PDA

View Full Version : Antialiased depth buffer?


JMarc
11-06-2002, 03:06 PM
Can anyone tell me which of the many available depth buffer savers is antialiased?

Jean Marc

JMarc
11-06-2002, 04:36 PM
I just found this other similar thread:

http://www.cgtalk.com/showthread.php?s=&threadid=15851&highlight=depth+buffer

But even with the hidden "Depth Buffer Antialiasing..." option set to either Minimum or Average (I don't see an On or Off so I assume it's always on) and Adaptive Sampling turned off in the Camera panel, I can't seem to get my Depth Buffer to be antialiased. This is a real pain for my current project which is using lots of motion blur.

The Depth for the mo-blurred objects is showing up as one solid shade of black when it is moving towards camera. I would expect it to incorporate the change in distance to the camera if the antialiasing passes are being rendered into the depth buffer.

It looks like I'll have to render the extra pass using the fog workaround but since this project is at film resolution this is a real deadline killer.

CG.p
11-06-2002, 05:21 PM
The Depth for the mo-blurred objects is showing up as one solid shade of black when it is moving towards camera. I would expect it to incorporate the change in distance to the camera if the antialiasing passes are being rendered into the depth buffer.


So how do you want a mostly "transparent" object to show up in a depth buffer? Do you want the object or the stuff behind it?

The fog 'hack' would be best for what you want.

JMarc
11-06-2002, 05:57 PM
I get your point. I am compositing this piece so I will be rendering the background separately. As for objects that overlap in depth, I assumed (wrongly, I suppose) that the depth info would translate similarly to the way that transparency info is written to the matte. I can see now that this doesn't really make sense when you consider the transparency of overlapping motion-blurred objects. If only there were some way of using object id's to determine what percentage of the depth channel to use.

I am trying to use the depth info to control depth of field in the composite in order to retain the most flexibility but the current depth channel doesn't match the render due to the lack of motion blur "stretching" of the objects. This results in some nasty fringes and hard edges in areas that are supposed to be soft and smooth.

I think I'll try the fog technique or maybe a luminosity gradient linked to distance to camera.

Thanks for helping me wrap my head around this stuff!

CGTalk Moderation
01-13-2006, 10:00 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.