occlusion render pass?

[visualboo]

- Create a new empty pass and rename it AO. Or whatever you like.
- In background_objects_partition, (to keep things simple) add a constant shader and plug an AO node into it's color channel.
- Bam!


If you have shaders that use displacement mapping or anything more than simple. You would want to approach this slightly different. Press F1 and read up on overrides.

2 side notes also:

- A quick search would probably turn up 400 threads on this. No biggie tho.
- Don't use the preset passes. Learn HOW to re-create those and others from scratch so you understand this stuff. Presets are sometimes nice but not a replacement. Especially if you are new, as it hurts more than it helps in the long run.

[Rens]

Another quick side note; you can plug the AO shader directly into the surface port, no need for the constant shader. :)

[visualboo]

This is true... although I don't think it's a good habbit.

[JDex]

This is true... although I don't think it's a good habbit.

Could you elaborate, please?

[vortex]

hello,

there's also another nice way to get the AO.

it's covered here: http://www.xsibase.com/forum/index.php?board=12;action=display;threadid=33983

[Jeroen]

wow, these answers helped me a lot.
Thanks, you saved my skin,

[visualboo]

Could you elaborate, please?
Ah sorry Jason.

I would say continuity's sake? When you start adding overrides and such to partitions later in a project, you don't want to override something and have it NOT affect certain objects because it doesn't know where to plug the param into. If the shader doesn't exist, how can it override the param?

IMO, it's best to just throw a constant shader in there (in this case) and have everything be setup the same. Constant shaders are about as light render wise as you can get anyway, so imo it's worth it.

I'm sure someone from soft could mention a few other reasons it's a bad idea but this is my main reason.

[Atyss]

Ah sorry Jason.

I would say continuity's sake? When you start adding overrides and such to partitions later in a project, you don't want to override something and have it NOT affect certain objects because it doesn't know where to plug the param into. If the shader doesn't exist, how can it override the param?

IMO, it's best to just throw a constant shader in there (in this case) and have everything be setup the same. Constant shaders are about as light render wise as you can get anyway, so imo it's worth it.

I'm sure someone from soft could mention a few other reasons it's a bad idea but this is my main reason.


Nothing wrong with continuity, but I'm a strong believer in the hierarchy of priorities. Continuity is great and needed, but the problem is that it can affect rendering, and when it comes to rendering, I believe speed stand higher in the hierarchy than continuity.

Why adding more data and overhead to something that doesn't need it? Okay, perhaps just a tiny-tiny amount of those, but like in car racing, every tiny-tiny bit counts. Looking at it on case-by-case basis, in this case there really is no need for the constant shader. All we want is occlusion, and the shader is applied at partition level.

Don't get me wrong, there are cases where I *may* include shaders in render trees that are meant to be used through overrides only.

But hey I guess it all comes down to priorities. The amount of user time saved vs the render time saved is important. If speeding up render means more work for the artists, well, it may not be a win. It'd be interesting how much work it saves you to add this constant shader in there.



Cheers
Bernard

[CGTalk Moderation]

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.