On 14 Oct 2003 at 0:04, Helmut Walle wrote: > Yes, but I would not expect any overall processing time advantage from > this. If you have a parallel port scanner (?)
Sorry, should have mentioned the hardware: Athlon 2k and USB2 scanner. It is normally *fast*, but what makes it slow in this case, is the image size. I scan each negative frame at the scanner's max resolution (1600dpi). This slows the scanner down significantly. Then, when I start with correction, the GIMP takes its time too, also due to the image size. So what I get is the scanner working for a minute or so but the GIMP is idle, and then the GIMP does it bit for a minute or so while the scanner sits idle. Which is why I want the scanner to start with the next scan, before starting the GIMP. > And, just out of curiosity, how do you do colour correction? Do you > have a densitometre? Do you really correct finely (10 to 20 > interpolated steps per colour)? Or is it more a quick visual > inspection, and "turn up green a bit"? A number of my negs were scanned on a Fuji mini digital lab, and the colours of these scans are fairly accurate. I used these images as a baseline, and then scanned the same images with my own scanner. A friend and I sat down and worked out the best "levels"-based correction for the whole lot, which gave me four sets of correction numbers (overall gamma first, and then gamma & highlight- point & shadow-point for R, G and B). So, yes, the process is visual and not perfect, and I will have to fine-tune it significantly. But it gives *much* better results than Xsane's attempt at correction for negs. rgds Jaco
