Re: [LAU] MOTU AVB discussion from LAC

From: Len Ovens <len@email-addr-hidden>
Date: Sun Mar 31 2019 - 17:53:28 EEST

On Sat, 30 Mar 2019, Fernando Lopez-Lezcano wrote:

> On 3/29/19 4:56 PM, Max wrote:
>
>> 3. An endless card aquisition loop between Jack and Pulseaudio caused by
>> the long time the card needs to switch sampling rate.
>> 4. Seemingly erratic behavior, opening the device fails, fails again,
>> again, then works suddenly.

Not really related to just the motu devices but when using both pulse and
jack with any device:
         1) The device jack uses should have it's pulse profile set to
                 "Off"
         2) Pulse uses 44k1 as default SR if Jack uses 48K there will be
                 an SR change switching from one to the other. Setting
                 Pulse's default (and secondary) SR the same as jack
                 might help. (The internal HDA audio seems to be designed
                 around 48k anyway and so probably sounds better there)
         3) If using the pulse-jack bridge, All device profiles in pulse
                 should be turned to "Off". Better is to unload the alsa
                 and udev modules from pulse and use jack as the only
                 pulse device. This will allow jack to freewheel and
                 create less xruns.

My comment about AVB and the motu devices is that the HW in these devices
does not seem to be up to the latency AVB requires. One thing the LAC talk
on the AVB jack backend pointed out to me is the 6 sample buffer size/irq
rate. Can these devices be set to run at 12 samples? Or does that drop the
channel count as well due to packet size? (I wouldn't think so as each
stream is up to 8 channals) From what I have read, the motu devices are
just barely AVB compliant and do not allow all the various AVB rates.
There is one that is aes67 compatable with 1ms packets that may work
better for PC (any OS) to AVB network.

Also, for high channel work on the motu devices, I would suggest no
internal audio processing be used aside from routing. The internal mixer
with eq and effects may be what has caused trouble and decided people
fewer channels was a better choice for the firmware.

Sometimes advertizing gets in the way of usablility.

--
Len Ovens
www.ovenwerks.net
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@lists.linuxaudio.org
https://lists.linuxaudio.org/listinfo/linux-audio-user
Received on Sun Mar 31 20:15:01 2019

This archive was generated by hypermail 2.1.8 : Sun Mar 31 2019 - 20:15:02 EEST