[pvrusb2] pvrusb2-mci-20070114 radio support - stereo?

[Pantelis]

> Thanks for the reply. I don't quite understand the possibilities of a
> hardware limitation, because:
>
> * On the P450 laptop running 2.6.15, I can get stereo FM from the
> PVR-USB2. I can yank the USB cord, plug it into the 2.6.18 desktop, and
> it's mono. I can move the cable back to the laptop, and it's still
> stereo. Therefore, the PVR-USB2 29032 device itself clearly handles the
> stereo.
>
> Where could you see a hardware limitation being?

So you have a 29xxx device after all, sorry I thought you had a 24xxx device.
That rules out the conflict possibilities. Also this means we actually have
the same setup, so I maybe can help more :-)

When you say that you move the cord from one computer to the other,
are they both tuned to the *same* station? are they so close that it is
almost 100% sure they are getting the same signal?

I 'm only saying that even on windows I 've seen it sometimes *reporting*
to be stereo but the sound being mono + hisss. This may indicate a hardware
limitation where the support chip detects stereo incorrectly.

>
> Some things I could try:
>
> * remove the PVR-150 from the desktop
> * upgrade the laptop to 2.6.18-3 to match the desktop
> * I've been meaning to plug something into the PVR-USB2 Audio Input
> jacks. Would that be a useful test?
>
> If the cx25840 module is different and applying a stereo threshold, that
> might make sense. However, wouldn't the threshold refuse to set the
> tuner to stereo? For it to allow the tuner to process the stereo signal
> but mix it to mono would be strange. If it's unintentional (ie a bug),
> then anything goes.

forget the cx25840 module, if you have a 29xxx device you don 't have it.
You have an msp34xxx instead. This one certainly does have a stereo_threshold
option.

What you can try:
Have the two computers close enough, so that whether you connect to one
or the other, the antenna stays on the same place.

Tune to *exactly* the same station. If you still hear mono+hiss to one while
it is stereo in the other, then play with stereo_threshold to see if that 
changes anything.

If nothing of the above works, then it is probably a regression in the
msp34xxx module, but I 'm not sure how much it has changed from 2.6.15
and whether there are relevant changes. If you are an experience linux user
(i.e., feel comfortable with git) you could look for stereo-relevant commits
and try to reverse at will. Maybe you could even do that with the v4l-dvb
repository using mercurial.

I could do this myself, but it takes a lot of time and there are much bigger
fish to fry right now (both on and off pvrusb2 development), sorry :-(

You could also try to talk to the author of the msp34xxx code or your
tuner code to see if your problem rings any bells.

Just my 0.02c,
I could be totally wrong, that won 't be the first time ;-)
-Pantelis