On Fri, 2008-11-28 at 04:36 +0100, . . wrote: > Andy, > > I tried this and still got no audio: > v4l2-ctl --device=/dev/video0 --set-input=1 > sleep 5 > v4l2-ctl --device=/dev/video0 --set-ctrl=audio_mute=0 > sleep 5 > v4l2-ctl --device=/dev/video0 --set-input=0 > sleep 5 > v4l2-ctl --device=/dev/video0 --set-ctrl=audio_mute=0
> I am splitting my signal 5 ways with the aid of a signal booster, but > when I directly plug in the originating source feed coax cable I still > get no audio. OK, so there's something going other than the audio standard detection microcontroller being picky and signal strength problems. > The only 2 things that have changed are: > 1) Switched the PVR-150 with a new HVR-1600 both had FM and no IR > 2) Ran another coax cable from an empty terminal on my splitter to the > HVR-1600, one was already run because of the PVR-150 that was removed > > I kind of thought if something turned out to not work it would be with > the new HVR-1600 (/dev/video1) not the HVR-1600 that was already in > the system that was already working well (/dev/video0). My suspicion is PCI bus errors that are going undetected by the driver. My recent changes to the driver did 2 things WRT this: eliminated PCI read retries for simple reads (they never helped fix things AFACIT) and added functions named cx18_write_*expect() which does retries conditioned on a mask. I think that those *expect() functions may be missing PCI readback errors if the value they are expecting is all 1's. I need to inspect these and fix them if that is going on. In the meantime could you leave both cards in the machine and check for audio using one card at a time: # modprobe -r cx18 # modprobe cx18 cards=0,-1 (test card) # modprobe -r cx18 # modprobe cx18 cards=-1,0 (test card) If they both work individually, this would be further indication of PCI bus errors being the problem. My only other guess would be something changed with the tuner driver. My HVR-1600 unit with the TCL M2523_5N_E tuner is on my other computer downstairs - I'll try to do some testing with that. > These logs were made with 2 instances of mplayer running and both tuners > tuned to the same > channel. > > Thanks! Ah, rats. I'm sorry, we have different versions of v4l2-dbg. Can you repeat your data collection with explicit register ranges this time? # v4l2-dbg -d /dev/video0 -R type=host,min=0x2c40000,max=0x2c409c7 # v4l2-dbg -d /dev/video1 -R type=host,min=0x2c40000,max=0x2c409c7 Regards, Andy _______________________________________________ ivtv-users mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-users
