[alsa-devel] issues with usb-audio and maudio quattro
Hi,
It seems that some changes were made to the usb-audio driver over the years that have resulted in partially broken behaviour with maudio quattro.
We have run some tests on two different devices with different machines and found the following items:
1: There are no longer three pcm devices - this breaks previously working asoundrc files 2: pcm0 and pcm1 have different behaviours - this results in audio input working for pcm0 but white noise in pcm1 (probably due to endianess le vs be) 3: We can record 4 tracks with arecord. Tracks 1/2 have audio and track 3/4 have white noise 4: We can no longer run 4 playback tracks in JACK2 as it complains about the ports not matching (this might be a seperate issue with JACK2 and route/plug configs for multichannel devices but that would probably affect the delta 10/10, RME Hammerfall or other multichannel devices too.)
jackd: ../common/JackGraphManager.cpp:44: void Jack::JackGraphManager::AssertPort(jack_port_id_t): Assertion `port_index < fPortMax' failed.
So the main question is who made the decision to drop/consolidate the third pcm and break previously working asoundrc files?
There must be a good reason in which case maybe we can get things working again relatively easily?
Cheers
-- Patrick Shirkey Boost Hardware Ltd
At Sun, 8 Dec 2013 01:47:19 +1100 (EST), Patrick Shirkey wrote:
Hi,
It seems that some changes were made to the usb-audio driver over the years that have resulted in partially broken behaviour with maudio quattro.
We have run some tests on two different devices with different machines and found the following items:
1: There are no longer three pcm devices - this breaks previously working asoundrc files 2: pcm0 and pcm1 have different behaviours - this results in audio input working for pcm0 but white noise in pcm1 (probably due to endianess le vs be) 3: We can record 4 tracks with arecord. Tracks 1/2 have audio and track 3/4 have white noise 4: We can no longer run 4 playback tracks in JACK2 as it complains about the ports not matching (this might be a seperate issue with JACK2 and route/plug configs for multichannel devices but that would probably affect the delta 10/10, RME Hammerfall or other multichannel devices too.)
jackd: ../common/JackGraphManager.cpp:44: void Jack::JackGraphManager::AssertPort(jack_port_id_t): Assertion `port_index < fPortMax' failed.
So the main question is who made the decision to drop/consolidate the third pcm and break previously working asoundrc files?
There must be a good reason in which case maybe we can get things working again relatively easily?
I don't know of such intentional breakage. Bisection please.
Takashi
On Sun, Dec 08, 2013 at 10:21:20AM +0100, Takashi Iwai wrote:
At Sun, 8 Dec 2013 01:47:19 +1100 (EST), Patrick Shirkey wrote:
Hi,
It seems that some changes were made to the usb-audio driver over the years that have resulted in partially broken behaviour with maudio quattro.
We have run some tests on two different devices with different machines and found the following items:
1: There are no longer three pcm devices - this breaks previously working asoundrc files 2: pcm0 and pcm1 have different behaviours - this results in audio input working for pcm0 but white noise in pcm1 (probably due to endianess le vs be) 3: We can record 4 tracks with arecord. Tracks 1/2 have audio and track 3/4 have white noise 4: We can no longer run 4 playback tracks in JACK2 as it complains about the ports not matching (this might be a seperate issue with JACK2 and route/plug configs for multichannel devices but that would probably affect the delta 10/10, RME Hammerfall or other multichannel devices too.)
jackd: ../common/JackGraphManager.cpp:44: void Jack::JackGraphManager::AssertPort(jack_port_id_t): Assertion `port_index < fPortMax' failed.
I'm not sure what this warning is supposed to tell us.
So the main question is who made the decision to drop/consolidate the third pcm and break previously working asoundrc files?
There must be a good reason in which case maybe we can get things working again relatively easily?
I don't know of such intentional breakage. Bisection please.
Yes, please. Also, please describe how the audio interface was exposed before to userspace, and how it is now. The output of 'lsusb -v' would also help.
Thanks, Daniel
participants (3)
-
Daniel Mack -
Patrick Shirkey -
Takashi Iwai