Thanks everyone for the help.  My crowd doesn't really need to move, but I do need to pan over it in a stadium setting in 3-D, so I'll take a shot at creating it in Vue.   
SS

Given the tricky tasks and processing power -- and render time -- demands of trying to render over video, I'm not sure why ANYONE would ever opt to do so compared to just dropping Poser PNG images output into a video editor and laying in the background video beneath it.  I've had great success with it, and even succeeded in matching the shadows and lighting of the Poser foreground objects to the lighting, color tone and shadows of the background video.
SS

Yes, the image is from the music video of "Take On Me" by A-Ha. I was fascinated by that video from the first time I saw it, and it made the song unforgettable. And I spelled Brazil as Brasil out of habit (lotsa multinational 'furiners in my company). I have a feeling that I'll have to settle for second-best images, or maybe a stutter-style video (not unlike the A-Ha one) where I hand-edit each frame. I also found a Python script in Content Paradise that automatically applies Art Materials Cartoon Shaders to the chosen materials in a scene: kinda like a more precise version of the Toon Shader in Poser 7. SS

I also have some Toon Actions that I bought a couple of years ago. Miss Nancy's image sample reminded me of what it can produce. I also seem to remember some sketch actions from the same maker. SS

In Final Cut Studio (not sure about Express), you can set the duration of any still image that you import. For example, if your animation was at 15 frames per second, you would set the duration of each imported frame for a hair less than 0.067 seconds. Then you don't have to mess with the clip speed unless you really want to -- I, for one, don't like the effect that changing the speed has on the look of the video. As for using compositing software on the Quicktime output from Poser, I don't yet have Shake (but I'm thinking seriously about it now that Apple has cut the price), but I'm still leery about getting a clean knock-out, especially when the figures are outlined or where there are dark shadows on the figures or props. But then, I'm also a rank amateur when it comes to properly dealing with alpha channels. SS

The image above is an onion skin charcoal sketch done by hand atop live video, frame by frame, but I'm WAY too lazy and incompetent for that. That's why, in order to achieve the same look, I was thinking of Z-tooning (flattening the 3D figures into 2D-like images using a surrogate primitive as the parent object) a Poser animation, then exporting the animation (as a series of PNGs) either to Final Cut Pro (where a video filter or two would be applied to the PNG still-image sequence) OR using Photoshop's automation feature to apply a more complex sequence of filters to each individual frame (uh, all 18,000 of them - I figure that task alone might take up to 2 days of steady computer time). So yeah, I could probably get close to where I want without actual frame-by-frame postwork in a couple of ways, but I'm hoping that someone has stumbled upon an ideal setting in the custom sketch render setup (perhaps along with some light settings, etc.) that does a good job. I could take it from there. Incidentally, the image is from a well known 80's pop music video, which in turn used a storyline from the movie Brasil. I haven't seen anything quite like it in the 20+ years hence. When you finally think of it, you'll say the name of the band. SS

Well, this isn't a fruitful search. Aren't 3D animators also occasionally 2D animators? Z-toon and Photoshop, here I come!

Actually, you need to export them as PNG images, not tiffs. They go to a new folder, in order, so you simply select all of them and import them into a new bin (in Final Cut). Once imported into Final Cut, select them all again and drop them into the timeline. You can then nest that sequence into any timeline you want, where it can be edited, etc., or you can export it to Quicktime to create a self-contained movie. Works perfectly for me, every time, and I've used it with very long animation. Better yet, if your computer should freeze or otherwise blip at any point in the rendering process, you can restart from the last frame, rather than having to go back and start all over. That alone makes it superior to any other technique. And no, using a black background and then trying to key it out doesn't work - the keying picks up too many edges and shadows, no matter what. Not sure what they did in the tut. SS

Thanks, guys. And I'll try the bitch bending tools, too... SS

Thanks, Bar-Code. Is there something by Cepstral? SS

ernyoka1, you have one of the best avatars I have ever seen. Hilarious enough to print out and stick on the company bulletin board. Good job!

Any novel by William Gibson, but especially Idoru, which is about a virtual pop music star followed by a rabid bunch of cyberfans linked through a Second Life-type community, although I understand someone may be developing an anime version. Gibson's most recent novel, Pattern Recognition (a bestseller), is apparently being made into a movie. Gibson is one of the founders of the cyberpunk literary movement. Bono and The Edge are big fans. Some of you may know Gibson as the author of Neuromancer (another good topic for a CG movie, it was the first novel to win all three major science fiction book awards). I think it was as visionary as any novel published in the last 25 years. In their 1990 novel The Difference Engine, he and Bruce Sterling helped popularize the "steampunk" idea -- advanced science mixed with 19th-century design and mentality. The theme has made its way into a number of Poser props, costumes and figures as well as the actual name of a Poser vendor. SS

There is a new book on Animating with Poser 7, supposedly out right about now. 400+ pages. Looks really comprehensive. I plan to buy it.

I'll have to try your technique, tbop. I often need more than one female voice in an animation, and changing just the pitch is not enough differentiation. SS