These seem like a distinct improvement, and trying the minf box will surely elicit comments on the PDAM if there are problems.
The contribution expresses concern that some simple descriptions may involve a lot of structure (sample groups, and so on). However, it does seem to overlap with existing solutions, and we are not totally happy about that. It’s also not clear what a file author should write if he has no particular display in mind: what constitutes an ‘associated view’?
Perhaps if we are indicating suitable stereo views (only) to have a view information box that details which views make suitable stereo pairs? Do we need view disparity for non-adjacent views?
We agree on adding the view_ids to the camera parameter boxes.
In general, we should inspect the PDAM to see: are simple things simple, and complicated things possible? There are almost certainly some simple things that are complicated to do at the moment, so the thrust of this proposal (that there should be simple ways to do simple things) is exactly correct.
We put the view association box into the multiview group container box for now, but fully expect comments on the PDAM on the overlap in functionality.
Technical Work in Progress.
Dostları ilə paylaş: |
