<<>> Something from San Diego Supercomputer Center: Tool Description --------------- ------------------------------------------------ imadjust Adjust image colors, desaturate, brighten, etc. imcat Concatenate images into multi-image files imcltroll Cycle a color lookup table imcomp Digitally composite images imconv Convert between all supported image file formats imcopy Copy a portion of an image into a new file imfile Display image attributes imfill Fill a region of an image with a color or gradient imflip Flip an image horizontally or vertically imformats Describe image file format support imgray Convert to grayscale imhist Compute an image histogram immono Convert to monochrome impaste Paste an image into another imroll Cycle an image horizontally or vertically imrotate Free rotate an image imscale Scale an image up or down imshear Shear an image horizontally or vertically imsplit Split apart multi-image files For large ais files: -clear causes a bounds fault -center fails (it shouldn't shift the context, it should shift the AIS) ColorTrixDoc: need section on Warp and Replication Ct Copy: option to choose source from Preview Resample ct resamplefile works ok if src origin is not (0,0) but uses wrong translation(?) and leaves junk data at right edge of picture otherwise. Debug it on small pictures for speed. The problem is probably in ColorTrixFilterCommands.CtResampleFile or there's a small chance it's in ColorTrixFile.ReadFile. The window logic in CtResampleFile is tricky -- there are three windows to reconcile: the src file, the src window from the src file, and the dst window size. The allocated sample map has to be big enough for both source and dest. Call me if you have questions, as I'm taking a listing. One of the weird things about ResampleFile (note: the 's' is small because "resample" is a word but "re", in that context, is not. I think I'll puke next time I see SqRt, UnDo, or xSize. Now that I've got that out of my system...) is that I'm bending over backwards to accomodate Resample having a single maps argument, instead of two (a src and a dst), which would be more natural. In my version here I've got two maps (or the equivalent) and two windows. You might want to make that change to Resample (and the relevant sub-procedures). It should be quite easy. Just change: ScanlineRead[maps, ...] -> ScanlineRead[srcmap, ...] ScanlineWrite[maps, ...] -> ScanlineWrite[dstmap, ...] and then change the definition of overlap to: overlap: BOOL _ SameMaps[srcmap, dstmap] AND ax0