For those of you who use Gmask Tracer, I find myself wanting to apply a matchbox to an individual gmask or geometry in gmask tracer, like blur, grunge, or any matte deformer.
Is this something you would use?
Is it at all possible @fredwarren @FrancisBouillon
If yes, I’ll make a feature request…
Hi Sinan,
To be able to apply a Matchbox we need to handle pixel images which GMask or Geometry are not until they reach the GMask Tracer output. One way I was able to do what you are asking for is to split your objects in multiple GMask Tracer output and apply Matchbox connected to the Camera to the appropriate output that would give you the ability to apply processing as you wish per output with the draw back to have to recombine the multiple outputs down the pipeline.
Is it what you are looking for?
Otherwise to it all in the same output we would have to do something like Source Node or introduce a Selective Matte extraction concept to GMask Tracer.
Let me know what you think!
Francis.
Hey Francis,
I was sort of expecting this answer, as I know masks and geometries are not treated as pixel images. Still one can hope 
…
I did look for Source Nodes in the bin, but (well perhaps, luckily) they weren’t there. I know how source nodes in action can cause problems.
The workaround you suggested is not that different than doing the same operation downstream in the batch tree. Plus the matchboxes don’t need to support cameraFX.
This was me being lazy actually. Thanks for taking the time to explain…
Cheers
Haha, I’ve had this same theoretical feature request in my head for quite a while (more for gmasks in Action), and never asked because of the same expected theoretical answer.
Love the idea, but you can just make individual outs for those masks, apply the matchbox and recombine after.
Are those mask usually in the same plane as the input?
I was thinking, would you find it useful to have Matte Selective in GMask Tracer that would behave a bit like Modular Keyer? That way you could use Selective Keyer and GMask and apply Matchbox on the result per Selective.
But granted that doesn’t solve the problem of geometry or Gmask that are in 3D using a camera track for example.
If you’re going to make changes to Tracer, can you disable the Comp output view while you’re at it? (And maybe implement my feature request for a command to automatically separate masks into their own outputs)
Having different outputs is the way fow now. I’ll use the matchboxes downstream.
My bug report about the gmask tracer is more important for me.