Hi guys,

I've been doing a few tests and come upon a weird behavior. 

If I want to output 2D images with proper holdout matte from deep images, I get 
different results from using a DeepHoldout node vs using the holdout option in 
the DeepMerge node connected to a DeepToImage node. 

The Alpha of an image coming from a DeepMerge with a houldout operation is full 
of noise in the edges, mostly visible in motion blurred images, where the alpha 
generated from a DeepHoldout node is much cleaner. The DeepHoldout node though 
takes about twice as long to process then the DeepMerge and DeepToImage 
combination. 

Is this behavior normal? What's the difference between the two approaches and 
why is the DeepHoldout process so much longer then the DeepMerge process? I'm 
using Nuke 7.0v5 with Exr 2.0 deep images.

------------------------
Mathieu Leclaire
Director of R&D
Hybride Technologies, a Ubisoft division



_______________________________________________
Nuke-users mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Reply via email to