Differences in scan dimensions

[Doug Nelson]

I was just experimenting with a heavily-textured print to see if any sort of standard methodology could be developed for removing said texture when I noticed something peculiar:

I scanned the same image twice, using the same settings, the second time at 90 degrees from the first. I then carefully rotated them both so they had exactly the same aspect. Then I dragged one over to the other and applied difference mode for precise alignment. I found I couldn't align all sections of the image at the same time. It turns out one of the images was slightly narrower than the other. Not so much as you'd notice with a single scan, but enough to make precise alignment impossible.

Has anyone else noticed this? Is this from the scanner, or from the rotation in Photoshop?

[Doug Nelson]

The scanner is a new Epson 3200. I'm still playing with this as I type this, but right now I'm suspecting it's something in the rotation.

I did discover this one little tip for anyone trying this: for alignment scans only, and not for general work, scan at your maximum-rated scanner resolution. It's slow, sure, but at lower resolutions a one-pixel shift can be way too much for precise alignment, and once you have the layers aligned you can downsample to your normal working resolution.

[jeaniesa]

Originally posted by Doug Nelson
I did discover this one little tip for anyone trying this: for alignment scans only, and not for general work, scan at your maximum-rated scanner resolution. It's slow, sure, but at lower resolutions a one-pixel shift can be way too much for precise alignment, and once you have the layers aligned you can downsample to your normal working resolution.

EXCELLENT tip Doug! I never thought of that, but it makes perfect sense. I've run into similar problems as you're describing and usually just use Free Transform to get "close enough". It works OK, but given that I'm a bit of a perfectionist I'm always looking for better ideas!

Jeanie