[FFmpeg-trac] #6279(ffmpeg:new): can't request rgb24 (uvc over dshow, unknown compression type)

FFmpeg trac at avcodec.org
Thu Mar 30 22:51:54 EEST 2017

#6279: can't request rgb24 (uvc over dshow, unknown compression type)
             Reporter:  Djfe    |                     Type:  defect
               Status:  new     |                 Priority:  normal
            Component:  ffmpeg  |                  Version:  git-master
             Keywords:          |               Blocked By:
             Blocking:          |  Reproduced by developer:  0
Analyzed by developer:  0       |
 ffmpeg doesn't seem to know some of the color space id's that dshow uses.
 In particular the rgb24 color space:
 "[dshow @ 0000000000756320]   unknown compression type 0xE436EB7D  min
 s=1920x1080 fps=15 max s=1920x1080 fps=59.9999"
 (I provided a log output, containing more similar lines)

 A quick google search for the id (0xE436EB7D) in question reveals, that it
 represents rgb24:

 I bet ffmpeg doesn't know the other RGB GUIDs either, but I can't verify
 it, since my device only supports outputting in those two color spaces.

 I'm unsure if it is allowed to simply take the information from the linked
 opencv file (regarding the opencv license (vs LGPL)) though.
 But it would be nice if other GUIDs apart from rgb could be "imported"
 into ffmpeg's knowledge as well.

 But now to the interesting part:
 is it enough if ffmpeg knows, that the data is of color space rgb24, or

 until someone finds the time: I'm looking forward to rgb24 support for uvc
 over dshow :) (maybe that will give you some motivation ^^)
 The reason being, that I'm recording to qtrle (lossless, but only rgb
 color space).
 Asking the grabber to give me rgb avoids unneccessary color space
 conversion on both sides.

 I'm unsure if the type of this bug report is a defect or an enhancement,
 feel free to correct it, if you think it isn't a defect ;)

Ticket URL: <https://trac.ffmpeg.org/ticket/6279>
FFmpeg <https://ffmpeg.org>
FFmpeg issue tracker

More information about the FFmpeg-trac mailing list