1 |
On Thu, Jul 14, 2022 at 4:53 PM Mark Knecht <markknecht@×××××.com> wrote: |
2 |
> |
3 |
> |
4 |
> |
5 |
> |
6 |
> On Thu, Jul 14, 2022, 4:25 PM Wol <antlists@××××××××××××.uk> wrote: |
7 |
> > |
8 |
> > On 14/07/2022 18:42, Mark Knecht wrote: |
9 |
> > > If instead I'm in Linux with a Win10 VM running I can run the same |
10 |
> > > software in the VM, and it will always see the external DSP when first |
11 |
> > > started, but at random times, generally 5-20 seconds but never more |
12 |
than |
13 |
> > > 5 minutes, the software will tell me the USB connection has been |
14 |
> > > interrupted and I am forced in the software to reestablish a |
15 |
connection. |
16 |
> > > I am always able to reconnect but I am so far unable to keep it |
17 |
connected. |
18 |
> > |
19 |
> > Is there a setting in VB to say "take over the USB port"? ISTR something |
20 |
> > of the sort, which would basically let Win10 take over the port and |
21 |
> > drive it, with linux out of the picture. |
22 |
> > |
23 |
> > Don't trust me on this, I don't play with USB and it's ages since I |
24 |
> > stumbled across this, but it's worth taking a look. |
25 |
> > |
26 |
> > Cheers, |
27 |
> > Wol |
28 |
> |
29 |
> |
30 |
> Good questions. I'm not sure with USB. As I understand it there |
31 |
> are options in Virtualbox to pass through complete pieces of hardware. |
32 |
> I've read that people do this with complete graphics cards. In theory |
33 |
> possibly it could be done with USB but my guess is it might be |
34 |
> difficult as most USB controllers are part of the chipset. Still, it's |
35 |
worth |
36 |
> some study. |
37 |
> |
38 |
> As this DSP processor is identified as an audio device I wondered |
39 |
> today if possibly pulseaudio might be trying to grab it. I intend to |
40 |
> look into configuring PA not to touch it. |
41 |
> |
42 |
> There's also (possibly) blacklisting certain USB device IDs. In the |
43 |
> VM I configured it to understand the device ID and hook to it. |
44 |
> Possibly I can tell the Linux USB stack to ignore this device so that |
45 |
> PA or some other part of the system just stays away. |
46 |
> |
47 |
> But being that the DSP device is an guitar amplifier modeler |
48 |
> and my new Tele came back from my guitar tech at lunchtime |
49 |
> I spent the afternoon playing the guitar! Nonproductive but fun! |
50 |
> |
51 |
> Thanks for the ideas, |
52 |
> Mark |
53 |
|
54 |
> |
55 |
More or less solved this morning apparently. We'll see after more time |
56 |
has elapsed. |
57 |
|
58 |
My desktop environment is KDE. MY sound environment is then, by |
59 |
default, pulseaudio. By disabling the DSP processor as a sound |
60 |
device in KDE System Settings I've now been running the VM |
61 |
and support software for about 90 minutes with only 1 disconnection |
62 |
which came the very first time I did something in KDE (started a |
63 |
new chrome window) after starting the control software in the VM. |
64 |
|
65 |
Since then I have had no more disconnections. |
66 |
|
67 |
I will have to see how it does over the rest of the day and weeks |
68 |
ahead but it's certainly acting better right now. |
69 |
|
70 |
Cheers, |
71 |
Mark |