See my post above, it explains why your setup does not work.
The loopback module creates two connected alsa devices, sharing a common memory buffer. That means no resampling, no sample format change, only one writing pointer and one reading pointer (it is a bit more complicated because each device can setup different buffer/period size).
The first side to open defines the params for both sides. The other side must use the same params if it wants to open its-side device. Change can happen only when both devices are released/closed.
Use the /proc/asound files, they provide key information for alsa troubleshooting.
Actually there is the module parameter pcm_notify and PCM control pcm_notify (available in amixer).
IIUC the module param should allow closing the capture side (with error) when params change on playback side. That would be easy to use in camilladsp - just re-opening the device with correct params.
IIUC the module param should allow closing the capture side (with error) when params change on playback side. That would be easy to use in camilladsp - just re-opening the device with correct params.
set "queuelimit" to 1
Where to set this? in camilladsp ?
I have to figure out howto grab the stout from squeezelite into camilladsp also.
I guess i will have to change th .yml file for that to happend.
Will be back
Jesper.
phofman
I understand this now.
What i do is to close/reload the camilladsp through python, samplerate is set there at output.
That part is working.
The part where i need to stop/reload/close<>open the squeezelite is not working.
As you wrote i need to stop booth devices.
Will go thinking about this
Jesper.
I understand this now.
What i do is to close/reload the camilladsp through python, samplerate is set there at output.
That part is working.
The part where i need to stop/reload/close<>open the squeezelite is not working.
As you wrote i need to stop booth devices.
Will go thinking about this
Jesper.
Good tip! That options seems quite useful. Breaking the capture should make CamillaDSP exit, so then you can just restart with the right config.
Henrik
Someone at slimserver forum (LMS) asked me this :
With more explanation :
This
I hope it's okay to ask you if this could be possible to do what he suggest?
Regarding the other suggestions phofman and you told here (stdout & snd-aloop pcm_notify, sry.. i really havent figured it out yet, and i think the suggestion from the other forum was important)
Jesper.
Someone at slimserver forum (LMS) asked me this :
With more explanation :
This
is just to configure for them using LMS howto transcode (flac etc...)
I hope it's okay to ask you if this could be possible to do what he suggest?
Regarding the other suggestions phofman and you told here (stdout & snd-aloop pcm_notify, sry.. i really havent figured it out yet, and i think the suggestion from the other forum was important)
Jesper.
Hmm this would require big changes to how CamillaDSP handles the configuration internally. It would also need either a set of config files (one for each sample frequency) or a way to autogenerate them.Henrik
Someone at slimserver forum (LMS) asked me this :
With more explanation :
This is just to configure for them using LMS howto transcode (flac etc...)
I hope it's okay to ask you if this could be possible to do what he suggest?
Regarding the other suggestions phofman and you told here (stdout & snd-aloop pcm_notify, sry.. i really havent figured it out yet, and i think the suggestion from the other forum was important)
Jesper.
But I think the same end result can be achieved by using some supervising process, and let that process launch CamillaDSP with the right settings and then feed it data. Since the config is standard YAML it's very easy to autogenerate a valid config file. Python should be more than fast enough for this.
Cool enough Henrik... You descide
Regarding the use of the snd-aloop option pcm_notify. How do i use that?
I can't seem to find much about it.
ilename: /lib/modules/4.19.97-v7l+/kernel/sound/drivers/snd-aloop.ko
license: GPL
description: A loopback soundcard
author: Jaroslav Kysela <perex@perex.cz>
srcversion: 3B8A8D841D62048BC909BB7
depends: snd-pcm,snd
intree: Y
name: snd_aloop
vermagic: 4.19.97-v7l+ SMP mod_unload modversions ARMv7 p2v8
parm: index:Index value for loopback soundcard. (array of int)
parm: id:ID string for loopback soundcard. (array of charp)
parm: enable:Enable this loopback soundcard. (array of bool)
parm: pcm_substreams
parm: pcm_notify:Break capture when PCM format/rate/channels changes. (array of int)
Jesper.
I tried to use the pcm notify param as well the PCM control, to no avail. I sent a question to the mailing list, I believe it will be resolved soon Functionality of pcm_notify in snd-aloop?
Like
I created a file /etc/modprobe.d/ - sndloop.conf
In that file i pasted :
options snd-aloop pcm_notify=1
And after a reboot i got this :
But lack of dokumentation i can't tell if pcm_notify=1 is the right to do.
Jesper.
Hello here.
So i'am playing some with the engine.
So trying the stdio thing as suggested by Henrik some posts ago i modified my rate_44100.yml file into this :
So while trying i started the squeezelite player with output to stdout into a file "rawfile.dat"
Trying to playback like this:
cat rawfile.dat | camilladsp rate_44100.yml
Does not work; i'am pretty sure my .yml file is wrong?
Jesper.
So i'am playing some with the engine.
So trying the stdio thing as suggested by Henrik some posts ago i modified my rate_44100.yml file into this :
So while trying i started the squeezelite player with output to stdout into a file "rawfile.dat"
Trying to playback like this:
cat rawfile.dat | camilladsp rate_44100.yml
Does not work; i'am pretty sure my .yml file is wrong?
Jesper.
So like this :
I must be dumb
...Running squeezelite with output STDOUT, and camilladsp like this :
I cant make it work, camilla just exit's
Jesper.
Working like this :
squeezelite -n XioPlayX -o - -a 32 -z (n=name, -o - =stdout, -a32=buf -z=daemonize)
camilladsp -v /home/pi/Cfilters/rate_44100.yml
But no sound, and squeezelite is consuming like 100% CPU...
also squeezelite acting very unstable like this... not camilladsp's fault
I created an issue on the squeezelite git for this a couble of hours ago allready.
Let's see if there will be any correction on that?
Jesper.