On 10/13/10 20:01, Luis Garrido wrote: > On Wed, Oct 13, 2010 at 2:02 PM, Robin Gareus <ro...@gareus.org> wrote: >> With the USB2 hub I also get >> ALSA urb.c:856: cannot submit datapipe for urb 0, error -28: not enough >> bandwidth >> ALSA midi.c:214: urb status -32 >> messages and jackd won't start. > > Does the hub cause this problem also through a regular USB port (not PCMCIA)?
yes it does. >> but connecting the UA-25 directly to the PCMCIA/USB2 card works if I >> also connect some external power supply to the PCMCIA card. Did you try >> that? > > Believe it or not, that combination hadn't crossed my mind yet: using > the brand new shiny and otherwise useless hub as a 16 € glorified 5V > regulator between its power supply and the PCMCIA card. Why the heck > not, eh? One fewer layer, too. > > Well, there's my workaround, thanks Mr. Gareus! But (it couldn't have > been that easy, could it?) only works on one of the two USB/PCMCIA > ports. Weird. Well, just use the one that works :-p > Trying to use the other one gets me this nice message: > > ALSA sound/usb/usbaudio.c:1348: 40:1:1: usb_set_interface failed > > Bleh. > > Anyone interested in a moderately used Edirol UA-25 interface with > some occasional -74 dB noise at 13.1 kHz, and a penchant for serial > (bus) killing? what are you going to do? get a Firewire device? > Anyway, I think this hub issue is worth looking into. Old USB 1.1 hubs > don't mess with realtime audio and new USB 2.0 do? It looks like (in my case) the hub is actually translating from USB2 to USB1.1 and thereby causes some overhead. If I connect both an external HDD and a cheap (usb1.1) mouse to the hub (which is connected to the USB port of my laptop) the HDD still gets max throughput (>30MB/s). > Now that's not > exactly what I'd call an improvement. What can we expect from USB 3.0? > > Cheers, > > L _______________________________________________ Linux-audio-dev mailing list Linux-audio-dev@lists.linuxaudio.org http://lists.linuxaudio.org/listinfo/linux-audio-dev