Yes it’s possible, basically the problem come from the fact that the post-processing is camera based. Therefore you need to setup a multicamera scene. One camera for post processed meshes, the other for no postprocessing. I had explained a way in this topic.
Here is what is needed :
Creating 2 cameras : camera, noPostProcessCamera
Putting noPostProcessCamera on another layer (0x10000000), same for the unprocessed meshes
The above is quite simple and gives this result : Playground
Now the problem is that there is a depth issue with PostProcessing and multicameras. You need to setup a shared depth texture for the postprocessing pipeline :
Set up a RenderTargetTexture
Set up a PassPostProcess for the noPostProcessCamera
Share the depth with pipeline.imageProcessing.inputTexture.shareDepth(rtt.renderTarget)
As far as I know, if you want to remove specific objects from PostProcessing, yes…
Otherwise an other way would be to do two render passes with the same camera, but I’m not sure it will ease the stuff ^^