Have you looked at Panorama Tools? they have a gimp plugin and also a stand alone java app, very cool byt hard too get a handle on . hope this helps http://www.fh-furtwangen.de/~dersch/ Jason
On Thursday 07 March 2002 09:30, you wrote: > Greetings, > > I am a reletive newcomer to GIMP. I have searched the Manual and the > plugin registry to see if anything like this has already been done, but > without success. So I'm asking if anyone out there can help or point me > in the right direction. > > The PROJECT > ============ > I am in the process of archiving my collection of 300 or so vinyl > records. This includes recording them digitally, filtering out clicks, > rumble, etc. and burning to CD. I am also scanning the covers, labels > and inserts and burning them on a second session of the CD. And so the > whole record and packaging is preserved for posterity. > > The PROBLEM > ============ > The average album cover is over 12 inches square while my (average) > scanner can only handel about 8.5 x 11. So each side must be scanned as > four separate quadrants and then be merged back into one image. > > I have been using GIMP to reassemble the images with excellent results. > The problem is that it is taking me several days to complete each > images. Do the math, and it will take me several years to complete > this. I'm sure the process could be made faster with a little help from > the experts :-). > > The ISSUES > =========== > Why does it take so long just to load two images, move layer and > flatten? Well... > > -) The layers must line up pixel-perfect. Album covers were printed > with the same technology which prints pictures in the newspaper using > different sized dots of color to create a grayscale. If these tiny dots > don't line up perfectly then you get an interferience pattern visable > from across the room. > > -) After 30-40 years, the album covers are no longer square. So the one > image is usually rotated slightly from the other and must be corrected. > It takes me 2-3 tries to get this right. > > -) Since the scanner performs a new gamma correction on each scan, one > image is usually lighter than the other and must be brightness/contrast > corrected. It can take a few tries to get this right. > > -) Publishers tended to cram the backs of album covers with very fine > text. This means a resolution of 600ppi to adaquately reproduce. At > that resolution the images are quite large and it can take two minutes > for each of the steps above to be performed. And reset. And tried > again. Point is, everything happens real slow. Nothing happens in > real-time. > > The ULTIMATE SOLUTION > ====================== > I would like to start with a GIMP file containing a background and a > second layer. > > I would like to identify eight pixels within this file: > 1) A pixel from the top/left of the background > 2) That same pixel from the top/left of the layer > 3) A pixel from the bottom/right of the background > 4) That same pixel from the bottom/right of the layer > 5) A light pixel from the background > 6) That same light pixel from the layer > 7) A dark pixel from the background > 8) That same dark pixel from the layer > > I would then like to feed those eight pixel coordinates into a Script-Fu > which would then automaticly perform the translation, rotation, contrast > and brightness corrections. I could then give eyeball approval, flatten > and save. > > HELP > ===== > That all being said, what do I need? > > 1) Has any of this been done already? For example, does anyone know of > a Script-Fu which does the translation/rotation bit and I would only > have to add the contrast calculations to it. It's always easier to edit > than to create :-). > > 2) How do you get the RGB value of the pixel the cursor is pointing at? > I assume it is possible... > > 3) Given the last four pixels listed above, just what are the equations > used to calculate the brightness/contrast adjustments needed to > equalise? > > Thanks in advance for all your help. > > Bob Washburne > _______________________________________________ > Gimp-user mailing list > [EMAIL PROTECTED] > http://lists.xcf.berkeley.edu/mailman/listinfo/gimp-user _______________________________________________ Gimp-user mailing list [EMAIL PROTECTED] http://lists.xcf.berkeley.edu/mailman/listinfo/gimp-user
