Use different pixel formats for depth and IR 16-bit streams |
|
Issue descriptionThis task is related to discussion started at [1] chfremer@chromium.org "With that, for Z16 and INVZ, it seems like we pretending to clients that the format is Y16 when it is really not. I don't think that is okay. If we require pass-through of these formats, then we should add them to the PIXEL_FORMAT_**** enum to allow the client code to decide what to do with it." Patch [2] adds also D16 format. The importance to tell apart depth pixel format (used for stream formats INVZ, Z16, D16) from infrared (Y16 stream) comes from videoKind API: https://www.w3.org/TR/mediacapture-depth/#def-constraint-videoKind when videoKind constraint is set to "depth", a depth stream track should be returned. I don't see the use case where we need to have different pixel formats within the depth stream format category (D16, INVZ, Z16) - the only difference there is in fourcc code - bitness and interpretation is the same. The plan: rename Y16 to D16. There is no use case yet for infrared stream so it might be better to remove the Y16 stream capture and let the D16, Z16 and INZV produce D16 pixel format. [1] https://chromium-review.googlesource.com/c/chromium/src/+/918001/3/media/capture/video/win/video_capture_device_mf_win.cc#199 [2] https://chromium-review.googlesource.com/c/chromium/src/+/944627 |
|
►
Sign in to add a comment |
|
Comment 1 by aleksand...@intel.com
, Mar 2 2018