[Ffmpeg-devel-irc] ffmpeg-devel.log.20140502

burek burek021 at gmail.com
Sat May 3 02:05:02 CEST 2014


[02:35] <cone-259> ffmpeg.git 03Michael Niedermayer 07master:0f9f24c9cfd2: avfilter/vf_yadif: fix filtering a single image
[02:53] <ShadowJK> michaelni,  did you see the garbled video from SpaceX, and their request for gurus to help try clean it up? :-)
[02:54] <michaelni> ShadowJK, no, where ?
[02:55] <ShadowJK> sec
[02:56] <ShadowJK> http://www.spacex.com/news/2014/04/29/first-stage-landing-video
[02:59] <ShadowJK> you can only really see two moments in that, heh. The "Free fall" part, and then I guess they get an I-frame when the computer decides it's time to fire up engine to slow down.. and hten they land in water and the water destroys data :)
[02:59] <wm4> michaelni: thanks for looking into the yadif thing
[02:59] <michaelni> wm4: np
[02:59] <Timothy_Gu> shadowjk: that was fun ;)
[02:59] <ShadowJK> I'm somewhat wondering if they wouldn't be better off using extremely cheap chinese video cameras that record in mjpeg, for data integrity :)
[03:01] <Timothy_Gu> I don't have a computer to check, but are they using MPEG-2?
[03:02] <ShadowJK> http://www.reddit.com/r/spacex/comments/24bsn2/first_stage_landing_video/
[03:03] <ShadowJK> " MPEG 4 Part 2, P/I 15, 15fps, NTSC, fixed bitrate"
[03:04] <Timothy_Gu> Yeah I would rather use MJPEG for this kind of video.
[03:04] <J_Darnley> What?!
[03:04] <J_Darnley> mpeg4
[03:04] <ShadowJK> "we took a pass on the data to align every MPEG packet on a 188 byte boundary and set the packet start byte to 0x47. Then we identified blocks within keyframes that contain bit errors, and then manually flipping bits in those corrupt blocks to see if it recovers more of the image."
[03:05] <ShadowJK> mpeg4 vs mpeg2?At the time I am sure it made sense to try to use the more effiecient video codec, as they had concerns about the available bandwidth
[03:06] <Timothy_Gu> michaelni: I want to rename a wiki page but then I found out that I don't have the permission.
[03:07] <ShadowJK> "Q: Why is the video so bad? A: This was recorded over a very lossy RF link."
[03:11] <baptiste> http://www.spacex.com/sites/spacex/files/raw.ts
[03:22] <Timothy_Gu> michaelni: see http://trac.edgewall.org/wiki/TracPermissions
[03:24] <michaelni> Timothy_Gu, try again, should work
[03:45] <Timothy_Gu> MichaelNi: it works, thanks!
[06:12] <cone-158> ffmpeg.git 03Matt Oliver 07master:1ccd1a3860b1: Enable opencl wihtout pthreads.
[07:49] <Zeranoe> Is ffmpeg running a shortage of bandwidth? I'm getting ~50KB/s consistently downloading a git clone or snapshot
[10:09] <plepere> kurosu, we haven't noticed any real improvement. it might be because of a bottleneck somewhere else.
[14:42] <ubitux> wbs: just in case, ad8fbddb94c2d7e0fbcd46d508e962bbfd939ba4
[14:45] <Daemon404> confusing message is confusing
[14:48] <wm4> bunch of commit hashes, some point to merge commits
[14:48] <wm4> at this point, cherry-picking really would be easier, I guess
[15:15] <Daemon404> still no gcc 4.8.3 eh
[15:50] <Daemon404> hmm
[15:50] <Daemon404> ffmpeg has no GRAY16A
[15:50] <Daemon404> i.e. gray with alpha
[15:58] <kierank> http://www.spacex.com/news/2014/04/29/first-stage-landing-video
[15:59] <ubitux> yeah it was already shared
[16:02] <Daemon404> needs libpostproc
[16:52] <ubitux> http://aeroquartet.com/wordpress/2014/04/24/why-mxf-format-will-eventually-win/
[16:52] <ubitux> > MXF is an open standard promoted by the video industry. Thats a guarantee that a single vendor cannot and will not play dirty tricks.
[16:52] <ubitux> lol
[17:06] <kierank> ubitux: yeah hillarious
[17:08] <kierank> mateo`: know anyone interested in testing an sxs driver for linux?
[17:11] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:c60b266eac18: avcodec/mpegvideo: support disabling motion compensation
[18:38] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:41a679000f22: avcodec/mpegvideo: fix gray*() parameter types
[19:02] <cone-788> ffmpeg.git 03Lukasz Marek 07master:41e4c6d8c5fb: lavd/opengl_enc: use flag to mark inited context
[19:02] <cone-788> ffmpeg.git 03Lukasz Marek 07master:45601854b148: lavd/opengl_enc: add window size param
[19:02] <cone-788> ffmpeg.git 03Lukasz Marek 07master:db4b03146cc3: lavd/opengl_enc: fix window size correction code
[19:02] <cone-788> ffmpeg.git 03Lukasz Marek 07master:ba52fb11dc63: lavu/opt: add av_opt_set_dict2() function
[19:02] <cone-788> ffmpeg.git 03Lukasz Marek 07master:330d547ef3cc: lavf/mux: pass options to nested structs of priv data
[19:02] <cone-788> ffmpeg.git 03Lukasz Marek 07master:fa4f7b17bc12: lavd/xv: reident after previous commits
[19:02] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:cea9ee5dbd93: Merge remote-tracking branch 'lukaszmluki/master'
[19:13] <mateo`> kierank: nope, it is a driver for camera, right ?
[20:16] <kierank> mateo`: it is a memory card for sony cameras, yes
[20:16] <kierank> has mxf stuff on it
[20:16] <kierank> but I don't know many people who are interested in testing it
[22:59] <cone-788> ffmpeg.git 03Michael Niedermayer 07master:64e448994cb4: avcodec/h263dec: print MB position for each MB when TRACE is enabled
[00:00] --- Sat May  3 2014


More information about the Ffmpeg-devel-irc mailing list