Help with collections, compositing, masking please. It could be any one of these I need help with, or something else…
I want my final render to include two objects.
One object has motion blur. The other object is the same dimensions, material, etc as the first object and is in the same location but it has no motion blur. I want the object with no motion blur to be “in front” of the the object with motion blur and I do not want any optical interaction between the two objects (no blending of the two and no shadows from from object to the other, etc.)
There is a also a background. This can go in the same collection as the object with motion blur.
I am using Cycles.
I have avoided mentioning Collections, etc, because I do not have any preferred method. The best method will do!
Here is a crude illustration of what I want to achieve in Blender:
Picture 1: the object with motion blur and background.
Picture 2: the object with no motion blur (has a transparent background).
As long as you save the image w/transparent background to a file format with an alpha channel, you can use the Compositor to superimpose it atop the motion-blurred image using an Alpha Over node. No need for Collections.
If you’re wanting to return to this Composite, later, after closing Blender, those renders will need to be saved as image files. Otherwise you’ll have to re-render b/c those render layer nodes will be empty.
No big deal for a single image, but you’ll definitely want to save the image sequences.
This will actually happen automatically when you Render Animation. Those image files will be saved to your drive wherever you specify in the Output Properties tab. That’s also where you’ll set the file format. For renders with transparency you need a format that supports an alpha channel. PNG, RGBA, 16 bit will work for what you’re doing here.
You then load those sequences into the Compositor via Image Sequence nodes. Those will replace the Render Layer nodes in your node tree.
Please excuse my slowness. Are you saying I render a sequence of images for the background (we’ll call it). Then I render a sequence of images for the objects that appear in front (with a transparent background). Then bring them back in one from each sequence so two at a time in an image sequence and composite them together as desired? That sounds much better and more robust than the set-up I have been trying.
If I have understood it correctly, can you please give me the details I will need?
Pretty much. After rendering the two image sequences, in the Compositor add 2 Image Sequence nodes. You’ll be prompted for their location on your drive. Then simply connect them to the Alpha Over node as above. Connect it to the Composite node as above.
You can, and probably should, do this in a new Blender file, setting a new location for the composite render. But if you use the same one you rendered the sequences in, you’ll want to disable Use for Rendering in the View Layer Properties tab so that you don’t waste time re-rendering your previous sequence.
Then Render Animation.
You could render directly to a movie file format, but best practice is to render another image sequence and then import it into a NLE to export a movie file.
Wow. Everything as you said, but I added a link from Alpha of the top image to Fac of the Alpha Over.
Then I set number of frames in the Timeline and smashed Render Animation. Is that really it??
Apart, this case, in youtube Pixeltrain channel has a very good and ordered explanation in usage of separated collections in rendering and compositing.
Seconding @simonwebnet. I learned Blender compositing from this very series. I wish I had just linked to it instead of subjecting you to my rambling, hard to follow explanation.
