Hopefully I’ll explain myself clearly enough…
I need to do a text effect on a chalkboard inside a classroom.
The overall rendering will be done with YafaRay (Photon mapping). But the text effect will be done with Blender.
This generates the following 2 questions:
Thanks in advance guys 
JDL
There’s a tutorial on “cutting steel with a blowtorch” that talks about how to use paths in a very similar way.
The source-files that you will produce in Blender will consist of the text, and maybe the piece of chalk following the text as it is gradually exposed. It will all be on an Alpha = 0.0 transparent background, anti-aliased and ready for comping.
I suggest that you generate the lighting and the shadows all as separate layers … that is to say, as separate outputs, so that when you get to “comp” you can adjust all of these things separately.
From a pure-compositing point of view, for example, the thing that you need for Blender to tell you about “light #3” is… where it falls on the chalkboard surface (which can be described in a single frame if the light doesn’t move) … and, separately, how it is occluded from the chalkboard plane by the traveling chalk … and, separately, how it falls on the chalk as it moves. An entirely different color or treatment or texture of the light, or the shadow, or both, could be added into the shot in “comp,” as long as you have available to you this information about just how and where it falls. As long as you have it all in clean, detailed, independent channels… a directory full of carefully-labeled EXR files.
You need exactly the same thing from Yafray. (But, heh, you just might discover that you don’t need Yafray.)
If you are going for “complicated photorealism,” as I guess that you are, then you’re going to be putting the shot together in comp, not in the individual source renders. You’ll be building-up the effects one step at a time, running through it over and over again, changing this and bumping that, just like they do at a mixing panel in a recording studio. (“Let’s try a little bit more chalk-dust in light #3…” “Okay, just a sec… got it. Here goes…”)
Thanks a lot for your valuable input sundialsvc4
It does help a lot to sort out my ideas.
I might just go the extra mile and do as you suggest.
JDL
Believe me … when your shots get complicated (and they very quickly do…) this sort of detail is worth it.
Now, it should also be said that for a very simple render you might be able to get away with a little bit less paperwork, and yet still accomplish the same basic workflow, through the use of compositing nodes that draw from “upstream” rendering stages rather than from actual files. A series of nodes can break-down the individual stages in such a way as to make them individually “tweakable” even though it is necessary to re-run that blend-file each time you do a tweak. (This might, of course, be one of several such files.)
The key concept here, is that there is a data flow whereby many different channels of isolated information flow separately, “gently down the stream,” to finally combine into one or more finished (or intermediate) outputs … and that you consciously control and design all of that data flow.
You should be looking for “the right mix”: [ul]
[] Avoid making the computer do redundant effort that can reasonably be avoided. [] Avoid an excessive “paper-work” burden for yourself. [*] Above all, provide yourself with fine-grained control and the ability to “fine-tune” the finished product, without subjecting yourself to long delays. [/ul]
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.