Anamorphism and 0,0 point errors in SCAL

[Brian Lawler]

Hi all,

Having solved my jigsaw puzzle problem, I have another one for you...

I tested SCAL for precision and positional accuracy, and I have found errors on both counts.

1. A rectangle that should measure 11.5 x 23.5 in. comes out slightly smaller, specifically 1.1% short on the 11.5 in dimension (short), and .5% on the long dimension.

2. The 0,0 point is not even close. I find an error of .19 in on the X dimension, and .41 on the Y dimension. So I now put those values into my SCAL files to put the upper-right corner at the right place.

Am I alone in experiencing this behavior, or is this "normal?"

(I fix the anamorphic problem by scaling my placed SVG files to 98.9% and 99.5% respectively. With those values, I get results that are nearly perfect.)

[jasenj1]

Hmm.. That's interesting, I never tried scaling things. I figured SCAL was just "drifting". There's a thread around here somewhere where people have pointed out the Cricut and/or SCAL is not very accurate across the length of the mat. When I first got the Cricut, I tried doing several "print & cut" projects, but due to the inaccuracy I gave up.

Here's a thread discussing the cutting inaccuracies.

- Jasen.

[esheen]

I asked a question a few days ago about the size issue. I was trying to make a double mat for a Christmas card and was using the rectangle shape. When I would put the length and width into the program and cut, it was a bit shorter on the one side. I should have been able to input the dimensions and have two perfect mats but instead I had to put one on top of the other and keep nudging the length. I thought that by cutting the mats on the cricut I would have perfectly sized mats! What frustration!

[Brian Lawler]

The way I tested this was to make a rectangle in Adobe Illustrator at exactly 11.5 x 23.5 inches in size.

Then I saved the rectangle as an SVG file, placed it in SCAL, and ran the cutting on the Expression. Then I measured the resulting cut and calculated the difference.

It's almost exactly 98.9% on the short dimension, and 99.5% on the long dimension. You can enter those values into SCAL, and it will correct for the error.

By making my dielines distorted in Illustrator (before saving them as SVGs) it solves the problem.

The 0,0 point issue is no big deal, except that you would expect 0,0 to be the exact corner of the mat. It's not. As I mentioned in my posting here, the error seems to be .19 from the top, and .41 from the edge. If SCAL allowed for greater numerical precision in the software, this would be 0.1875 on the top, and 0.4125 on the edge.

I have tried both center-aligned cutting, and right-top edge aligned cutting. The right-top technique is more predictable.

Expression is very accurate, if you start the mat carefully on the left edge of the machine. I also tested that. I cut the same pattern several times (4 or 5 times), ejecting the mat, then replacing it again, and cutting again. It cut virtually in the same place each time, so that's very impressive.

There is no way for the Expression to know what mat is in it. There is no bar code, nor any reading mechanism in the Expression. It just uses left-top start as its beginning point, and it counts from there. I would guess it has an optical encoder on the end of the top feed roller; that's the logical method to keep track of the mat's location.

I would also guess its mechanical precision is about ±0.025 in. It may be coarser.

My only real gripe is that sharp edges become rounded edges because the machine drags the blade around corners.

More expensive machines like the Gerber vinyl cutters, and the Kongsberg cutters ($85,000) are much more precise. When they make a sharp corner, they lift the blade, turn the blade, and start cutting again. It's a much better approach to precise cutting.

But, I paid just over $200 for my Expression, and I got much more than I paid for it. It's an exceptional machine with many uses. SCAL makes it really valuable to me.

Brian

[talanhart]

Brian,
We use CNC mills etc, here at work, and for the price we paid for our Cricuts, I think it is an awesome little machine. I have had mine for over a year now and I have cut sheets and sheets of vinyl that take almost an hour to cut. Love my Cricut.

[firstcut]

Brian,
I did a lot of work on this and my solution was to shift and scale a virtual mat in Inkscape so that i didn't have to do any math. But these solutions have their limits because the "creep factor" with the Cricut isn't linear. The error builds as you get further from the blade starting point. When I ran similar tests to your I got varying scale factors depending on the size and location of the cut.
The linear type of calibration you and I are using could easily be added into SCAL, however, and I put it on the wishlist a good while ago and would appreciate any "amens" that might add it to Todd's to do list.

Here are some threads you might want to see
http://forums.surecutsalot.com/viewtopic.php?f=8&t=2956&

http://forums.surecutsalot.com/viewtopi ... 220#p24220

http://forums.surecutsalot.com/viewtopic.php?f=6&t=4489

Most of us don't get very good repeatability between mat loadings so I am wondering if that is something that has been improved on new machines perhaps. That would be good news indeed.

BTW, welcome to the craftedge board! Pleasure to have you here.