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

[Mike Isely]

On Sat, 20 Jan 2007, Rick Macdonald wrote:

> Mike Isely wrote:
> > Rick:
> >
> > The actual handling of mono vs stereo is up to the tuner core, not the 
> > pvrusb2 driver (that is, once the pvrusb2 driver actually hands the 
> > proper commands to the tuner core which is what this bug fix does).
> >
> > Are you sure you were running the updated pvrusb2 driver?  Look at the 
> > banner it prints in your log when it starts; the snapshot ID will be a 
> > part of that banner.
> >   
> Here's the log from boot-up:
> 
> Jan 19 23:06:18 timshel kernel: Linux video capture interface: v2.00
> Jan 19 23:06:18 timshel kernel: ivtv:  ==================== START INIT 
> IVTV ====================

   [...]

> Jan 19 23:06:18 timshel kernel: ivtv:  ====================  END INIT 
> IVTV  ====================
> 

That dump is a mixture of ivtv and pvrusb2 related messages.  While we 
can pick that apart, it'd be better just to run that 'cat 
/sys/class/pvrusb2/*/debuginfo' command and THEN look at the log.


> > We might be using different versions of the I2C modules with different 
> > behavior.  For my testing last night I was running in kernel 2.6.18.1, 
> > using the V4L core that was a part of that kernel.
> >
> > When I tested, the way I determined I was getting stereo was by flipping 
> > the audio mode back and forth while listening (that control works now).  
> > I could definitely hear the sound "spread" when I switched to stereo 
> > mode.
> >   
> I tried this just now, and the sound does change, but neither is Stereo. 
> When I tell it stereo, the volume drops a bit and seems thinner. It 
> almost as if it was giving me just left or right through both sides, but 
> this is hard to tell until I can catch some music or commercial with 
> definite separation, and that's long enough to switch back and forth a 
> few times.

Well I'd like to hear more about what you think you are hearing.  I can 
tell you that with every stereo FM receiver I have ever used that 
"stereo" usually does sound somewhat lower in volume and "thinner".  I 
am definitely hearing the sound as being more "spread" when stereo is 
enabled.  But being unable to create a controlled test for this, it is 
rather difficult to really know what is going on.

One thing you could try is running the audio through a 3D expander (many 
sound cards have this ability).  If the L and R are really different 
then expanding the audio will emphasize that difference.

> 
> > I have also asked my contact at Hauppauge about this, when there was a 
> > question about whether or not Stereo is even there.  The answer is that 
> > it is definitely there, but I was also warned that some combinations of 
> > tuner types have been "harder" to get working in FM stereo mode.  He 
> > didn't know which ones those were, but a difference like that between 
> > our model types might have something to do with it.  After plugging in 
> > the PVR USB2 device and starting the radio, do this command from the 
> > shell:
> >
> >   cat /sys/class/pvrusb2/*/debuginfo
> >
> > That will dump info to the screen about the associated I2C modules.  
> > Send me that output.  In addition, the system log will get a status dump 
> > from everything (it will be obvious in the log); send that as well.  I 
> > can use that to look for differences in the hardware between what you 
> > are running and what I tested.
> >   
> timshel:~# cat /sys/class/pvrusb2/*/debuginfo
> big lock free; ctl lock free
> driver flags: initialized ok connected
> Subsystems enabled / configured: +ENC_FIRMWARE +ENC_CFG +DIG_RUN 
> +USB_RUN +ENC_RUN
> Subsystems disabled / unconfigured:
> Attached I2C modules:
> tveeprom @ 0x50 [v4l2_standard v4l2_audiomode v4l2_bcsh v4l2_volume 
> v4l2_freq v4l2_size v4l2_log]
> MSP3445G-B8 @ 0x40 (handler: pvrusb2-audio v4l2) [v4l2_standard 
> v4l2_audiomode v4l2_bcsh v4l2_volume v4l2_freq v4l2_size v4l2_log]
> saa7115 @ 0x21 (handler: pvrusb2-video-v4l) [v4l2_standard 
> v4l2_audiomode v4l2_bcsh v4l2_volume v4l2_freq v4l2_size v4l2_log]
> tda9887 @ 0x43 (handler: pvrusb2-tuner) [v4l2_standard v4l2_audiomode 
> v4l2_bcsh v4l2_volume v4l2_freq v4l2_size v4l2_log]
> Philips NTSC MK3 (FM1236MK3 or FM1236/F) @ 0x61 (handler: pvrusb2-tuner) 
> [v4l2_standard v4l2_audiomode v4l2_bcsh v4l2_volume v4l2_freq v4l2_size 
> v4l2_log]
> timshel:~#

Right.  Looks good.  Now look at your system log, AFTER you've run that 
command.  Why?  Because running that command also has the side effect of 
causing all the I2C chip-level drivers to dump their internal status to 
the system log.  That will tell us things like if those driver think 
they are tuning a stereo signal.

> 
> > However, the fact that you got stereo while running under WinXP would 
> > suggest that this isn't the issue.  But it could be a different enough 
> > tuner that the tuner core isn't driving it correctly.  There has been at 
> > least one recent commit into the v4l-dvb repository regarding proper 
> > suport of FM stereo for certain tuner types.  This leads to another 
> > suggestion: Check out and build the latest v4l-dvb repository and test 
> > that.  The pvrusb2 driver is also a part of that repository, and in fact 
> > as of this morning, all the recent changes have been pulled up into it.
> >   
> It might be a day or three before I can look at v4l-dvb (three birthdays 
> in the family this weekend).
> 
> Question: I'm playing the radio with this:  mplayer -vo null /dev/radio0
> and setting the station as root with this:   echo 89300000 > 
> /sys/class/pvrusb2/sn-7570160/ctl_frequency/cur_val
> 
> Is there anything else that works?

You can try kradio to tune the radio (but you will still have to use 
mplayer or some kind of mpeg-consuming app to hear the audio).

> 
> It seemed like it worked OK the other day, but now after 5 or 10 seconds 
> it dumps a bunch of stuff and gets all choppy (after it prints "too SLOW"):

   [...]

Yes, I've seen this as well.  Not sure what is going on.  I know that 
stereo reception is not as robust against noise as mono is.  So if a 
station is marginal, then there might be dropouts when in stereo, and I 
know that if the input to the mpeg encoder breaks up, then the encoder 
will have a hard time with it and cause strange things in the output.  
See if the symptom is influenced by stereo vs mono mode.  (It's also 
possible that some tuning issue in the software is keeping the hardware 
from getting a strong signal.)

  -Mike

-- 
                        |         Mike Isely          |     PGP fingerprint
     Spammers Die!!     |                             | 03 54 43 4D 75 E5 CC 92
                        |   isely @ pobox (dot) com   | 71 16 01 E2 B5 F5 C1 E8
                        |                             |