People: Render or Post-Process?

[mjm223]

I'm curious to know, when it comes to putting people in your scenes, which method do you think works best? Putting the people directly in the sketchup model or adding them later in post-processing? I've done it both ways and have found pros and cons to both. Putting people directly in the model allows me to accurately scale them to the perfect size, and they respond to the GI and lighting much more accurately. But adding them in Photoshop allows me to move and re-arrange people or swap them out for new people without having to re-render my scene.

How about you guys (and gals)? Which method do you prefer? Do you have any strategies that makes one method more successful than the other?

[Jackson_cg]

I prefer rendering in clipmapped people in the scene, so that VfSU takes care of scale, lighting and shadows, but I sometimes add a couple of people in the foreground in post. A simple trick to get the best of both worlds is to run one render with people and one without. Layer them in PS, apply adjustments to both layers and then you can easily delete out anyone you want by erasing through the "people layer" revealing the "background layer" beneath. You don't even need to render the whole scene twice as people obviously tend to be in the lower half of the scene so run a Bottom-to-Top final pass and cancel it when it's cleared all the people.

I often do the same thing with trees so I can fine tune their transparency to reveal features behind.

/Jackson