Refine your workflow. First of all, each panel must be a standard output size if you are sending the images to a series of projectors. You must figure out how the show is going to be put on, then follow the requirements for multi stream projection from the supplier of the gear. If you don't check first your entire project may have to be redone.
Secondly, I'd learn about proxies, buy all the ram that will fit in your system, and think about your workflow very hard. You could set up 5 cameras on your huge comp, you could set up guide lines and then duplicate your master comp and use the guides to break up the copies into smaller comps using the crop composition to region of interest feature.
Another option is to set up a single comp at a more reasonable size to produce a version of your comp that contains all of the elements but at an HD frame size. Then you could use the scale comp script file to create a full size comp that could be divided up into final output frame sizes using the crop to region of interest option on duplicates.
If it were me I'd set up an HD comp, cut it up into 5 equal pieces, then set up the animations. If the footage (images) used in the final are huge, use proxies, if there's artwork, make it vector art, if there's video, use it at full resolution.
None of this advice will mean anything if you don't have the output specs from the folks providing the projectors in the first place. I'm a little concerned by the line "5 screens of different dimensions." Every multi screen system I've worked with has the same output options, NTSC Normal, NTSC Widescreen, or HD.
Well thanks Rick. There's a lot of good stuff in there to think about. Really appreciate it.
So I work closely with the lighting company and we have a proven workflow in place. Regardless of the panel size all my outputs are 1920x1080. All machines have to run this output so I supply the content within this frame size. For example, the centre screen is 880x600 which I just place into a 1920x1080 comp.
So the output workflow is all in place its just finding the most efficient way to refine the workflow in AE.
I'll try the 5 cameras and the ROI idea, that could help. I use the scale comp script all the time but Im not having trouble in the construction stages, just the final rendering. I also use proxies all the time but that's not what Im having trouble with.
I'm just trying to find out how AE deals with these procedures. Does it deal with ROI differently than placing a camera in a comp that only sees 1/5 of the comp. Will it still try and render the 4000x comp?
Thanks again for your time.
If you have camera moves you don't need to add additional cameras. All you need is your guides and the ROI tool. Duplicate a comp, set the region of interest, then go to the menu Composition>Crop Comp to Region of Interest. You can also see the position of the ROI in the info palette.
The cameras don't change the comp size. Multiple cameras is just a way to set up multiple views easily. You would still need to resize the comps. It sounds like multiple cameras would be an unreasonably complex solution to your workflow.