Deinterlace on Progressive Scan

[sean@…]

I believe that this point has been brought up before and the ticket was closed because apparently mythtv detects type of scan. Actually I believe it is the detectInterlace function in NuppelVideoPlayer? that does this detection so I hope this can be sorted: apparently the scan is detected as progressive or interlaced and this is meant to deduce whether the deinterlacer is in effect or not.. I have NEVER seen the deinterlacer NOT being used - some channels I watch are progressive and some are interlaced and ALL channels are deinterlaced by whatever deinterlacer I have chosen - this results in flickery results for the progressive channels.

I did a mythbackend -v all ... and grep for Interlace so that all the detectInterlace calls are shown.. some very strange results below... not sure why there are so many calls to detectInterlace and why the results don't make sense (maybe I'm reading the logs incorrectly) but definately the deinterlacer is in effect all the time (note: I have actually disabled the deinterlacer for these logs, I have only given logs to show the detection, let me know if you need more logs)

*TUNE INTERLACED CHANNEL

mpegts_add_stream: at pid 0x21 with type 2 mpegts_add_stream: stream #0, has id 0x21 and codec MPEG2VIDEO, type Video at 0x898b510 2006-01-08 16:21:57.799 detectInterlace(Detect Scan, Detect Scan, 50, 576) ->Progressive Scan 2006-01-08 16:21:57.799 Interlaced: Progressive Scan video_height: 576 fps: 50 Input #0, mpegts, from '/recordings/1010_20060108162153.mpg':

Duration: N/A, start: 9527.499844, bitrate: N/A Stream #0.0[0x21], 1/90000, 50.00 fps(r): Video: mpeg1video, yuv420p, 768x576, 1/50, 1152 kb/s

2006-01-08 16:21:58.152 detectInterlace(Detect Scan, Progressive Scan, 50, 576) ->Progressive Scan 2006-01-08 16:21:58.152 Interlaced: Progressive Scan video_height: 576 fps: 50 Closing PAT Filter: pid=0x0 Closing PMT Filter: pid=0x20

*TUNE PROGRESSIVE CHANNEL

mpegts_add_stream: at pid 0x200 with type 2 mpegts_add_stream: stream #0, has id 0x200 and codec MPEG2VIDEO, type Video at 0x898da50 mpegts_add_stream: at pid 0x28a with type 4 mpegts_add_stream: stream #1, has id 0x28a and codec MP3, type Audio at 0x898d860 mpegts_add_stream: at pid 0x28b with type 129 mpegts_add_stream: stream #2, has id 0x28b and codec AC3, type Audio at 0x897dc10 mpegts_add_stream: at pid 0x240 with type 6 mpegts_add_stream: stream #3, has id 0x240 and codec DVB_VBI, type Data at 0x898c140 [mpegts @ 0xb764b360]Parser not found for Codec Id: 94211 ! 2006-01-08 16:22:30.636 detectInterlace(Detect Scan, Detect Scan, 25, 576) ->Interlaced Scan 2006-01-08 16:22:30.636 Interlaced: Interlaced Scan video_height: 576 fps: 25 Input #0, mpegts, from '/recordings/1010_20060108162154.mpg':

Duration: 00:00:35.7, start: 74557.286856, bitrate: 7750 kb/s Stream #0.0[0x200], 1/90000, 25.00 fps(r): Video: mpeg2video, yuv420p, 704x576, 1/25, 12000 kb/s Stream #0.1[0x28a](eng), 1/90000: Audio: mp2, 48000 Hz, stereo, 256 kb/s Stream #0.2[0x28b](eng), 1/90000: Audio: ac3, 48000 Hz, stereo, 256 kb/s Stream #0.3[0x240], 1/90000: Data: 0x0000

Closing PAT Filter: pid=0x0 Closing PMT Filter: pid=0x20 mpegts_add_stream: at pid 0x21 with type 2 mpegts_add_stream: stream #0, has id 0x21 and codec MPEG2VIDEO, type Video at 0x8989a40 2006-01-08 16:22:40.669 detectInterlace(Detect Scan, Detect Scan, 50, 576) ->Progressive Scan 2006-01-08 16:22:40.669 Interlaced: Progressive Scan video_height: 576 fps: 50 Input #0, mpegts, from '/recordings/1742_20060108162230.mpg':

Duration: N/A, start: 9527.499844, bitrate: N/A Stream #0.0[0x21], 1/90000, 50.00 fps(r): Video: mpeg1video, yuv420p, 768x576, 1/50, 1152 kb/s

2006-01-08 16:22:40.783 detectInterlace(Detect Scan, Progressive Scan, 50, 576) ->Progressive Scan 2006-01-08 16:22:40.783 Interlaced: Progressive Scan video_height: 576 fps: 50 Closing PAT Filter: pid=0x0 Closing PMT Filter: pid=0x20

*TUNE INTERLACED CHANNEL AGAIN

mpegts_add_stream: at pid 0x602 with type 3 mpegts_add_stream: stream #0, has id 0x602 and codec MP3, type Audio at 0x898ebd0 mpegts_add_stream: at pid 0x601 with type 2 mpegts_add_stream: stream #1, has id 0x601 and codec MPEG2VIDEO, type Video at 0x8b4e1a0 mpegts_add_stream: at pid 0x604 with type 6 mpegts_add_stream: stream #2, has id 0x604 and codec DVB_VBI, type Data at 0x898eec0 [mpegts @ 0xb764b360]Parser not found for Codec Id: 94211 ! 2006-01-08 16:23:09.115 detectInterlace(Detect Scan, Detect Scan, 25, 576) ->Interlaced Scan 2006-01-08 16:23:09.116 Interlaced: Interlaced Scan video_height: 576 fps: 25 Input #0, mpegts, from '/recordings/1742_20060108162231.mpg':

Duration: 11:48:24.0, start: 34352.902633, bitrate: 5 kb/s Stream #0.0[0x602](eng), 1/90000: Audio: mp2, 48000 Hz, stereo, 256 kb/s Stream #0.1[0x601], 1/90000, 25.00 fps(r): Video: mpeg2video, yuv420p, 720x576, 1/25, 15000 kb/s Stream #0.2[0x604], 1/90000: Data: 0x0000

Closing PAT Filter: pid=0x0 Closing PMT Filter: pid=0x20 mpegts_add_stream: at pid 0x21 with type 2 mpegts_add_stream: stream #0, has id 0x21 and codec MPEG2VIDEO, type Video at 0x8981900 2006-01-08 16:23:15.106 detectInterlace(Detect Scan, Detect Scan, 50, 576) ->Progressive Scan 2006-01-08 16:23:15.106 Interlaced: Progressive Scan video_height: 576 fps: 50 Input #0, mpegts, from '/recordings/1010_20060108162308.mpg':

Duration: N/A, start: 9527.499844, bitrate: N/A Stream #0.0[0x21], 1/90000, 50.00 fps(r): Video: mpeg1video, yuv420p, 768x576, 1/50, 1152 kb/s

2006-01-08 16:23:15.290 detectInterlace(Detect Scan, Progressive Scan, 50, 576) ->Progressive Scan 2006-01-08 16:23:15.290 Interlaced: Progressive Scan video_height: 576 fps: 50 Closing PAT Filter: pid=0x0 Closing PMT Filter: pid=0x20

[sean@…]

Looked at the code... VERY simple progressive/interlaced detection.. I may have a look and see if there's a better way (unless someone more knowledgeable can do so) ... but what bugs me is that this detection seems to be ignored and it deinterlaces anyway.. The code is very difficult to follow. Has the deint method changed over time as some of the code in VideoOutBase? regarding deinterlace never gets called? .. Will look further into it.. -- Sean

[sean@…]

What was I on last night!? Of course the logs above are screwed.. they're the _backend_ logs.. my apologies, I can't believe I did that!