Adds full frame implementation to this node operations.
No functional changes.
No performance changes.
Details
Details
- Reviewers
Jeroen Bakker (jbakker) - Commits
- rB5deb3229a075: Compositor: Full frame Mask node
Diff Detail
Diff Detail
- Repository
- rB Blender
- Branch
- cmp-node-mask (branched from master)
- Build Status
Buildable 16151 Build 16151: arc lint + arc unit
Event Timeline
| source/blender/compositor/operations/COM_MaskOperation.cc | ||
|---|---|---|
| 180 | Don't we need to cleat the buffer in this case? | |
| source/blender/compositor/operations/COM_MaskOperation.cc | ||
|---|---|---|
| 180 | At method start it's cleared. Then if there is no handles it will be already cleared here. And if there is only 1, we don't want to clear that handle rasterization or averaging it. By looking at tiled implementation it should produce same results. | |
Comment Actions
- Merge branch 'master' into cmp-node-mask
- Merge branch 'master' into cmp-node-mask
- Use iterators and only clear buffer when there are no handles
| source/blender/compositor/operations/COM_MaskOperation.cc | ||
|---|---|---|
| 180 | Not for this patch, but we could move this responsibility to the operator. If I understood correctly we could reduce unneeded clearing when we reuse or allocate the buffers. Normally you would overwrite the whole buffer. | |