Ultralite AVB

Talk about your MIDI interfaces, microphones, keyboards...

Moderators: khz, MattKingUSA

User avatar
Be.
Established Member
Posts: 36
Joined: Mon Mar 16, 2015 4:51 am

Re: Ultralite AVB

Postby Be. » Sat Nov 24, 2018 7:53 am

agoode wrote:I am pretty sure this device is extremely sensitive to interruptions in the USB stream.


FWIW, I used mine with a DJTechTools Chroma Cable with a ferrite bead on each end. That did not make a difference.

When I had my Ultralite AVB, sometimes it would go a few weeks without crackling and sometimes it would crackle multiple times per day.

agoode wrote:I found some forum posts that show this happening under Windows over 2 years ago, with audio recordings that sound the same as what I can reproduce:
http://www.motunation.com/forum/viewtop ... 14&t=61286


That is similar to what I heard, but not quite the same. In that recording, the noise cycles in volume. In my case it was at a steady level for a random amount of time, never more than a few minutes.

I found an old forum post that seemed to be talking about similar issues on macOS, but it did not have any recordings attached and I don't remember how I found that. So I'm not sure that these issues never happen with Macs.

Unfortunately it seems that MOTU spent too much effort on fancy features and not enough on getting the basics rock solid. Now that I'm back to using my RME Babyface Pro, I'm glad I don't have to wait a minute for my audio interface to boot up. I think something is quite wrong when an audio interface has a nontrivial boot time. All in all, I recommend staying away from these MOTU interfaces.

Fortunately Sweetwater is awesome and took my returned Ultralite AVB for almost a whole refund (minus a $15 fee) months after I bought it and after the screen broke carrying it in a backpack. :D

Drumfix
Established Member
Posts: 229
Joined: Mon Jan 26, 2009 5:15 pm

Re: Ultralite AVB

Postby Drumfix » Sat Dec 22, 2018 7:15 pm

I did some research today with the 1.3.4-xxx firmware and created a patch (against linux-4.19.5).

- Changing the sample rate will still take about 10 seconds to complete due to the interface completely reorganizing internally.
So, if you are using jack, set the timeout value to 10 seconds.
- In case of an xrun the interface will go silent for a short moment now.
- Crackling and moving around of channels should not happen anymore.

Untar the patches into your <linux-src>/sound/usb directory
and apply them with

Code: Select all

patch < clock.c.patch
patch < endpoint.c.patch
patch < pcm.c.patch


and recompile/reinstall your kernel.

feedback welcome

Drumfix
You do not have the required permissions to view the files attached to this post.

ahellquist
Established Member
Posts: 23
Joined: Mon Jul 01, 2013 12:28 am

Re: Ultralite AVB

Postby ahellquist » Sun Dec 23, 2018 10:34 am

baptiste wrote:Hi !
Newbie here. I own an Ultralite AVB, and was pretty happy about it until this progressive bitcrushing bug on Linux. Before that was the BSOD bug on W10. I contacted MOTU about it and they kindly sent me a beta firmware which resolved the issue. Very confident after that first contact, i was almost happy to report the bitcrushing bug, but got no answer from them, because i mentioned being on Linux i guess.

I'm just here to thank you people because i downgraded to 1.2.9+479 and while i can no longer enjoy the touch interface, my Ultralite AVB is now running smoothly !

FWIW

I have tested all firmwares for the Ultralite AVB including the betas for the touch mixer and one that works pretty well for me is 1.3.4+558
That gives me Touch mixer and no bit crushing drifting. The newer ones are unusable and I have filed a report back to motu of my experiences and the have been friendly but firm that I am running unsupported OS. They claim there is no difference in the handling of sample-sync and timing but for me it is not true.

I have also asked them to bring back configurable "number of audio i/o channels over usb compliant mode" that they removed in the beginning of the firmware life cycle.

Please note that I mainly use my AVB card for low latency live gig running software synts so stability is key but multi tracking is not my regular use case.

In case someone needs a beta or two for troubleshooting, let me know

PS: Great to hear you guys are narrowing down and possibly working around the issue.

/Anders

Drumfix
Established Member
Posts: 229
Joined: Mon Jan 26, 2009 5:15 pm

Re: Ultralite AVB

Postby Drumfix » Sun Jan 13, 2019 10:12 am

Started development of an AVB driver for my intel i210 ethernet controller. Today i was successfully playing/recording my first 8 in/ 8 out AVB stream with my 16A. :D

stanlea
Established Member
Posts: 550
Joined: Wed Apr 25, 2012 9:49 pm

Re: Ultralite AVB

Postby stanlea » Sun Jan 13, 2019 10:28 am

Drumfix wrote:Started development of an AVB driver for my intel i210 ethernet controller. Today i was successfully playing/recording my first 8 in/ 8 out AVB stream with my 16A. :D


Congrats ! Would it work for Ultralite AVB ?

Drumfix
Established Member
Posts: 229
Joined: Mon Jan 26, 2009 5:15 pm

Re: Ultralite AVB

Postby Drumfix » Sun Jan 13, 2019 1:29 pm

It should work work any Motu AVB device.

Musinux
Established Member
Posts: 38
Joined: Sun Jan 14, 2018 8:38 pm

Re: Ultralite AVB

Postby Musinux » Sun Jan 13, 2019 5:10 pm

Drumfix wrote:Started development of an AVB driver for my intel i210 ethernet controller. Today i was successfully playing/recording my first 8 in/ 8 out AVB stream with my 16A. :D


Cool.
I've compiled your patches in already, just have to find the time to update the interface to 1.3+ and try it out.

Any chance that switching the sample rate with AVB works faster (i.e. to play music with different sample rates seamless) ?

Drumfix
Established Member
Posts: 229
Joined: Mon Jan 26, 2009 5:15 pm

Re: Ultralite AVB

Postby Drumfix » Sun Jan 13, 2019 6:00 pm

The delay when switching samplerates seems to be manly due to the internal mixer reinitialization taking so long, so AVB won't help here.
I did some latency measurements now:

reaper at 48000, 48 samples, 2 buffers:

Output to input latency:
Playing a track out analog 1 and recording back through analog in 1 is 78 samples (1.625ms).

Input to output latency through reaper:
Recording into analog in 1 pathing the signal through reaper to ananlog out 1 and recording again through analog 2 in:
The measured offset between the recorded analog in 1 and the recored analog in 2 (i.e. the true live input to output roundtrip latency through reaper) 170 samples (3.54ms).

ahellquist
Established Member
Posts: 23
Joined: Mon Jul 01, 2013 12:28 am

Re: Ultralite AVB

Postby ahellquist » Mon Jan 14, 2019 8:08 am

Drumfix wrote:It should work work any Motu AVB device.

Cool

Haven't checked the code yet and have to order a couple i210 to try this out asap.

Will the AVB implementation work with other vendors AVB streams as well, like for instance Presonus or L-ACOUSTICS stuff or is it specifically targeted towards Motu hardware or is it maybe too soon to ask..

Is there a discovery thing in there and will this work with jack or alsa or am I asking too many questions :)

I get exited since native AVB in linux kernel with some kind of discovery and jack connection would open up for so many cool things, I think.

/Anders

stanlea
Established Member
Posts: 550
Joined: Wed Apr 25, 2012 9:49 pm

Re: Ultralite AVB

Postby stanlea » Mon Jan 14, 2019 11:28 am

Is there a way to connect an i210 to a laptop ?

Musinux
Established Member
Posts: 38
Joined: Sun Jan 14, 2018 8:38 pm

Re: Ultralite AVB

Postby Musinux » Mon Jan 14, 2019 7:27 pm

Drumfix wrote:The delay when switching samplerates seems to be manly due to the internal mixer reinitialization taking so long, so AVB won't help here.

I had the hope that the interface handles it's own settings different when used with AVB. I'm wondering if there's anything in the AVB-spec which deals with sample rate switching, and the time it should be set.


Drumfix wrote:I did some latency measurements now:


Sounds like it's working great, did you do "long-term" testing as well by now?

murray
Established Member
Posts: 5
Joined: Mon Jan 14, 2019 3:51 am

Re: Ultralite AVB

Postby murray » Sat Jan 19, 2019 6:24 am

Drumfix wrote:I did some research today with the 1.3.4-xxx firmware and created a patch (against linux-4.19.5).

- Changing the sample rate will still take about 10 seconds to complete due to the interface completely reorganizing internally.
So, if you are using jack, set the timeout value to 10 seconds.
- In case of an xrun the interface will go silent for a short moment now.
- Crackling and moving around of channels should not happen anymore.

Untar the patches into your <linux-src>/sound/usb directory
and apply them with

Code: Select all

patch < clock.c.patch
patch < endpoint.c.patch
patch < pcm.c.patch


and recompile/reinstall your kernel.

feedback welcome

Drumfix


Hey, thanks for the patch. I've applied them to my older-than-4.19.5 kernel:
Linux cyperus 4.18.9-gentoo #1 SMP PREEMPT Sun Nov 4 11:30:20 PST 2018 x86_64 Intel(R) Core(TM) i7-8850H CPU @ 2.60GHz GenuineIntel GNU/Linux

Some hand-edits were necessary, but I believe I made the appropriate changes ( I've some kernel audio device debugging experience under my belt).
So far so good, but I just got my 16A today (after seeing this thread)--will update here with my continued experience.

Thanks again.

murray
Established Member
Posts: 5
Joined: Mon Jan 14, 2019 3:51 am

Re: Ultralite AVB

Postby murray » Thu Jan 24, 2019 3:37 pm

murray wrote:
Drumfix wrote:moving around of channels should not happen anymore.
Drumfix


Hey, thanks for the patch. I've applied them to my older-than-4.19.5 kernel:
Linux cyperus 4.18.9-gentoo #1 SMP PREEMPT Sun Nov 4 11:30:20 PST 2018 x86_64 Intel(R) Core(TM) i7-8850H CPU @ 2.60GHz GenuineIntel GNU/Linux

Some hand-edits were necessary, but I believe I made the appropriate changes ( I've some kernel audio device debugging experience under my belt).
So far so good, but I just got my 16A today (after seeing this thread)--will update here with my continued experience.

Thanks again.


Hey so looks like my first group of ADAT playback channels and the first eight analog playback channels are flipping pretty consistently after the Motu internals finish initializing and I start hooking things up in JACK. Not sure if this is due to my older-than-4.19.5 kernel or not but I'll upgrade to see if it fixes my issue.

jmccoy555
Posts: 2
Joined: Sat Jan 26, 2019 3:25 pm

Re: Ultralite AVB

Postby jmccoy555 » Sat Jan 26, 2019 4:11 pm

Hi all, finally made the jump from Windows to Linux for my desktop after running a number of headless servers for a number of years (trying KDE neon) and immediately ran into issues with my MOTU 8M running v1.3.4+1422, dropped back to v1.2.9+1280 which solved the audio playback but the web interface is locking up so more investigating to be done..... but;

Drumfix wrote:Started development of an AVB driver for my intel i210 ethernet controller. Today i was successfully playing/recording my first 8 in/ 8 out AVB stream with my 16A. :D


Sounds really exciting!! My motherboard (Supermicro X10SLM) has a onboard i210 so would love to use this!!

Is there any more info about this anywhere?

pemberley
Established Member
Posts: 24
Joined: Sun Jan 13, 2019 11:33 pm

Re: Ultralite AVB

Postby pemberley » Sat Jan 26, 2019 8:06 pm

I started another thread a few days ago, but it might be worth jumping in here as somebody who has applied these patches and still gets channels jumping around. I'm on Manjaro, kernel version 4.19.16, trying to get a MOTU 16A to work. The patches seem to make the audio less crunchy, and maybe make the channel jumps less frequent, but it's still an issue.

Thanks again to Drumfix for his work in creating these patches!


Return to “Computer Related Hardware”

Who is online

Users browsing this forum: mahak123 and 0 guests