I haven't read every post in this thread, so maybe it's been mentioned already: the CX23100 side of frames 12, 22, 24, 25, 54 of UnProcessedSample.avi look clear. The interference varies and sometimes only covers parts of the frames. Hmm.

It has to do with the way they digitize and decode the video signal. I'm not sure I understand it well enough myself to explain it, but I've posted some screenshots here to demonstrate such differences.

The RF noise isn't entirely consistent and is difficult to see on some frames. I don't see it at all with the other card, which has less warping as well, with bad scanline sync still visible but not as severe. If the herringbone was coming from the vcr it should be visible on both cards.

Assuming the BT878 isn't low-pass filtering its composite input.

I thought the BT878 used s-video, the other card used composite. Maybe I'm confusing this with so many similar posts lately.

Came across this older post again today: http://forum.videohelp.com/threads/3...=1#post2359552

The lines that user was getting don't look precisely the same, but it's interesting because he only got them on a Philips SAA7134-based card. His BT848 card in the same PCI slot was fine. Whereas the BT878 is what's being problematic in this thread...