Op zaterdag 22 januari 2005 12:07, schreef William: > AFAIF deinterlacing is only done to the screen. The file is not altered so > your file is stored interlaced.
Yes, in the bobdeint case, but no in general. All the deinterlace filters combine two consecutive frames in one so they halve the FPS (50 -> 25 on PAL). The special thing about bob (and weave) deinterlace is that it doesn't try to assemble a new image but just stores the first frame at the top and the second one at the bottom. The display algorithm then should fetch the top half, write it over the correct (odd or even) lines in the buffer of the last assembled full-height frame and then does the same with the bottom half, effectively doubling the framerate again. Obviously the last part isn't automaticly done when you put bobdeint in the per channel recording filters. This is probably done to not need to encode twice as much frames. The problem is, MythTV does not detect that my old recordings are deinterlaced already, so I can't move the deinterlacer to the playback filter chain. I've tried looking it up in the code but I couldn't find how and where it reads the interlace/progressive-scan state from NUV files (aka "the stream"). With progressive scan output you should be able to select a deinterlace filter in the frontend and have MythTV find out if it need to apply it to this particular recording or not. Like I said "Is it possible to edit my existing nuv files so Myth knows it's deinterlaced already (aka. no guesses)?" Henk Poley <>< PS: I'm sorry if this boils down to my question some months ago about where and how MythTV stores the info about if a recording is interlaced or not. If I remember correctly it was seen as a bogus question back then.
_______________________________________________ mythtv-users mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
