Re: [Mjpeg-users] problems with lavrec and lml33

2003-03-14 Thread Ronald Bultje
Hey Andrew, On Thu, 2003-03-13 at 22:03, Andrew Stevens wrote: Ronald: can you take an Eyeball to the bits marked LEAVE_OUT_FIX in zoran_device.c - I think this fixes an obvious inconsistency in the code... Looks good, probably one of the minor weirdities that need tweaking before we can call

Re: [Mjpeg-users] problems with lavrec and lml33

2003-03-11 Thread Ronald Bultje
Hey Paul, On Sun, 2003-03-09 at 23:08, Paul Forgey wrote: Before I reverted from v4l2, I did try ZORAN_VIDEODEV_2 with the same results. Actually, that might be correct, since I can't find Andrew's changes either. Andrew, did you actually commit them? Ronald -- Ronald Bultje [EMAIL

Re: [Mjpeg-users] problems with lavrec and lml33

2003-03-11 Thread Andrew Stevens
Hi Paul, How old is your LML33 a recent R10 upwards or an oldie like mine. I'm currently working (i.e. its on my list for a free evening when I'm not knackered) on tuning the LML33 video decoder setup. Feel free to feed me your Results - it useful to have different video sources they

Re: [Mjpeg-users] problems with lavrec and lml33

2003-03-09 Thread Ronald Bultje
On Sun, 2003-03-09 at 07:46, Paul Forgey wrote: I've got an lml33 card and am noticing the following problems with it using lavrec: LML33 + zoran-0.9.1 is broken. It'll be fixed in 0.9.2. In the mean time, use the latest 0.8.x release or current CVS (ZORAN_VIDEODEV_2 branch). Ronald --

Re: [Mjpeg-users] problems with lavrec and lml33

2003-03-09 Thread Paul Forgey
Before I reverted from v4l2, I did try ZORAN_VIDEODEV_2 with the same results. 0.8 doesn't compile on 2.4: Making Zoran 36060+36057/67 driver for 2.4.20 kernel... DIR=`pwd`; (cd /usr/src/linux; make SUBDIRS=$DIR modules) make[1]: Entering directory `/usr/src/linux-2.4.20' make -C

[Mjpeg-users] problems with lavrec and lml33

2003-03-08 Thread Paul Forgey
I've got an lml33 card and am noticing the following problems with it using lavrec: 1) despite selecting -iN, it's reading the composite input. -in also reads the composite input, so they aren't reversed. 2) the picture is shifted either to the right or to the left (it's random) by about