Definitive reference on 'Kerf' setting?

I’ve been using Lightburn a bit over a year now and I’m starting to make some mechanisms that require high tolerances. So I’m going through and trying to calibrate my laser to a higher degree of accuracy. I calculated it’s kerf using one of the well known methods to the 1/1000 of a mm. The thing I can’t definitively seem to figure out is if we should put the entire calculated kerf value in and it gets cut in half in software, or should we divide the kerf in half prior to entering it??

I ask this because I’m assuming that, on a microscopic level, the kerf should be split between the left and right of the travel path. And the name Kerf really means the entire width of the removed material. The input field doesn’t say half of the kerf, just ‘Kerf’, hence my confusion. Any definitive answers from the developers would be super appreciated. Thanks.

The kerf adjustment value should be half the measured kerf. It’s stating how much outward/inward the cut needs to be to account for the kerf.

This would be an extraordinarily narrow kerf. I would argue to the point where no kerf adjustment would be necessary for most materials.

What material are you testing this for? Kerf values for wood materials are typically in the tenths of millimeters range.

Those tolerances are pretty extreme. Even my pro cnc machines have a repeat accuracy of .004".
1/1000 of a mm is orders of magnitude higher than that at .00004" well beyond the repeat accuracy of most laser and cnc machines.

1 Like

What well known method…?

This is mdf and caliper is reading 10 times the actual kerf… of 0.164mm on my OMTech China Blue 50W (or so they say)… 2" lens…

The fiber has a much more capable with engraving and it’s dot is ~50uM…


Boy, I really miscommunicated here, apologies. No, my kerf is definitely NOT 1/1000 mm (boy don’t I wish!). If that was the case, I wouldn’t even be bothering. What I meant to say is the tool I’m measuring this all with is accurate to that place. And ‘jkwilborn’, you ask what the well known method is and then post the picture of said well known method. I was measuring the kerf of cut acrylic on a 60watt chinese laser exactly as you have illustrated. My value was similar to what you are showing in your photo.

My main confusion just came down to the input field in Lightburn being labeled ‘KERF’ which seems to indicate that one would input the actual kerf value. Nowhere in the docs does it state any detailed description of calculating it, nor that the value should be input as half the calculated value. Maybe if the input field was called ‘Kerf Offset’ or ‘Kerf Compensation’ it would be a bit more intuitive. That is all. Thanks everyone for the replies.

Nothing I have done to date has really ever required things to be so accurate, but now I’m trying my hand at making small mechanisms with gears I am hoping to cut on the laser. So at the scale I’m working at, being off by 0.5mm would be a problem. So just looking for any gains I can get.

The wording is actually “Kerf offset”. The offset is the clue here. It basically uses the same mechanism as Offset Shapes tool. Which implies that it’s going inward or outward the amount specified, which would need to be half the measured kerf.


Depending on your use case of cutting and fitting parts, you may also explore “Offset Shapes” to make your parts fit better. The setting for “Inward” and Outward" distances and whether the settings apply to inner and/or outside shapes are also available on a part-by-part basis.


totally agree… I have had plenty of times I have to offset by a slight amount for fitting parts. I do mostly cnc cutting and a fair amount of laser cutting and sometimes just making that part as many times as it takes to dial in my offsets… I have most of these offsets memorized for my workflow now and fitment is not as much of an issue after years of practice.

If the kerf is computed correctly, it allows me to make very snug parts without much of a problem. I used to cut small parts out to figure out a snug kerf to make it fit, but found using 1/2 the kerf on both halves does an excellent job in a myriad of applications.


I don’t suppose there is a mechanism in Lightburn to manage rectangular beam shapes?

I can get one axis perfect but the other is sloppy. Right now I’m either adjusting the model after I know layout orientation (tedious) or setting a happy medium kerf offset that leave one axis a bit too tight and the other a bit too loose.

For kerf, no. But there’s a similar concept called “dot width” that is meant to adjust for width of dot when doing engravings. It accounts for the bluntness of the dot rather than assuming that the dot is infinitely small.

Another strategy that I’ve employed is to rotate your design such that you’re cutting at a 45 degree diagonal. This effectively averages the rectangular kerf. It’s most effective when most of your cuts are at right angles.

I’ll have to check into the “dot width” control. I haven’t noticed it.

I have considered model rotation and will explore that as this project moves from prototyping to production. But, as you said, it’s highly dependant on model geometry.

Most machines have the X axes or head/drag chain riding on the Y carriage meaning the Y axes can’t be as fast as the X axes. When you change the X axes in some other direction where Y needs to move, it usually slows down the whole process…

About 20:20 minutes into this video they speak to the Lightburn developer that created the dot width adjust function. He explains it very well…

The whole video is very good.

And a quick Oz explanation…

Good luck


Noted, and thank you. Also for the link. Will watch later.

Where careful kerf adjustments are required, I’d prefer precision at the expense of speed. Of course, some projects don’t require 0.1 mm precision, so “Full ahead, Mr. Sulu.”

Your digital calipers may well show increases of 0.001 MM, but I would defy anyone to actually measure precisely any item either internal, or external to that precision. Many factors involved here, temperature, humidity, error of measurement of instrument, operator feel, condition of instrument , period of calibration, to name but a few.
Personally, I would use a good set of feeler guages to measure the kerf width, but even at that you simply wont get any better than 0.01 mm, if even that precise, with standard readily available measuring equipment,

1 Like

I don’t think it matters if you can measure to the thousands or not, that’s probably more accurate than you can use anyway…

Maybe you forgot what it took you to learn how to actually use feeler gauges with good repeatability and accuracy?

Unlike a caliper, there is much more of a feel to a feeler gauge that you only learn by trail and error…

They are of no use if you can’t physically get them into the area you need to measure… Measuring anything other than a gap/clearance, put the feeler gauges away and use some other tool…

You can be a total klutz and drive a digital gauge with usable accuracy… pretty much out of the package. All you need to is figure out how to turn it on… then you can easily select metric or imperial…

With a feeler gauge, you have to pick metric or imperial steps, not to mention basic addition when it’s somewhere in between…

The low end cost of feeler gauges are around $9 US, of course you probably need two to cover both unit.

A digital caliper can be had for about $8 US…

I totally understand that it’s hard to change when something has worked so well for you for decades… been there done that…

Couldn’t imagine using digital calipers when I set the valve rocker arm/stem gaps…

Whatever works best is probably a good choice…


Here is a great video on Kerf