Cuts more than 1/4" off / warp in preview image


#1

cuts are more than 1/4" off where the screen predicts and the bottom of my image is showing as wicked warped.


#2

same file, different piece of leather, more weird results.

waste is mitigated by my alignment box, I run it at 500 speed 1 power and it barely marks the masking letting me know where the laser is actually going to fire.


#3

Hmmmm… So a couple of things…

  1. Sorry to hear you’re having a problem.
  2. Please include the rulers and magnification like you did in the 1st pic so we can get an idea of scale.
  3. Can you test on something more rigid? If you don’t want to waste any non-scrap material, how about a nice flat sheet of cardboard held down with magnets? Thankfully Amazon supplies me a seemingly never-ending supply of it. I’ve managed to get some good single 20" x 12" sheets out of it that are always perfect for testing for me.

#4

Just some things to check:

  1. Often the preview image in the GFUI will be shifted after the cut. People have noted this when doing multi-part cut/engraves. It would be interesting to see what would happen if you did this:
    1. Manually mark on the masking where you are expecting the cuts to be.
    2. Align the preview to your marks, focusing on the center.
    3. Run your light engrave.
    4. Then measure the difference in position between your manual markings and the actual engraved marks.
  2. I see “Standard Natural Leather” selected, but with the blue painters tape, I’m assuming (please correct me if I’m wrong) that this is not proofgrade.
  3. Have you used micrometer calipers to measure the thickness of the leather + masking and entered it in the GFUI. Not having the exact thickness will cause larger than normal differences in placement vs. cut.
  4. What browser are you using? Chrome seems to be the best-supported.
  5. You might want to try placing some (thin) neodymium magnets in the middle of the leather to hold it down across the full bed. Warping of the material will cause problems with visual alignment–and can also affect the cut/engrave.

#5

To your points
1, I have a bounding box cut that I run at 500 speed and 1 power that marks the material where the GFUI thinks it is going to cut, depending on how off this is, I then adjust the material accordingly. It’s the same amount off as the final cuts.
2. the leather isn’t proofgrade, but the thickness is the same.
3. I measured it with dial calipers and it’s within a 1/32 of an inch in thickness to proofgrade leather
4. Using chrome
5. Leather sits flat and isn’t buckled. I can press on the sheet the full width of the bed and get no defection (when over the honeycomb), the leather is super pliable

Tom
I can try a new test, but I shut my GF down after each session and the alignment issues aren’t consistent as the calibration changes every time the unit starts up.


#6

This is exactly how my glowforge performs and I’ve had two support tickets on it. Both times they said it was behaving as expected and that software would eventually resolve the issue.


#7

Just an FYI for terminology (so that people don’t get confused) a micrometer and a caliper are different things; its not a type of caliper.


#8

While I was using “micrometer caliper” in a way that was not correct for what I intended, micrometer is more formally a subset of caliper.

From Wikipedia:

A caliper using a calibrated screw for measurement, rather than a slide, is called a micrometer caliper or, more often, simply a micrometer. (Sometimes the term caliper, referring to any other type in this article, is held in contradistinction to micrometer.)

What I was meaning to convey is using a caliper that is accurate in the micrometer range (micrometer in this use being the unit of measure :blush:)

So maybe we both learned something today.

[edited to add link for the Wikipedia citation]


#9

From what I’ve read, 1/32 is enough to throw the fisheye correction off significantly at the edges.

Try selecting the proofgrade as a starting point and then switching to manual mode and entering the accurate thickness.


#10

Also knew that, but I assumed you didnt actually mean you wanted them to pull out a 0-1 mic and get a measurement to 0.0001". It could just cause confusion for someone googling and thinking they need to buy an expensive tool to measure with.

Edit: realized that sounded a like dicky; I was just thinking because many people on the forums are new to this kind of work and have never used anything except a tape measure so they wouldnt question whatever google tells them


#11

No problem. I didn’t take it that way.


#12

Here is how my view changes going from the default standard leather (.069") to 1/32 thicker (.1"):


#15

Calm down Frances*, all I was suggesting is that being a tad bit less pedantic could help your reception.
The distinction is important when you got to the tool crib or order one but for general conversation, it is not polite.
Also, when someone says “clip” when they mean “magazine”, … the list goes on.

*I know your name is not Frances, it’s a movie quote.


#16

Reminded why I generally no longer participate in forums except to ask questions.


#17

(That’s why I just typed “dial calipers” micrometer was wrong but not worth correcting)


#18

Looks like this shifts your view by ~1/16" which doesn’t account for my 1/4"+9 offset.


#19

What is the +9?

No - it doesn’t account for your offset but it seems that it would certainly contribute if you’re already starting 1/16" off in the view, no?

I don’t see anything exceeding 1/4" in your first image posted? Every which way I was able to measure it came out to maybe right at 1/4", but looks just a hair under.


#20

That was meant to be 1/4"+ the 9 was just a fat finger typo.

Yellow lines show the gap of ~5/16" and they also show the weird warping effects I’m getting, the cuts in the file as well as the resulting cuts were flat an true.


#21

Has anyone done some optical alignment testing? that is, put a known grid onto the cutting bed and see how the camera sees it at various heights/material thicknesses? The place a drawing in a known position on the graph ( zooming and tweaking), cut it, and see were the cut actually is compared to the position of the drawing in the GFUI? I’d think we could map the error pretty quickly - or has that already been done?


#22

Yep. @cynd11 I think. She did a whole write up on how she figured out how many clicks to move her designs to match her material view. She probably knows when/title otherwise it might be a bit of a search to find it but it’s out there.