Midi glitching with the new Torso T-1

Hi all,

I think this issue is on the T-1 side (and I’ve reported it to the Torso team), but I’m getting some midi glitches and a kernel error using the new T-1 over USB — reporting it here in case other folks are seeing the same thing.

The configuration for this test case is an E-RM midiclock+ sending clock in over DIN to a Torso T-1, connected via USB on a hub, with no other devices on the hub, to a Nord Drum 2, also connected over DIN. This audio file is an example of the glitching I’m getting — it’s pretty noticeable. When I connect the T-1 directly to the Nord over DIN, there are no glitches and it’s completely regular; likewise when I use my SEQ in the same configuration. There are a fair number of routes defined in my Bomebox and a Midi Translator Pro project is loaded, but the project is just used for aliasing and everything else is unplugged, and the Bome has been rebooted after unplugging stuff.

In the kernel log, I’m seeing constant spamming of the line “urb status -32” when the T-1 is plugged in, which is a broken pipe in the midi driver.

Will update when I get a response from Torso.

glitch-demo.zip (34.8 KB)

Hi, could you share your MT Pro project file? Also maybe some screen shots of device and routing on BomeBox? Is BomeBox connected via WiFi or ethernet or is it just stand alone?

Steve Caldwell
Bome Customer Care


Also available for paid consulting services: bome@sniz.biz

Also, have you tried this using a powered USB hub? Perhaps your T-1 is drawing too much power? How to you have your BomeBox powered?

Steve Caldwell
Bome Customer Care


Also available for paid consulting services: bome@sniz.biz

It’s on a 60W powered hub (and still shows this problem even with nothing else plugged into it). Bomebox is powered from an true-2.5A USB power supply, and also connected via ethernet. Without the T-1 plugged in, everything is stable; with the T-1, everything is unstable. There are a lot of routes and such configured, but it seems cut-and-dry enough that this is mostly an awareness post at this point. @FlorianBome might be interested that this failure mode can disrupt midi routing reliability, but without a bunch of kernel hacking, it feels like it’s unlikely to have a fix.

I’ll let folks know when I hear back from Torso.

Thanks Eleanor, makes sense. We’re also working on a firmware update with an updated kernel and newer USB drivers, which may fix this issue.

Glad to hear it — BTW, this is happening on the Beta that you sent me, still.

If you have a POE adapter handy, you might try that for power instead of the USB power supply. I’ve found that they generally provide more stable power. With that said, maybe a new kernel driver would do the trick. In the meantime, maybe hook up via the T-1 MIDI port to BomeBox.

Steve Caldwell
Bome Customer Care


Also available for paid consulting services: bome@sniz.biz

Just to close this out, the new 1.0.2 update for the T-1 has fixed this issue.

Thanks for keeping us informed!

Steve Caldwell
Bome Customer Care


Also available for paid consulting services: bome@sniz.biz