On Tue, 2009-04-21 at 17:00 -0400, faginbagin wrote: > Hi Andy, > > I think I have good news regarding the mono out test. And I'd really like to > know if it would be a total waste of my time to test my HVR-1600 in the old > computer using WinXP and Linux with the latest drivers. If not, it's > something I > can do before I hand the machine over to my sister this weekend. >
> Setting to mono: > > # v4l2-dbg -c host0 -g 0x2c72014 > ioctl: VIDIOC_DBG_G_REGISTER > Register 0x02c72014 = 320h (800d 00000011 00100000b) > > # v4l2-dbg -c host0 -s 0x2c72014 0xb05 > Register 0x02c72014 set to 0xb05 > (Although I wonder if the value should have been 0xb00 since the > original register value was 0x320, not 0x325.) > > # v4l2-dbg -c host0 -g 0x2c72014 > ioctl: VIDIOC_DBG_G_REGISTER > Register 0x02c72014 = 305h (773d 00000011 00000101b) > > Results: > > No noticeable static on any channel my TV can tune! This is great news to > me. Good. > It means I might be able to work around the problem, meaning I'll have > a usable myth system and a system that I can also use to contribute some > testing on VBI, for example. You said that if the static went away, you'd > know it's a CX23418 digitizer front-end problem. Yes. You should test the patches at http://linuxtv.org/hg/~awalls/cx18-init-debug I had these patches in place for symptoms of "no audio". I didn't occur to me that they might help you, until you confired tuner baseband audio out was static free. > I'm not quite sure what > that means. Does it mean this particular card might be sub-par and that > perhaps I should exchange it for another? If so, should I try the older > model 1178? There's a remote chance. If you test the card in a Windows machine and audio works fine, it is not a bad card. More likely it's the Audio Standard autodetection microcontroller firmware that isn't getting uploaded properly. Testing with the cx18-init-debug repo will eliminate that possibly as it does a complete readback of the firmware after it's uploaded and will gripe if it's wrong. It could also be that the the Automatic Gain Control for SIF audio in the CX23418 front end is being driven by the video signal level instead of the SIF signal level. I'll have to check to make sure the driver's doing the right thing there. > > BTW, this test was done going through the inline amp to 8 way to 4 way > splitters. My earlier tests convince me that they are not the root cause > and I don't need to hurry up to the attic to bypass them. I'll still need > the 4 way no matter what. Agree. This is not an input signal problem. > There was one small issue. The sound was a lot lower with this configuration. Yes. I had you use a baseband analog audio output that comes straight out of the tuner, to a CS5345 digitizer/mux chip, to the first I2S (serial digital audio) port on the CX23418. The CS5345 has a volume control that defaults to 0 dB (a multiplier of 1.0) that the driver provides no way to change. The volume level recorded in the MPEG stream is at the same level as it comes out of the tuner - not very loud. > Restoring stereo: > # v4l2-dbg -c host0 -s 0x2c72014 0xb20 > (not 0xb25, to restore original value) > Register 0x02c72014 set to 0xb20 > > # v4l2-dbg -c host0 -g 0x2c72014 > ioctl: VIDIOC_DBG_G_REGISTER > Register 0x02c72014 = 320h (800d 00000011 00100000b) > > So what are bits 0 and 2 to that register? Should I try setting the value > to 0xb25 and seeing if that allows me to get stereo without static? No. I have almost no documentation on that register. I can tell you about bits 4-5 as the driver manipulates those in cx18-audio.c: 00 - I2S Input 1 audio samples are routed to the MPEG encoder (CX18_AV_AUDIO_SERIAL1 in cx18-cards.c) 01 - I2S Input 2 audio samples are routed to the MPEG encoder (CX18_AV_AUDIO_SERIAL2 in cx18-cards.c) 10 - Audio samples from the SIF audio/broadcast decoder are routed to the MPEG encoder. (CX18_AV_AUDIOn in cx18-cards.c) As for the rest of the bits, from experimentation I found that the CX23418 will change the low nibble all on its own and that messing with the low nibble will screw up good audio. > I understand (at least the bit about life being busy). And let me take this > opportunity to say I really appreciate just how responsive you've been to my > particular issue. You're welcome. > One thing I'm unclear about is whether you now think we have > similar root causes. No, we don't. My problem may have a root cause in analog tuner configuration. I'll figure it out when it really bugs me, but I rarely watch channel 32. > If not, is there a chance that my problem is something that > can, or should, be addressed in software, either kernel or user space. IOW, > might there be an opportunity for me to help test a patch for this issue? Yes. I have high confidence that we should be able to fix your problem with software. > > I also need to look at your other email, and ask some questions about > > the cable box behavior. > > Whenever life's not too busy <vbg>. Although I would like to know if it would > be > a total waste of my time to test my HVR-1600 in the old computer using WinXP > and > Linux with the latest drivers. If not, it's something I can do before I hand > the > machine over to my sister.\ this weekend. Just to conclusively eliminate the possibility of a defective card by testing under Windows XP. I suspect the card is *not* defective. -Andy _______________________________________________ ivtv-users mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-users
