[LAU] Motu M64 Madi interface (not) class compliant?

Fernando Lopez-Lezcano nando at ccrma.Stanford.EDU
Mon Apr 19 19:59:08 CEST 2021


On 4/18/21 12:09 AM, Peter P. wrote:
> * Paul Davis <paul at linuxaudiosystems.com> [2021-04-18 03:43]:
>> On Sat, Apr 17, 2021 at 4:22 PM Peter P. <peterparker at fastmail.com> wrote:
>>
>>> * Peter P. <peterparker at fastmail.com> [2021-03-12 18:43]:
>>>> Hi list,
>>>>
>>>> can report that the Motu M64 Madi interface does work with a limited
>>>> channel count of 24 for input and another 24 for output (instead of 64
>>>> each). So if you are coming across a venue or employer that already has
>>>> one, you are good to go. Sample rate needs to be set on the device
>>>> before starting jack. Of course, buying the card for oneself for so much
>>>> money getting only one third of the channels seems useless linux users
>>>> IMHO.
>>>
>>> I can add now that the card does not work reliably in class compliant
>>> mode with external sync clocks. It will randomly switch back and forth
>>> output channels in banks of eight, cause clicks and distortion and
>>> aliasing artifacts.
>>
>> The bank switching is a known problem with other MOTU USB devices, and has
>> been "solved" by using specific firmware releases (e.g. for the MOTU
>> Ultralite AVB) which may not be available anymore or for your specific
>> device.
 >
> Thanks Paul, that is good to know! Does anyone have more info on this
> back switching and associated firmware updates?

(the old firmware versions are very old - they used to allow up to 64 
channels in class compliant mode - and do not run in the newest hardware 
that you can buy today)

There is a VERY long thread about this here:

https://linuxmusicians.com/viewtopic.php?f=6&t=18046&sid=3586a531a296e4134cd4c7418f429395

A lot of information - I was just looking at new pages at the end since 
last time I looked, and there is a way to get a telnet root login into 
the boxes, ha ha - running linux internally of course (I expect this 
will be gone in the next firmware release :-)

Anyway - not really a resolution to the issue in the thread, but lots of 
ideas to try.

I am running a couple of motu interfaces using a patched kernel that 
enables the use of the alternate usb (proprietary) endpoints and that 
works mostly fine (64 channels over usb with current firmware, 16A & 8M 
tested with this). But the patch does not patch over current kernels and 
a LOT has changed in the audio usb driver, I still have to figure out 
where to do things now. Sigh.

-- Fernando


>> I do not believe that the aliasing is caused by bank switching, but by
>> (possibly) an ALSA driver bug. It generally goes away within a few minutes
>> or on device restart.
> But wouldn't that driver bug affect all class compliant interfaces then?


More information about the Linux-audio-user mailing list