On Sun, Dec 21, 2008 at 8:20 AM, Al McIntosh <[email protected]> wrote:
> > > On Sat, Dec 20, 2008 at 10:02 PM, Andy Walls <[email protected]> wrote: > >> On Sat, 2008-12-20 at 19:19 -0600, Mark Jenks wrote: >> > On Sat, Dec 20, 2008 at 6:46 PM, Al McIntosh <[email protected]> >> > wrote: >> > Oh, wow! Okay, I was wondering what kind of testing or data >> > would be >> > helpful but sounds as though it's all covered. Impressive >> > work Andy. >> > >> > Al >> >> Al, >> >> Don't thank me, thank Mike and Jeff for the reporting and debugging. >> The "hard" part for me was building a spreadsheet to compute the correct >> PLL parameter values - that was pretty mechanical. >> >> > > Jeff Campbell and Mike Bradley have been doing extensive (!) >> > > investigation and pointed out problem areas too me. Due to >> > their >> > > prompting and what they've found, I've got some fixes in >> > progress this >> > > weekend that should be good enough for the average user. It >> > should make >> > > SVideo and CVBS watchable with buffering. >> > > >> > > The problems are mostly audio and clock related. Maybe by >> > Sunday night >> > > I'll have something checked into my v4l-dvb repo. >> > > >> > > Regards, >> > > Andy >> > > >> > > >> > >> Al >> > >> > Hey Al! >> > >> > I was working on setting up a hvr-1600 today with svideo/audio >> > capturing to mpeg for mythv from a SA4250HD. Everything went great >> > except for the video skipping issue that I could not get rid of. I >> > was very happy to find this thread and I am extremely glad to see that >> > a fix might be right around the corner for this. >> > >> > I would of picked up a pvr-250, if they were easier to find, but I >> > heard that this card would work. (it does, except for that one >> > glitch). >> > >> > I am awaiting directions to try out your patch when it shows up. If >> > you need me to run a test on it, it might take me a few days, but I >> > can get it for you. >> >> Mark and Al, >> >> If you really want to test something, look in >> >> http://linuxtv.org/hg/~awalls/v4l-dvb<http://linuxtv.org/hg/%7Eawalls/v4l-dvb> >> >> the two latest changes should give you something decent for SVideo and >> CVBS. The first analog capture after modprobe will be goofy as always. >> (Still working on that one...) >> >> If you set the stream type to TS instead of the default PS, you may be >> able to run with unbuffered playback Although you may want to set the >> enc_mpg_bufsize=16 parameter, as the default of a 32 kB individual >> buffer sizes may seem a little jumpy if played back unbuffered. >> >> This might be the only changes I get done this weekend, due to holiday >> plans and obligations. >> > > > Andy this is very useable. > > > At first glance, it's a remarkable improvement. I was simply testing with > MythTv "Watch Tv", I watched for several minutes and did not notice any > dropped frames. Typically, frames were dropped almost immediately and > noticeably quite frequently throughout the recording. I'll leave this card > up and running to record from and report back if I notice anything wrong. > However, it's the holiday season and there is not much in the queue to > record over the next 2 weeks. In anycase, since joining this list by the > time you post patches to this list, it typically fixes the issue so I don't > expect to find any thing related. :) > > I do have one observation to point out but will post it in a separate > thread for ease of tracking. > > To everyone involved great work! > > Al > > > > Just noticed, it appears to only record at 480x480. My recording profile for LiveTv in Mythtv is set to record at 720x480 I also tried 704x480. I have attached dmesg and `v4l-dbg --log` out put. I am not sure yet if it's a myth issue or the cx18 driver I'll continue looking. I have noticed the following in mythtvbackend.log:"Not ivtv driver??" at beginning of recordings.
dmesg.log
Description: Binary data
v4l-dbg.log
Description: Binary data
_______________________________________________ ivtv-users mailing list [email protected] http://ivtvdriver.org/mailman/listinfo/ivtv-users
