Below is the current log on the tvtime bug report about swapped fields from the saa7134 card. See the note from Tom Zoerner. Sounds like his fix will help us out as well. Hope this helps, -Billy ----- Forwarded message from "SourceForge.net" <noreply@xxxxxxxxxxxxxxx> ----- To: noreply@xxxxxxxxxxxxxxx From: "SourceForge.net" <noreply@xxxxxxxxxxxxxxx> Subject: [Tvtime-devel] [ tvtime-Bugs-694863 ] fields swapped with saa7134-based card Date: Wed, 19 Mar 2003 18:30:24 -0800 Bugs item #694863, was opened at 2003-02-28 16:39 You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=506987&aid=694863&group_id=64301 Category: Capture and tuner related Group: None Status: Open Resolution: None Priority: 7 Submitted By: David Atkinson (psykax) Assigned to: Billy Biggs (vektor) Summary: fields swapped with saa7134-based card Initial Comment: I have a flyvideo 3000 pal-bg card, using the saa7134 driver 20030218 with card=2. I am using tvtime 0.9.7. The picture appears jittery in any bob-based mode, it looks like the odd/even fields are swapped. Is this a driver issue? ---------------------------------------------------------------------- >Comment By: David Atkinson (psykax) Date: 2003-03-20 13:30 Message: Logged In: YES user_id=722807 I can confirm that tomzo's hack/work-around fixes the artifacts. Video in tvtime looks a lot better now (like it should) with this change applied. ---------------------------------------------------------------------- Comment By: Tom Zoerner (tomzo) Date: 2003-03-20 05:22 Message: Logged In: YES user_id=396970 I've had the very same problem with VBI since at least driver version 0.2.2 (the first version I worked with). With a teletext stream swapped field output is very obvious, since you get the bottom half of the text pages before the upper half (and even before the page header, which renders the data stream broken and useless.) I've sent Gerd a work-around last week which solves the problem for VBI (it's already in the snapshot). You could apply the same work-around for video; maybe you could try this hack just to see if swapped fields are causing the artefacts: in saa7134-video.c::saa7134_irq_video_done() change "if ((status & 0x10) == 0x10)" into "if ((status & 0x10) != 0x10)" Note this inversion will only apply to interlaced capturing, not to single fields. This is only a work-around because it does not comply with the definition of the status bits in the data sheet. Of course the spec may be wrong, but this has to be investigated. BTW it would help if someone could name a specific driver version which did not have the swapping. ---------------------------------------------------------------------- Comment By: Billy Biggs (vektor) Date: 2003-03-19 11:14 Message: Logged In: YES user_id=153320 Stian Jordet claims to see this also, but also notes it did not use to be like this. zytta on #livid also claims this behavior, and I believe this person is using the march 13th snapshot. Time to post to video4linux-list.. ---------------------------------------------------------------------- Comment By: David Atkinson (psykax) Date: 2003-03-01 01:48 Message: Logged In: YES user_id=722807 For output of tvtime -v see bug 694870, file tvtime-err1. It's the same machine. ---------------------------------------------------------------------- Comment By: Billy Biggs (vektor) Date: 2003-02-28 23:05 Message: Logged In: YES user_id=153320 It could be a driver issue. I'll ask about it on the video4linux-list but I'd like to see the output of 'tvtime -v' on your machine. Also, while tvtime is running, hit 'd', it will print out some debug stats. Include those in your reply. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=506987&aid=694863&group_id=64301 ------------------------------------------------------- This SF.net email is sponsored by: Does your code think in ink? You could win a Tablet PC. Get a free Tablet PC hat just for playing. What are you waiting for? http://ads.sourceforge.net/cgi-bin/redirect.pl?micr5043en _______________________________________________ Tvtime-devel mailing list Tvtime-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/tvtime-devel ----- End forwarded message -----
