Vegas throw-togethers

Two copies of one Faye and the finished product
This media arose from a discussion about the possibility of simulating Technicolor in Vegas. I took a quick crack at 2 and 3 strip simulation in Vegas (though, I spent so little time that the 2-strip version isn't even done in pseudo linearized light space), and cooked them in to a comparison file here, and here are the files all at once. I'll make the projects available for people if there's any demand (who reads this stuff?), but it's simple enough to say that the easy way to fake 3-strip Technicolor is to use the channel blend to do the following:

Turn on row normalization.

R = 1*R + -.2*G + -.2*B, etc...

Do this between gammas of .666 and 1.5 to give the look more punch.

And now I've made the presets available as FX packages. I've managed to cook 2-strip in to an FX chain. Two versions of 2-strip (one yellow, one idealized) and one 3-strip are available here. You'll have to use the preset manager to load them over. You might find that the 2-strip is too magenta feeling. If that is the case, take the R channel contribution in the second channel blend and push it to a larger negative number, like -0.7. If 3-strip is too colorful, take the negative contributions (currently at -0.2) and change them to something of a lower magnitude, like -0.1. If you've already pushed the color in your footage, it will over-blow with these chains. Astute pixel pushers will observe that they can yank the first 2-strip channel blend and make the G and B lines XXX, 1, 1, 0, 0 instead of XXX, 1, 0, 0, 0. This loses some adjustability (and obviousness of the process), but it does the same thing numerically. This is left as an exercise for the user (can't give away everything)...

So, not willing to leave this alone, I've put together another preset here. You can see the results in .
All images are copyright of their respective owners. In this case, presumably, the Aviator effects crew or their employers.
This was put together for Eric with footage taken with a PDX10 placed on an office chair. Everything came from one take, and total editing time for the two versions was roughly one hour. Note that features of interaction with the scene (shadows, reflections, etc.) show up in difference layers.

Two of the same person in one scene and the thermoptic version

The projects (as a zip file with media) are here.

To do this kind of work, make a point to disable auto-focus and auto-exposure features on your camera if you can. It makes life a lot easier. Avoiding linear lights (such as the ones used for this footage) is a good idea if you want to localize the effect of your shadows and scene interaction. As with chroma-keying, it is important for there to be a difference between the foreground and scene content if you intend to establish a good mask. Since these areas are generally so similar as to not be visible in motion video, one can be fairly lazy and still show decent results.
This project was generated entirely in Vegas 5.0 using the new 3D compositing features of Vegas. The project file from which this video was generated is less than 200KB in size.

3D titles

If you want to make use of these projects, you will be best served by opening your internal preferences (ctrl+shift) and changing the "Max Track Cache Elements for 3D" value to something substantially higher than 8, like 100.

A version of the project that does not make use of effects is here. This shows the 3D layout of the project. Unless you are absolutely certain that you want to sit and wait quite a long time for each individual frame, this is your project.

The original project as it was composed for a 10-machine/30-hour Network Render is here. Note that this project was constructed to intentionally take an absurd amount of time to render. Changes in the construction of this project (pre-renders) could yield shorter render times with no changes in the final output.
This is an example of creating shadowed text from a single image. The source project is here. I will put together a step-by-step tutorial when I have more time. A revised version adds some subtle changes. The project for this one is here.