On the decoding complexity, there is a worry that this is expressing a ratio between this track and an (unidentified) external track. Actually, it’s decoding complexity ratio on an arbitrary scale. In a DASH context, one would insist that all the representations in an adaptation set were on the same scale, i.e. that the decoding complexity numbers were comparable.
The green sub-type is exposed as a subtype (after a “.”) of the codec type. The setup information (in the sample entry) gives a map of types, and the constant record size in each sample for the data of each type (padded up, if necessary). These could perhaps be clarified.
It should be clear that the metadata could be ignored – e.g. the display metadata doesn’t describe a situation in which the luma values have been adjusted, but a situation where they could be adjusted (and a suggestion as to how much).
Can we merge the quality metadata and the green metadata into one WD? We need a new suitable title, perhaps “WD Carriage of Timed metadata metrics in ISOBMFF”?
We agree to go to CD at this meeting (with an editing period).
Dostları ilə paylaş: |
