[mythtv-users] HVR-950q analog/NTSC capture has no sound

[R. G. Newbury]

On 02/10/2017 12:30 AM, James Miller wrote:
> On Thu, 9 Feb 2017, James Miller wrote:
>
>> Now that I've rebooted, I'm bumping into another minor issue. For some
>> reason, every time I reboot and/or power-cycle the machine, only one
>> of the HVR-950q NTSC tuners gets detected and configured: only video0
>> appears under /dev (that's not the one through which I'm trying to
>> record analog audio/video). It's an easy fix: all I have to do is
>> unplug and re-plug the second HVR-950q from/to the usb cable and
>> immediately video1 appears, along with the udev symlink I've created
>> for it, under the /dev directory. I'm not sure why I cannot get the
>> system to detect and initialize both of these tuners on boot/re-boot;
>> perhaps this is an indication that I need to compile snd-usb-audio
>> into the kernel? Or maybe there's a kernel parameter to be specified?
>> For what it's worth I have no initramfs and simply use the kernel's
>> built-in command-line to boot this system. Probably something I'll
>> need to inquire about on the Gentoo forums.
>
> A perhaps parting comment on this new issue, before I head off to seek
> help on what seems some more a strange hardware issue that anything
> directly connected with MythTV. What is really odd here is that the dvb
> tuners in both cards seem to initialize just fine on reboot/power-cycle.
> So far as I can discern, the analog audio component of both cards
> initializes fine as well--at least the udev rules I created for them
> seem to implement normally (arecord -l displays the names I specified
> for each). It is somehow only the analog video component of one of the
> two that fails to initialize properly on reboot/power-cycle. In order
> for that second analog video tuner to properly initialize and appear
> under /dev, an unplug/replug is required. Weird.

It's a little off the wall, but try plugging the two units into USB 
sockets *which are on different hubs*. Say, one on a rear socket and one 
on the front of the box. It is quite possible that the usb sub-system is 
getting confused about the number of units actually available *on one 
hub*, since they obviously have the same usb identifier.
Have you determined whether the *same* unit is always recognized on 
reboot? If that changes it would point to an initialization "lottery" 
condition (which is probably worse than a race condition)

Geoff