how to apply motion blur to one or a group of objects without applying it to the camera texture?
by compositing multiple render target.
send all of those objects you want to blur and their occluder (if any) to a dedicated render target which you’re going to apply motion blur, and then another render target of stuffs you don’t want to blur. and then combine them to a new/final render target that will put them together as a composite which then you can apply global post processing on top of everything, like color correction for example.
I roughly understand this, but when I try to assemble it in a project, nothing works as it should.)
have you double checked that they are in their proper layer too?
can you make a simple file with these render targeting settings?
Here’s a demo project. There are a lot of little things that can trip you up in this situation.
If you need your render target to be transparent, you need to set the clear color to transparent. Otherwise it will render with a color or texture background.
Your 3D materials need to include the alpha channel in the color mask, or they won’t render into the transparent render texture for some reason.
Double check your layers for cameras and objects. Confusingly, an object’s child can be on a separate layer than it, so you need to select the whole subtree and assign the layer to the whole hierarchy.
snap-render-targets.zip (5.5 MB)
Now that I’m posting it, I realized that there is some color bleed in this blur. I think it can be fixed simply by applying the blur effect to the rendered output, but you might need to make your own shader instead of using the post process component.
its cuz object add nearest black pixels. here is fix
snap-render-targets 2.zip (5.7 MB)