Add colorimetry information API in Picture #94
Conversation
src/lib.rs
Outdated
| DAV1D_COLOR_PRI_SMPTE431 => pixel::ColorPrimaries::P3DCI, | ||
| DAV1D_COLOR_PRI_SMPTE432 => pixel::ColorPrimaries::P3Display, | ||
| DAV1D_COLOR_PRI_EBU3213 => pixel::ColorPrimaries::Tech3213, | ||
| DAV1D_COLOR_PRI_RESERVED => unreachable!(), |
There was a problem hiding this comment.
Is that indeed unreachable? Also below in the same places
There was a problem hiding this comment.
With current dav1d git master yes, I've checked these appear only in the header file. Maybe we could make these methods return an Option<>?
There was a problem hiding this comment.
@lu-zero What do you think?
There was a problem hiding this comment.
The hacky way is to return pixel::ColorPrimaries::Unspecified as well, since the user would have no mean to deal with it anyway and add a mean to get the raw bitstream value in case the user knows how to use it since there are few "for future use" values. that assuming PRI_RESERVED maps to the bitstream value.
Anybody willing to read ISO/IEC 23091-4/ITU-T H.273 and doublecheck?
There was a problem hiding this comment.
This one? https://www.itu.int/rec/T-REC-H.273-202309-P/en
I'm not registered in TIES.
There was a problem hiding this comment.
https://www.itu.int/rec/T-REC-H.273-201612-S/en is freely accessible
There was a problem hiding this comment.
Well, in the spec the reserved value is specified as a range, for instance, for ColorPrimaries, it is 23-255, while DAV1D_COLOR_PRI_RESERVED is 255. Same goes for TransferCharacteristics (19-255) and MatrixCoefficients (15-255).
There was a problem hiding this comment.
I'd say that mapping it to Unspecified wouldn't feel too bad. Maybe I could future-proof av-data and add a Reserved(u8) but that's unrelated to this patch.
philn
force-pushed
the
colorimetry
branch
2 times, most recently
from
2ad25b7 to
2b29154
Compare
|
I'd land it |
|
Thanks for the reviews :) Is a new release planned soon? |
|
If dav1d-1.4.1 didn't bring anything more we can make one even today (or tomorrow). |
|
The release notes since 1.3.0 mention mostly optimizations. I've checked 1.4.1 works fine here (on x86_64) with current master branch of dav1d-rs and my WIP colorimetry branch of gst-plugin-dav1d ;) |
|
There are no changes in the headers since 1.3.0 so yes that seems fine. |
Fixes #93