Slice Transform - specify input region
Posted: Fri May 20, 2022 21:31
At the moment Slice Transform just doesn't cut it when trying to resize content into different kinds of slices.
The problem is having to map the entire Composition-shaped-texture into a slice. Compositions can be all kinds of whacky dimensions (to control multiple outputs, non-rectangular LED walls, etc). The Layer auto resize option is great for "normal screen" (single screen) shows, but otherwise can be very problematic.
If you want to put an image in a bunch of slices, one way to try that is to scale up the input media to FULL COMPOSITION size (which might be very large!), which Slice Transform then scales down to slice size. This is extremely wasteful. And also often bad results when there are mismatched aspect ratios.
One solution is to allow input dimensions to be specified for the Slice Transform. eg the comp may be 4Kx2K, but you just want 512x512 pixels from the middle of the screen to be mapped to your slices.
If you could specify it separately for different slices (and allow different scaling options per slice, as requested in another post), even better.
A more system-wide change to address this problem would be if Resolume actions were input-texture size aware. eg if you load a clip that is 512x512 and you use Slice Transform on the clip, the Slice Transform should know that the input is 512x512 and map that to the selected slices, with the output being the full size of the Composition, or perhaps also be otherwise specified.
-subpixel
The problem is having to map the entire Composition-shaped-texture into a slice. Compositions can be all kinds of whacky dimensions (to control multiple outputs, non-rectangular LED walls, etc). The Layer auto resize option is great for "normal screen" (single screen) shows, but otherwise can be very problematic.
If you want to put an image in a bunch of slices, one way to try that is to scale up the input media to FULL COMPOSITION size (which might be very large!), which Slice Transform then scales down to slice size. This is extremely wasteful. And also often bad results when there are mismatched aspect ratios.
One solution is to allow input dimensions to be specified for the Slice Transform. eg the comp may be 4Kx2K, but you just want 512x512 pixels from the middle of the screen to be mapped to your slices.
If you could specify it separately for different slices (and allow different scaling options per slice, as requested in another post), even better.
A more system-wide change to address this problem would be if Resolume actions were input-texture size aware. eg if you load a clip that is 512x512 and you use Slice Transform on the clip, the Slice Transform should know that the input is 512x512 and map that to the selected slices, with the output being the full size of the Composition, or perhaps also be otherwise specified.
-subpixel