AWS Thinkbox Discussion Forums

Shading KCM?

Deprecated Krakatoa MX 2.x Beta Discussion
#1

Ok, not something that will happen for 2.0, but it would be cool if we could get access to the shading and control it with a KCM.

We can do shading in a KCM now, but we can’t do a loop over the lights. So while it would be cool to allow us to do such a loop in a KCM, it would be nice if we could just make our own shading model that gets used during the lighting/shading phase of the render. This special KCM would just get called for each light. Or we can have 2 KCM’s, one called for each light and one called at the end to accumulate the results or do some other post-process.

If we are rendering to an image, you would just write out the color, but if we were rendering to a PRT, you could save the various lighting data out to channels.

#2

Cool idea(s). I think a KCM that is executed once per-light that is summed to make the final lighting value could be very useful.

#3

Chad is on a roll :smiley: more pain meds quick (or the cessation thereof, in that case no more for you)

#4

I think the summing might be a bit too generic, especially for NPR effects. But it would certainly be the default.

Ah, I see, you want to avoid accumulating lighting channels. So by summing them, you can just keep accumulating the results to the one buffer. (Or two, if we split out the LCache and the currently un-cached (and streamed, right) shading).

I was hoping we’d be able to save out intermediate channels to PRT’s, but I can see where that would involve a lot more re-architecture than just replacing the lighting/shading with a simple looping KCM.

  • Chad
Privacy | Site terms | Cookie preferences