Hello to all, I want to know, if it´s better to bake for a movie character ambiente occlusion and use it as texture or if it´s only good for game characters?
If your using it to start a texture then you’ll be using it somewhat differently, e.g. as the start of a stylistic colour texture.
In general you should only bake AO textures for objects which are static and in static lighting.
(I believe that’s right!)
So you think it´s not neccessary or not really good to use it for Characters?
Yes, use AO on characters, but since they are not static they need to be recalculated for each frame.
Oh, that´s a little bit too much work Ok thank you very much
You don’t do it! It does it automatically for each frame… You were on about baking them out to maps before!
Ah, ok, I know what you mean, but what things does a Charcter need? TextureMap, NormalMap and BumpMap I have already done. But will I need a SpecularMap for a Character in a Movie?
“It all depends.” There are no cut-and-dry answers to questions like these. These are all going to represent deliberate choices on your part, and most of those choices will be trade-offs. (And, none of them are irreversible.)
Usually, specular information isn’t mapped because it changes based on the camera position, and the camera is usually moving. There are a number of video tutorials out there that talk specifically about things like this. In general, think about an approach that is something like this …
- For things that are not moving, plan on “baking” as much of the information as you can. Stuff that is constant, that doesn’t depend on the camera position etc, is a good candidate for this.
- For computationally expensive things like AO, isolate only the things in the scene that actually benefit from it, and prepare maps for these. (If there is “bang for the buck,” then spend the bucks. But if there is a way to “cheat the shot” that is “almost as good,” bankroll your money=time. “Go ahead. Be creative.”)
- Things (like specularity) that reasonably do vary from frame-to-frame, can continue to be calculated frame-to-frame. (The computational load of calculating “only these things” will be drastically less than the alternatives.)
- Start simply. If you determine that an additional, perhaps computationally more expensive “fee-chur” is truly needed, it is often possible to calculate it separately, after-the-fact.
- Use render-nodes aggressively. Break out all the work-products into separate (MultiLayer file-format) channels, so that you can “tweak” them if you have to, without starting over from scratch. Tweaking a color-balance issue, for example, shouldn’t demand a re-render…)
- As usual, do all of your editing first. You should determine (via GLSL renders, etc.) exactly what footage you actually require for your show, before you start to refine any of it. Decide what you’re going to “leave on the cutting-room floor” in advance, and do not bother to actually render any of that material.