CompuServe Messages

My guess is to what is happening is this: 1. Imagine renders a continuous stream when it renders to regular frames. When it's done, it closes the file, and the PAR munches it, and everybody's happy. 2. When rendering fields, Imagine does its setup, renders the first field, but in a temporary format. It then pauses a bit, sets up again, and renders the second field, probably also in a temporary format. Then it combines the two temp files. Now the PAR is waiting for a frame from Imagine. When Imagine completes the first field, which is not a complete frame, the PAR software assumes that the frame is done, and looks at the file. It doesn't see a complete frame, and chokes. IMHO, the fix for this should probably come from Impulse — I think Imagine should render its temp files (*I am assuming it's rendering the first field to some sort of temp file, probably in the subproject's output directory *) to a TEMP directory (maybe where quickrenders are rendered?) and then after the second field is rendered and the final frame is assembled, write the final frame to the subproject output directory. This will probably involve some increased memory overhead, and a bit of processing time, but it need not be significant. As for ADPro, my thinking is that Sentry can wait for the frame to be completed (both fields) and then do the copying to DDR:woof or whatever your PAR directory is. Then the PAR won't get confused, because it's getting the final file. I don't have 3.1 for the Amiga — I have 3.1 for the PC, and copy rendered frames from the PC to the Amiga over a network, to the PAR — but I *think* Sentry, or possibly REND105 (or whatever number it's up to) should be able to wait for the completed frames and then send them to the PAR. I can try and cobble up some AREXX code (kinda blind, but I'll give it a shot) if you want, for use with Sentry. Jim Shields Regal Rodent Development