Testing focus, or what's 'autofocus' for?

Having just cleaned Glowfinger, as a new year treat, I thought I’d try to determine how fine a line I could ‘cut’ in the surface of a piece of mountboard, with a measured thickness of 1.45mm.
Set up a simple 5 lines piece of a/w, and entered a series of low power settings - 1 - 5, and run at 500 speed.
Before cleaning, the lines were a bit intermittent, and after cleaning much better. However, I wasn’t impressed by the width of the marked line, so I changed the focus height in a series of tests, from the 0.1mm up to 10mm, and found no difference !
What I’m trying to achieve is a bit more knowledge on just how fine a ‘cut’ line I can mark on the surface, and what effect de-focussing a given amount has on that width.
Has anyone else ‘been there and done that’, or any comments to throw into the pot.
John :upside_down_face:
PS Happy New Year to one and all.


Are you changing the focus on the cut setting, or the material height? The latter only affects how the image is dewarped, but it will still use the autofocus for cutting, if I understand correctly.

1 Like

Aaaaaagh !
You mean for the last year I’ve been setting the cut height only in the material box, and I should have been using the drop down settings window ?
I’d assumed that the material height overruled that !!!
Thanks for the clarification.

No, hang on - the material height does set the cuts to the right height.
I now see that I can also change individual lines, but that’s not what I’ve been doing.
I change the set of five lines for each test, and then I find no difference, so my blushes are slightly unnecessary !
J :upside_down_face:ohn

Maybe I just added confusion, but it is confusing. If material thickness and focus height are the same, the Glowforge will actually ignore the focus setting and use its autofocus. So for example, you could put in 3mm material, set the thickness to 20mm (which also sets focus to 20mm) and it would still focus correctly because the autofocus would lock on to the top of the piece regardless of what the settings say.

Manual focus only takes effect when the focus height you enter is different from the material height. There’s another possible explanation for your test results: the stepper motor that does the focusing moves in fairly course steps… I’m not sure how far you have to tweak it for it to make a difference.


OK, I think I have a handle on the possibilities now, with your latest reply.
I’ll give it another test with the 'material set to the correct 1.45mm, but set individual heights of the five lines differently, and let you know.

1 Like

Actual thickness 1.45mm
no obvious difference when the ‘Focus height’ is set to 0.5mm to 2mm, but 3mm gives finer line, 6mm thicker and 12mm thickest.
so more experimenting to do, but making progress now.
John :upside_down_face:

Pip explains it here:


Hi Jules.
Yes, that’s useful.
My test results show an undetectable difference in the ‘cut’ width, within a focus variation of + or- 0.5mm, so I’m not sure about the accuracy needed in measuring the thickness, to be honest.
I’ll try agin when I have time, to see if it makes the same, little or no difference, when I’m actually cutting through, or doing a ‘kiss’ cut through my material. So far, I suspect power/speed to be far more important.

John :upside_down_face:

1 Like

My final result shows that there is a slight difference in the cut though width for different manual focus hight settings, but that my finest cut is made at +0.5mm on the measured thickness.
I’ve been using the same power/speed setting for the last tests, 300/full power for picture mountboard of 1.45mm thickness, and I shall have to determine if the same variation is true for kiss cutting, when I increase the speed slightly.
John :upside_down_face:

The thinnest line will be seen with a score at the proper material thickness as a height measurement.
However, now that you know there is an override for the height, you can have some fun.

Here is an experiment I tried, just to see how the results came out.
The big deal in doing a large project like this is two fold.
Faster design (can even be a hand drawn design scanned in).
And way faster score versus engrave time.


Thanks for that link to your earlier thread. I’ve been moving in a similar direction, coming to the 'forge from a CorelDraw background. I’m used to manipulating nodes to get drawings done starting with simple shapes and simple elements, but defocusing to thicken up lines will be a good step forward, so thanks again.
John :upside_down_face: