand-off. The Pi5 hardware, I²S FIFO DAC, and kernel are fine; it is purely an ownership conflict between MPD and squeezelite.

So the right place to follow this up is in the plugin compatibility thread rather than the core system beta one:
Plugin Compatibility for Volumio on Bookworm

Kind Regards,

[nerd]

Hey @Axel,

To compare temperatures between 3.832 and 4.021 properly, we need to line up all factors. Please share your full setup, since differences in hardware or plugins can influence results:

• Pi5 model and memory size (4 GB or 8 GB)
• Cooling solution (official active cooler, passive heatsink, fanless, etc.)
• Power supply type and rating
• Storage medium (microSD, USB SSD, NVMe)
• Connected DAC or HAT (I2S, USB DAC, HDMI)
• Other peripherals (USB hubs, dongles, displays, etc.)
• List of installed plugins and which ones are enabled
• Any changes in config.txt, userconfig.txt, overlays

That way we can rule out differences like NVMe drawing more power than microSD, or plugins keeping CPU busy.

For the actual test, could you try the following steps on both versions (3.832 and 4.021) so we get quantifiable data:

1. From a cold boot, let the system idle for 5 minutes. Record:

   • vcgencmd measure_temp
   • vcgencmd get_throttled
   • cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
   • Log link from http://<volumioIP>/dev

2. Start playback of the same audio file/stream using the same DAC/output. After 10 minutes of continuous playback, record the same values again.

With identical conditions, we can see if the temperature rise is due to kernel, firmware, or workload differences.

Kind Regards,

[michel8166]

Okay, I changed the name. Now internet radio and Squeelite MC play fine with pause/play next, but not with Radio Paradise.
See Squeelite MC settings.

I think the best option is to simply stop and then play again, or disable it when one of the two is not in use.

log: http://logs.volumio.org/volumio/kgZOzWg.html

manual mode

[michel8166]

@nerd @Axel For the temperature: system v4.022.

Fluctuates between 61.5 and 63.7 degrees Celsius when it’s running.
When nothing is running (same)

Pi5 Model B Rev1 memory size (4 GB)
Cooling solution (passive heatsink)
Power supply: Ian Canada LiniaerPi MK2, 5 volts 2.5 amps
Storage medium (no storage)
Connected DAC or HAT (I2S Dddac mk2)
Other peripherals (USB, nothing)
Changes in config.txt, userconfig.txt, overlays (no changes)
List of installed plugins:
radio paradise, v2.0.16
squeezelite MC v2.0.0.
youtube2, v2.3.3.
system information, v3.010.

I’ll let it run for another half hour and post the temperature again.

[michel8166]

Second measurement: System online for 1 hour and 20 minutes, and the temperature rises slightly.
This seems logical to me because I built everything in a closed box.

now max 65.3 Celsius.

[Axel]

Hi & thx Nerd!
No, the changelog there is perfect, I just overlooked it.