Dimensional Scanner Distortion

[Josh T.]

Glad to be able to provide some useful info. And thanks for the link. For the time being I'll probably keep photos centered to keep distortion minimized.

But it may be interesting to so some experimenting with ImageMagick at some point down the road, probably scanning the same image horizontally and vertically but placed the same distance from the side of the platen, apply the approximate corrective distortion, rotate one image, then directly compare the results by flipping back and forth in the viewer.

 

[grepnold]

I'm using ImageMagick API calls in my application so I can include correction if I need to. I'll be sure to let you know if I do.

As a start, like you I'll just use the centre, and I'll align the longer dimension with the vertical axis, to minimise these errors.

 

[grepnold]

I recently bought a CIS scanner that doesn't have this problem, as a previous poster pointed out. CIS scanners use a sensor the width of the image and a linear trough-like mirror in place of a lens, so don't suffer from pincushion and barrel distortion. But in many other ways they're inferior, with very poor depth of field - if the image isn't in complete contact with the glass it will be out of focus. The sensor travel also sometimes judders or slips, so there's the occasional squashing of a couple of scanlines along the direction of movement.

 

[corkyg]

Why bring up a 2 year old necro post?

 

[grepnold]

Why bring up a 2 year old necro post?

CorkyG, because this was and I believe still is the only discussion on the web of a problem that will affect all who require dimensionally accurate scans. It has resulted in a workable solution for me, and has suggested other ways to correct for this error that others may wish to try and to report back on. In short, it is a contribution to knowledge that could result in improved techniques.