I had ever applied your patch one month ago. But I saw very strange result for my BT878A. For example, the original sequence of even field for NTSC in one second was: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 If I set btv->decimation to 10, 5 even fields should be dropped. Therefore the ideal output sequence (25 fields) should look be: 1 2 3 4 5 7 8 9 10 11 13 14 15 16 17 19 20 21 22 23 25 26 27 28 29 Right? BUT, here comes the surprise! The real result I got for my BT878A was: 1 2 3 4 3 7 8 9 10 9 13 14 15 16 15 19 20 21 22 21 25 26 27 28 27 That is, though the number of total even fields is still 25, 5 of them are repeated! I got actually 20 even fields. The worst thing is that you guys could image that what the above sequence looks like. And I had also tried different values of btv->decimation but it just didn't work at all. So, finally I give up the patch. Regards, Kevin ----- Original Message ----- From: "Benoit PAPILLAULT" <bpapillault@xxxxxxxxxxxxx> To: <video4linux-list@xxxxxxxxxx> Sent: Wednesday, June 19, 2002 8:38 PM Subject: Re: many wintv bt878 cards > On Tue, 18 Jun 2002 20:27:37 -0700 (PDT) > Trent Piepho <xyzzy@xxxxxxxxxxxxx> wrote: > > > > Brooktree made a nice databook for the bt848, you should be able to download > > it somewhere. In there it details the temporal decimation registers, which > > let you lower the frame/field rate to whatever you want. > > Available at: > ftp://telepresence.dmem.strath.ac.uk/pub/spec/brooktree/L879_a.pdf > > I have a patch to support temporal decimation in hardware. I can send you if you need > it. However, 2 notes: > > - lots of PCI bus (if not all) are unable to cope with several bt878 cards. (I wonder > if it does not come from the card itself, anyway, that can be solve in software, or maybe > by using the latest PCI patch for your mobo). This problem can "black" lines to appears > in the image (or no image). > > - temporal decimation works OK, but NOT as specified in the bt878 data sheet. I was unable > to choose to decimate on frame/field (it's always FRAME) and which field is first (when I > receive a frame, it's always EVEN field first! But, anyway, it reduces the needed PCI bus > bandwidth. > > Chris: keep us informed on your sucess/failure to use your 10 cards! (I'm curious: > what kind of PCI bus are you using? it's usually only 3 slots). > > Benoit > > > > _______________________________________________ > Video4linux-list mailing list > Video4linux-list@xxxxxxxxxx > https://listman.redhat.com/mailman/listinfo/video4linux-list >
