I looked at hw_params which shows the input (capture) format of the UMC1820 as 10 channels, S32_LE. The output (playback) was shown as 12 channels, S24_3LE. So i replaced my S24_LE settings with those from hw_params, and they worked fine with the "plughw:UMC1820" device, but CDSP failed to start using the "hw:UMC1820" device. Note that the channel numbers are 10 input and 12 output according to ALSA, yet the UMC1820 is described as having 18 input channels and 20 output channels by it's manufacturer. CDSP works fine with either setting, which seems to make sense since I only used a subset of the inputs and outputs. Still I cannot use the "hw:" device. I saw nothing else that made any sense to me. The hw_params did state that the output device had "interface = 1", whatever that means, and the capture device as having "interface = 2".
Interestingly, when using the new sample formats, the optical interface version which sounds like a dalek using the S24_LE format, sounded even worse with the "correct" formats.
I think I am still somewhat fumbling in the dark, and getting sick of hundreds of trial and error sessions, which usually either end in CDSP running but no sound, or CDSP stopping with a "closed channel" error message.
Interestingly, when using the new sample formats, the optical interface version which sounds like a dalek using the S24_LE format, sounded even worse with the "correct" formats.
I think I am still somewhat fumbling in the dark, and getting sick of hundreds of trial and error sessions, which usually either end in CDSP running but no sound, or CDSP stopping with a "closed channel" error message.
Those multichannel soundcards typically change number of channel by samplerate. For details see the stream0 file in proc.
Just to note - format S24_LE is 32 bit with 24 MSB bits, CDSP S24LE, not 24 bits. 24 bits is S24_3LE, S24LE3 in CDSP.
Also check play/capture using aplay/arecord -v plughw:xxx and see that transformations the plug plugin must do - that's the verbose switch -v. Ask e.g. copilot for details.
Just to note - format S24_LE is 32 bit with 24 MSB bits, CDSP S24LE, not 24 bits. 24 bits is S24_3LE, S24LE3 in CDSP.
Also check play/capture using aplay/arecord -v plughw:xxx and see that transformations the plug plugin must do - that's the verbose switch -v. Ask e.g. copilot for details.
I highly recommend - copy the command output to copilot and ask questions. It works reliably if the AI gets the data to analyze within the prompt and does not have to rely on factual details from its training.
My stream0 in proc is as follows below. I would very much like to know the significance or use of the "interface 1" statement.
BEHRINGER UMC1820 at usb-0000:00:1d.0-1.3, high speed : USB Audio
Playback:
Status: Stop
Interface 1
Altset 1
Format: S24_3LE
Channels: 12
Endpoint: 0x01 (1 OUT) (ASYNC)
Rates: 44100, 48000, 88200, 96000
Data packet interval: 125 us
Bits: 24
Channel map: FL FR FC LFE RL RR FLC FRC RC SL SR TC
Sync Endpoint: 0x81 (1 IN)
Sync EP Interface: 1
Sync EP Altset: 1
Implicit Feedback Mode: No
Capture:
Status: Stop
Interface 2
Altset 1
Format: S32_LE
Channels: 10
Endpoint: 0x82 (2 IN) (ASYNC)
Rates: 44100, 48000, 88200, 96000
Data packet interval: 125 us
Bits: 24
Channel map: FL FR FC LFE RL RR FLC FRC RC SL
BEHRINGER UMC1820 at usb-0000:00:1d.0-1.3, high speed : USB Audio
Playback:
Status: Stop
Interface 1
Altset 1
Format: S24_3LE
Channels: 12
Endpoint: 0x01 (1 OUT) (ASYNC)
Rates: 44100, 48000, 88200, 96000
Data packet interval: 125 us
Bits: 24
Channel map: FL FR FC LFE RL RR FLC FRC RC SL SR TC
Sync Endpoint: 0x81 (1 IN)
Sync EP Interface: 1
Sync EP Altset: 1
Implicit Feedback Mode: No
Capture:
Status: Stop
Interface 2
Altset 1
Format: S32_LE
Channels: 10
Endpoint: 0x82 (2 IN) (ASYNC)
Rates: 44100, 48000, 88200, 96000
Data packet interval: 125 us
Bits: 24
Channel map: FL FR FC LFE RL RR FLC FRC RC SL