Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
distortion when enlarging with a photocopier
#1
I tried scale up a plan to double the size with a photocopier (on a Xerox WorkCentre Pro 128). What I found is that it was distorted in one direction. The width was accurate but the length was increased by about 2% (there is an accurate grid on the plan). Given that there has to be a lens in there somewhere, I can understand that it happens. The question is whether there are simple ways of getting around this issue. Do any copiers with a zoom function have a way of correcting the distortion? I would guess that if the copier accepted an original in the center of the scanning area instead of at the edges that that would pretty much solve it.

(fwiw the plan is a guitar outline to scale and I would like to get it as accurate as possible as my idea is to stick the enlarged plan to thin plywood and cut along the line to make my template).
Reply
#2
Scale up 1.41x

Then rotate the new larger copy by 90 degree and enlarge again by 1.41x

The output should be the same size in both directions, roughly 2x, but perfectly equal.

The trick is that now you apply the distortion on the X axis the first time, then on the Y axis the second time.

Pretty slick, huh?
Reply
#3
[quote simonm]I tried scale up a plan to double the size with a photocopier (on a Xerox WorkCentre Pro 128). What I found is that it was distorted in one direction. The width was accurate but the length was increased by about 2% (there is an accurate grid on the plan). Given that there has to be a lens in there somewhere, I can understand that it happens. The question is whether there are simple ways of getting around this issue. Do any copiers with a zoom function have a way of correcting the distortion? I would guess that if the copier accepted an original in the center of the scanning area instead of at the edges that that would pretty much solve it.

(fwiw the plan is a guitar outline to scale and I would like to get it as accurate as possible as my idea is to stick the enlarged plan to thin plywood and cut along the line to make my template).
not necessarily, it's not a simple lens, remember, there is a slide that moves back and forth each time you scan/copy. There are probably one or more cylindrical lenses in there, it's not as simple as enlarging an image with a normal lens.

But the trick I posted above should work.
Reply
#4
That is slick.

Technically I think the formula is sqrt(magnification% - distortion%). It turns out that sqrt(200%) = sqrt(2) = 1.41 -- no distortion -- rounded to two digits is the same as sqrt(200%-2%) = sqrt (198%) = sqrt (1.98) = 1.41 -- your distortion.
Reply
#5
I won't be back in the office for a week, but I will give it a try then. I know I will have a battle with the automatic paper size recognition (rtm .. Wink) but if it works it would be great.

Thanks. I will post after I try it.
Reply
#6
[quote space-time]

Pretty slick, huh?

who taught you that?
Reply
#7
Most current Xerox copiers, and this applies to many other makes, no longer have a lens used in the process at all. They use a scanner bar and turn the image into a digital format similar to any ordinary flat bed scanner. Some have a fixed scanner bar for the document feeder as well, the motion is supplied by the feeder. Depending on the programming used internally to create the printed image, distortion can occur. Especially if the speed/displacement of the moving bar is any different from what is programmed in. Sometimes this is intentional, to make copies detectably different. All this makes it easy for adding options such as fax, document scanning to save to computer, and various other image modification.

In any case, space-time's suggestion should work. But the problem does come up in copying other material that must be exact. For that there are specific models designed from the ground up to give accurate renderings in both dimensions.
Reply
#8
[quote space-time]Scale up 1.41x

Then rotate the new larger copy by 90 degree and enlarge again by 1.41x

The output should be the same size in both directions, roughly 2x, but perfectly equal.

The trick is that now you apply the distortion on the X axis the first time, then on the Y axis the second time.

Pretty slick, huh?
Thanks for the excellent tip.

Finally got round to trying it today. Much better results than with my first method: more accurate in actual size and the distorion is essentially identical in each dimention. Well within the tolerances for my guitar project.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)