RPi4 + CamillaDSP Tutorial

[Nabussan]

Hi Michael, as for the Main 1-2 issue, a factory reset seems to have done the trick. I will check later whether the other issue is resolved, too.

 

[jdubs]

Michael, do you know if its possible to have the volume "reset" to -40 whenever there is a configuration change?

-Jim

 

[mdsimon2]

Michael, do you know if its possible to have the volume "reset" to -40 whenever there is a configuration change?

-Jim

Yeah that is pretty simple, just add the highlighted lines to flirc.py.

I'm debating if I want to make this default behavior. It probably makes sense where you are using config switching to change inputs but would make less sense where you are using config switching to implement different DSP profiles. I think I'll add it to flirc.py but have it commented out.

Michael

 

[jdubs]

Awesome, thank you!

-Jim

 

[somebodyelse]

I'm debating if I want to make this default behavior. It probably makes sense where you are using config switching to change inputs but would make less sense where you are using config switching to implement different DSP profiles. I think I'll add it to flirc.py but have it commented out.

Why not just make it conditional in that section, and add a config variable or two near the start where it's clear people need to edit things?

 

[mdsimon2]

Why not just make it conditional in that section, and add a config variable or two near the start where it's clear people need to edit things?

Good idea, I clearly don’t think like a programmer.

Michael

 

[Holmz]

Good idea, I clearly don’t think like a programmer.

Michael

Scripted or s/w engineer.
(It is not programming per se.)

 

[Wirrunna]

Michael, just noticed https://github.com/HEnquist/camillagui-backend/releases has v1.0.0-rc3 updated frontend with bugfixes and improved plotting.

 

[mdsimon2]

Michael, just noticed https://github.com/HEnquist/camillagui-backend/releases has v1.0.0-rc3 updated frontend with bugfixes and improved plotting.

Yes, I tried it out this evening. Unfortunately my two biggest complaints with the v1 GUI are still there. The logo disappears if you switch in / out of compact view and the level meters are very choppy.

With the new symlink configuration setup installing the new GUI is very easy, just delete the old one, download the new one and unzip it.

Michael

 

[BGV]

I can't set up MPD + CamillaDSP + alsa_cdsp. I want to configure audio output from MPD to CamillaDSP using alsa_cdsp (MPD stereo -> CamillaDSP+alsa_cdsp -> ASUS Xonar D1 out 8 channels). Maybe someone can help me with alsa_cdsp and MPD config. Everything seems to be clear from the description in https://github.com/scripple/alsa_cdsp and https://www.bitlab.nl/page_id=634, but it doesn't work. I did not find a full-fledged example on the web. I am using Intel ATOM D525, Debian11 Linux 5.10.0-14-amd64 #1 SMP Debian 5.10.113-1 server no GUI.
Earlier I configured MPD stereo -> Camilladsp -> ASUSXonar D1 (8 channels), but MPD output is resampled with ALSA to Camilladsp input at 96000Hz.

 

[Blew]

Thanks for the tutorial. I wonder how effective this would be when using a stereo DAC instead of multichannel, as that would eliminate the need to buy a new DAC unit.

Is there a measurable advantage to processing DSP before outputting separate individual subwoofer and stereo channels via a multichannel DAC unit like the M4 compared to processing DSP, outputting the resulting stereo channels via a stereo DAC then adjusting the crossover on the amplifier?

So
Pi4/CamillaDSP -> multichannel DAC -> individual channel outputs (stereo amplifier -> passive speakers, subwoofer RCA -> active subwoofers)
vs
Pi4/CamillaDSP -> stereo DAC -> multichannel amplifier with bass crossover -> active subwoofers & passive speakers

 

[mdsimon2]

Thanks for the tutorial. I wonder how effective this would be when using a stereo DAC instead of multichannel, as that would eliminate the need to buy a new DAC unit.

Is there a measurable advantage to processing DSP before outputting separate individual subwoofer and stereo channels via a multichannel DAC unit like the M4 compared to processing DSP, outputting the resulting stereo channels via a stereo DAC then adjusting the crossover on the amplifier?

So
Pi4/CamillaDSP -> multichannel DAC -> individual channel outputs (stereo amplifier -> passive speakers, subwoofer RCA -> active subwoofers)
vs
Pi4/CamillaDSP -> stereo DAC -> multichannel amplifier with bass crossover -> active subwoofers & passive speakers

A stereo DAC will work, you just lose some of the primary benefits especially if you are using subwoofer.

With a multichannel DAC you can apply individual correction to both the subwoofer and the main speakers, for example high passing the mains or applying a Linkwitz transform to the sub. If you have an amplifier with x-over capability you may be able to high pass your mains but I bet it is a lot less flexible than CamillaDSP in terms of frequency / slope / Q.

Michael

 

[Blew]

A stereo DAC will work, you just lose some of the primary benefits especially if you are using subwoofer.

With a multichannel DAC you can apply individual correction to both the subwoofer and the main speakers, for example high passing the mains or applying a Linkwitz transform to the sub. If you have an amplifier with x-over capability you may be able to high pass your mains but I bet it is a lot less flexible than CamillaDSP in terms of frequency / slope / Q.

Michael

Thanks, multichannel definitely sounds like the way to go then. My amp has a configurable crossover for high pass filter etc but doesn't have phase or anything else unless you disable pure direct, in which case it's noisy.

Do you know if there is anything that compares to the MOTU M4 in SINAD and price? I really want a coax or optical digital input, which the M4 lacks, and with the M8 being about twice the price of the M4 I may as well buy a MiniDSP Flex at around the same price.

 

[mdsimon2]

Thanks, multichannel definitely sounds like the way to go then. My amp has a configurable crossover for high pass filter etc but doesn't have phase or anything else unless you disable pure direct, in which case it's noisy.

Do you know if there is anything that compares to the MOTU M4 in SINAD and price? I really want a coax or optical digital input, which the M4 lacks, and with the M8 being about twice the price of the M4 I may as well buy a MiniDSP Flex at around the same price.

I am not aware of another 4 channel DAC that competes with the MOTU M4. There are options with SPDIF inputs that are similar in price (Behringer UMC1820 comes to mind) but they will be quite a bit noisier than the M4.

It is possible to add a SPDIF input with something like a miniDSP USBstreamer or a Hifiberry Digi+ I/O HAT but at that point you are getting closer in price to the Flex. For 4 channels the Flex is probably the best option unless you have a specific need that CamillaDSP does better like lots of FIR taps or volume control with dynamic loudness.

Michael

 

[igfarm]

Is anyone using librespot (Spotify) on this setup? I have a problem where the volume is super coarse when I connecting from iPhone. I start at 25% and that is already loud a couple of clicks down and is muted, a couple of clicks up and its very loud, and the volume bar on iPhone is not even half way.

Here is the command looks like to start the library:

/opt/librespot/librespot \
  --name "Living Room" \
  --backend alsa \
  --bitrate 320 \
  --device plughw:0 \
  --mixer softvol \
  --volume-ctrl linear \
  --initial-volume 25 \
  --onevent /opt/librespot/onevent

In the other hand, controlling the volume via Roon works as expected.

 

[mdsimon2]

Is anyone using librespot (Spotify) on this setup? I have a problem where the volume is super coarse when I connecting from iPhone. I start at 25% and that is already loud a couple of clicks down and is muted, a couple of clicks up and its very loud, and the volume bar on iPhone is not even half way.

Here is the command looks like to start the library:

/opt/librespot/librespot \
  --name Living Room \
  --backend alsa \
  --bitrate 320 \
  --device plughw:0 \
  --mixer softvol \
  --volume-ctrl linear \
  --initial-volume 25 \
  --onevent /opt/librespot/onevent

In the other hand, controlling the volume via Roon works as expected.

I have not but I am glad to hear that someone is using librespot as a few folks have asked me about it.

Are you saying that the CamillaDSP volume control is too coarse? Or are you using a volume control in librespot?

Michael

 

[igfarm]

I have not but I am glad to hear that someone is using librespot as a few folks have asked me about it.

Are you saying that the CamillaDSP volume control is too coarse? Or are you using a volume control in librespot?

Michael

This does not seem to be a CamillaDSP specific issue, as it seems to be working properly with Roon, which I have configured to do volume control in software (i.e. I am not using the CamillaDSP volume controls).

When using Spotify on my phone (or Desktop app) and connecting to the Pi running CamillaDSP and librespot, that when the Spotify volume controls seem to be incorrect. I assume there is something amiss with librespot configurations.

 

[mdsimon2]

This does not seem to be a CamillaDSP specific issue, as it seems to be working properly with Roon, which I have configured to do volume control in software (i.e. I am not using the CamillaDSP volume controls).

When using Spotify on my phone (or Desktop app) and connecting to the Pi running CamillaDSP and librespot, that when the Spotify volume controls seem to be incorrect. I assume there is something amiss with librespot configurations.

Just a guess but I imagine you want to set the volume control to log or cubic instead of linear.

Michael

 

[igfarm]

Yes, that seems to have done it. New setup is like this:

/opt/librespot/librespot \
 --name "Living Room" \
 --backend alsa \
 --bitrate 320 \
 --device plughw:0 \
 --mixer softvol \
 --volume-ctrl log \
 --volume-range 80 \
 --initial-volume 50 \
 --onevent /opt/librespot/onevent

Changed volume-ctrl to log as suggested. Also update volume-range to 80 to match the Roon -80db low volume set up there and initial-volume to 50, which is half way which now seems like a sensible value.

Thanks Michael for your ever preset help on this thread.

 

[igfarm]

On a side note: why Spotify? The kids all use it and now it is safe for them to use the main sound system