Hi All ,
.
Parameter DCM composite samples < What is this doing when checked on ?.

On a side note in Nuke. I can read the deep samples so the data is there, although I cannot see any data in deep.front when trying to view in the alpha channel. Is there any way just to view deep.front in EXR2.0 ? .

Rob

I don't have nuke here so I can't verify, but simply put: if you have three semi-transparent surfaces of 50%, you will get samples of 50%, 50% and 50% with the option disabled, and 50%, 25% and 12.5% (from front to back) if enabled.

For the side note, do you have a recent build of Nuke? The early 8.0 versions had real trouble with Houdini deep exrs..

Thanks for the answer, I thought for a moment it might be multiplying by a shaders opacity.
Running build v02. BUG ID 39845 DeepRead:

I thought this was fixed ….


Thanks

Rob

Would be nice to get an authoritative answer (as I can't confirm atm), but I think the premultiplied/unpremultiplied state of individual samples is unaffected by this parameter, it only affects the compositing resulting from “stacking” the samples in one pixel.

I'm sorry I don't remember what got fixed in which version, just that in the beginning of 8.0 cycle it was really broken. Like with Houdini, when in doubt, test the latest build..

Ive upgraded to v04 build, fixed the display issue, but now theres a ew other issues with workarounds. Noted the RFE for Houdini to output deep.back which would be nice. Of note I have getting constant crashes in Nuke when rendering with DCM - pre composite samples turned on. Ive got a few more tests to see what I can come up with in optimizing the files.

Rob