-
How could I have done this more efficiently?
Hey, I just finished a commercial spot using AE CS3, and everything worked great, but the render time was close to 3 hours, and I have a feeling that I could have done the same thing much more efficiently.
Here’s what the spot looks like:
https://video.ericbarker.com/Commercials/temp/Beaver_Sports_nosound.movIt’s a floating grid of hexigons. Here and there in the grid, I have photos of people doing various winter activities. The “camera” pans around the grid and zooms in on the pictures. Since the pictures are close to full screen when zoomed in on, I had to make the entire grid high resolution. To do this, I tiled 6 smaller grid segments, each 3700×3700 pixels, comprised of about 150, or so, hexigonal shape layers (polystars). All 6 grids were parented to one of them, so that they’d all move and rotate together, creating one large continuous grid.
To move the “camera” around, I animated the anchor point instead of the position, so that at any time, rescaling the grid would cause the viewpoint to zoom in on whatever was directly centered on the screen.
I suppose I could have used a real camera layer to animate the viewpoint, instead of moving the grid, but I’m finding it difficult to truck the camera around, the controls seem more aimed at panning and tilting than trucking left/right and up/down. Also, the fact that the grid is tilted at about a 20° angle would have made camera movement even more difficult.
This isn’t a very complicated concept, but it still took 3 hours to render on the Dual Xeon XP machine I have at work. What could I have done to make it more efficient?