Yes, this is a very inefficient way to do it, as it requires cropping AND overlapping images (12% overlap is about the minimum you can get away with due to GI inconsistencies), AND reassembling images in Photoshop.
@lightning: I bought 2 new Dell desktop PC's off of Craigslist a while back for a great price. Both are identical, dual-core 3.0GHz PCs with 1 GB RAM each.
Here is the process I use: I assemble a TG2 image on my main PC, keep all files, objects, textures, etc in a single image folder. Once everything is dialed in and I am ready to render the full version, I transfer the folder to the other 2 Dell PCs.
Then I open 2 sessions of TG2 on each PC and then open the image file under each instance (which means 4 instances total). These are split into overlapping quarters, then each is set to rendering. Once complete, I'll take all 4 image quarters into Photoshop and simply reassemble them (takes about 1 minute or less).
Sometimes, I'll have 2 separate projects going on, so I'll just render on 1 of the Dells, splitting the image in 2 halves. Alot of my past images have been done this way, and I'll put the wording "rendered over 2 cores".
This is all why I am anxiously awaiting multiprocessing support.
