Issues with precision

Two problems I’d like some input one;

  1. Visual alignment expectations. after a job is done and the bed image refreshes the placed art is never exactly over the cut, and as a result the ability to do any back to front or secondarily placed art is impossible (As in cutting a shape, flipping it and aligning another image with the shape by camera only. I understand the suggestion of a jig for this, but should I really expect the alignment to be off by 1/8 to 1/4? Why is there no way to manually calibrate the head?

  2. Scale issues. I am working on a project that requires 2 acrylic parts to be bolted back to back. In the design (Created in Illustrator CC) I created the proper art to engrave (countersink) the holes for the 6 sided nut and the bolt. I measured the nut and bolt with precision calipers and created the corresponding art to the exact dimensions (allowing for kerf and ability to slip inside), and despite working and placing the design at 100%, the holes are no where near the dimensions specified in the design. Its almost as if GFUI has shrunk my design by nearly 10% (I can’t be sure of the % because GFUI doesn’t report the re-scale attributes. Am I missing something? Some setting that would make a cut exactly as I created it? Is there a DPI or LPI setting I need to adjust for Illustrator to avoid an interpolation from design to cut?


Scale issues are almost always the result of the “save as” settings for SVGs in the design software. The Matrix has great info on this, but the most likely culprit is “Responsive” being checked – it must be unchecked for the scale to be unaffected.

Updated: Here’s the deep link…


If you set up your artboard at 12x20, it should fix your scaling issue.


A couple of things I have noticed:

Anytime you are depending on a wide view camera to look at something there is going to be parallax issues. At one point I tried to take a long distance photo of a building and draw an addition on it and discovered even, in that case, there were some real issues that you would not notice unless you were matching a drawing. the Glowforge folk have made an effort to keep it from being crazy but real precision of that type would be asking a lot. Add to that that calculations for distance from the camera and the tiniest amount off makes a big difference in the result. After you have cut something tell the GFUI that the thickness is something different and look at how far everything is off vs the correct thickness.

The most accurate view will be right under the camera, and where possible that is where I try to center my work but that is not always possible. so I try to do a 10pew500zoom trial run to see where the real math is. I am even thinking a rectangle around complex stuff just for that and put on ignore all the other times. Once I have that relationship a bit of “Kentucky windage” will sharpen the result and most of the time multiple runs will fall exactly on where they ran before. I have had it otherwise but I have discovered that the undo is too easy and discovered I “undid” an important move without meaning to do that or realizing it.

As noted above distance from the camera makes a big difference in where it thinks things are. I see that they have now included a scale in the GFUI that allows you to say how big your design is. It is limited in decimal points of accuracy but is closer to reality than what you see on the bed of the crumb tray. As before a light 30-500 test run will increase precision some but accuracy in thickness will be needed as well. short of a per design jig the setup I started in the Precision thread here wrks very well for repeatability.


If you are using Illustrator and you turn off Responsive you shouldn’t need to do this anymore. I still use the 12x20 artboard for placement, but it isn’t needed for scaling.

1 Like

Oh good to know, so just in the export?
On a side note, do you use the Save As option or Export option for saving a svg in illustrator?

“Save as…” is the preferred method, yielding a cleaner svg file.

1 Like

As @dwardio said, Save As is the preferred method, but I typically use the Export functions; you need to make sure that the options are correct when using Export as it seems to be more prone to break in the GFUI if not using very specific options. I’m on mobile at the moment, so don’t have a link, but @chris1 has posted a number of helpful hints on using Export if you want to try it.


I’ve been doing some double sided stuff lately. I engrave then cut on the first side, then pop out & flip the cut pieces & put back into the cutouts without moving the main part of the original material, then run my 2nd side engrave. I place the art for both sides right on top of one another in Designer, set to different colors & export all as a single SVG. When I engrave/cut side A, I set the side B art to ignore, then ignore the side A art when I engrave side B.

Of course this works for me because my cuts are just circles and placement of the the side B engraving isn’t exactly critical. But in a case like this, it works great.

You can get the orientation exact by wrapping the entire thing in a square and doing this:

Engrave 1, cut square.


Engrave 2 cut real outline (circle)

The square cutout keeps your orientation during the flip.

Of course you can do any outershape with the correct symmetry acrosss your flip axis. Rectangles, isosceles triangles, etc.

Pretty sure @Jules wrote this up in a thread somewhere. Like “perfect yes perfect” alignment… let’s see… aha here we go:


ok, thank you all for your suggestions - I will go through them each and run tests - HOWEVER, I will say that EXPORT from Illustrator, for me, has produced better results than SAVE AS. Also, I have always made sure to exclude ‘Responsive’ when exporting, but alas, the issue has persisted. I will try some of the more specific suggestions to rule out ‘human error’ though.

As for jigs and flips and centered art - All valid [mathematical] answers… However, a lot of what I do are small items cut in quantity from large sheets, so starting in the center sort of kills the plane to always cut in the center - I need to reliably cut the extreme edges. Also part of what I engrave on the back is “made by…” and isn’t always the very next thing, so a flip isn’t always happening as step two - And it sounds like I need to align the back side engrave with the front side engrave (rather than the camera view of the back. I totally get it, I understand parallax and the affect it has on alignment, I guess I just assumed that with known lenses and bed size the math would be able to correct for that (I mean, Photoshop can correct for wide angle parallax for a variety of lenses, shouldn’t GF be able to handle one?)

So, if I use the 12x20 hack, can I then place my art where I’d like it and expect scale to be correct? What’s the plan for smaller than 12x20 materials? Do I zero them out, or center them? Should I make a measured jig to accommodate the offset placement of the smaller material?

The assumption in the forum is that this will eventually be fixed; I believe it has officially been stated as being in the “hopper.” Obviously, there are many folks that are frustrated that this is not fixed yet, but workarounds have been developed. It’s unfortunate they are needed, but they are there.

Yes. You can then move it around manually in the GFUI if needed.

If I understand the question correctly, either approach would work. I don’t like to rely on the camera, so when I need precise placement, I tend to cut a jig out of cardboard and then place my object there. You still have the potential for a kerf-width of error in 2 dimensions, but for my purposes, that’s acceptable.

I have started to just put a square or rectangle around the limits of the work and first tie the work to the corne tjatbi want it to land on and print at 5-500 and run that. There will be a tiny mark in each corner and you grab the actual corner it lands on to the place that you want and put the rectangle on ignore .

Using that I have made many cuts within a mm of the edge.

1 Like

Just add my solution, though I think it’s been suggested many times before.
I cover the honeycomb with a piece of paper, taped down along all the edges, then run the file as a low power vector/score or whatever.
That gives me a registration for however many pieces (or 1 piece) I need to load, or covering the work area with bits of random scrap. Pin them down where necessary, and go.

1 Like

Do you set the elevation to zero? a change in thickness can have a huge effect.

Yes, or possibly 0.008", roughly the thickness of my large sheets of paper I use for this.
Then I generally add that to the goofgrade I use, in the manual settings.


I make gears from scrap frequently and this is where I made one from escutcheon button

So I’ve mentioned this concept before but here goes again:

As an economic argument in almost all cases where you’re lasering non-unique materials (I.e., not something like a laptop or your very last little bit of a certain material), it just doesn’t make sense to screw around with super fine placement.

Granted there are cases where you’ll need to do so, but take something like BB ply. I source mine for about a dollar for a full sheet. You’ll need to figure out your own hourly rate, but that entire sheet by itself is worth less than a minute of my time. A 4x4” piece? economically that’s worth literally a few seconds.

Edit: I did the math, my material cost for 4x4” piece of B.B. ply is about 7 cents US. Granted this is simplifying and true cost includes time to procure and dispose and so on, but… seven cents. If I wanted to cut something that was even close, I’d grab a larger piece of scrap or a whole new sheet and be done with it.

Now I’m a New Englander and as such frugality is in my bones, but there are definitely practical limits to how much time you should spend worrying about saving a millimeter of material here, a quarter of an inch there.

And yes one more time: I do understand that there are times where squeezing everything out of your material is important — that’s where jigs come in — I’m just saying that I find that it’s just not needed in the majority of cases. Ymmv.


I’m with you on this. I have the similar feeling about scraps - the time it takes me to hunt around for a piece of scrap that has the right size of unused material left to do a project just isn’t worth it. I use a piece multiple times but not much after there aren’t fairly big pieces of unused material left. I probably only touch a piece 2 or 3 times before it hits the toss it pile. I need kindling for the pizza oven anyway :slightly_smiling_face:

1 Like

Thanks for the help in this thread, all! I’ve moved this thread to Beyond the Manual so the discussion can continue there.