Have posted on this subject before in Yahoo forums, and got couple of good replies. Just wondered if anyone could improve on those suggestions, as they are not working perfectly...
I am using a dual core processor and want to speed up the render time of my images.
Current version of TG (not 2) doesn't support dual core, so I am trying to split the image into 2 parts, render each in a different Terragen instance - therefore using both cores to the max - and then stitch back together in Photoshop.
If I am rendering a 3000x2000 final image view, I change the camera pitch to say 15 degrees either side of the desired view and then render two images of 3000x1000 (with a small overlap to ensure correct alignment).
Everything looks fine until I come to stitch the images together and they don't line up properly

It looks as if one image is being taken with a slightly different camera zoom setting. I can only assume that changing the pitch is altering the camera zoom as well - although (and after plenty of maths) I have checked and when changing the camera pitch, TG changes the target coordinates to retain the distance between the camera and the target at the new pitch.
I have considered rendering two images with only changing the camera bank by 180 degrees, but this will produce an overlap of 33% and therefore probably not reduce the render time significantly.
Any other suggestions?
Cheers,
Jon.