Summary of discussions
m36457 On WD of Carriage of ROI coordinates
Thank you. Editors to prepare a CD based on the working draft.
m36613 On Carriage of ROI Information
We are concerned that a difference approach results in samples that are either 8 or 1 byte (and hence the sample size table enlarges), and we now need a sync sample table (to mark the samples that have absolute coordinates). However, this makes the simple case no worse (omit the sync sample table). We can use the compact sample size box, also. It seems it depends on whether there is enough usage to justify the slight added complexity. We would need the equations, please, given existing x,y etc. what are the new x,y etc.
The document also worked from something other than the working draft. We’d like an analysis of how much is actually saved in some realistic conditions. We would prefer not to incorporate this now; we would be prepared, of course, to add it in response to a ballot comment, given answers to these questions.
We go to a design that does linear interpolation and pre-roll, and go to PDAM.
Dostları ilə paylaş: |
