A Treatise on Alignment

As one of the members who has been around the glowforge community since the beginning, as well as one of the admins of the facebook group, and a general laser enthusiast, I have heard (and experienced) quite a bit of consternation over alignment.

I wanted to take this opportunity to go over laser interfaces in general and where the glowforge falls in.

From the very beginning I have been interested in achieving the most accurate alignment possible when using the glowforge (or any other laser). A good portion of my early post history was dedicated to this topic. It is one of the main reasons I purchased a glowforge, and has become one of the reasons I still love my glowforge and tout its capabilities whenever I can.

Alignment Types

Having used a number of lasers ive found there are generally 4 ways one can attempt to get accurate alignment of engraves/cuts on a piece of material or object. They are as follows:

  • 0,0 corner placement + the use of measured offsets
  • head/red dot alignment based on the current location of the red dot on the work area
  • camera based optical system which images the bed and materials
  • fiducial placement via registration marks gathered by the camera system(s)

The first is the most basic and common method of aligning something. You find the home position of the laser head (usually the top left corner of the bed), and used that as the origin point for a cartesian based grid where you can measure offsets from that position in order to determine that your material is where the design is set on the virtual bed in the application you are using. Practically every laser has this type of system as a base (the glowforge, however does not 100%).

The second builds on this by allowing you to use a laser head equipped with red dot output to be moved over your material via jog controls. The position of the red dot acts as an anchor to which you can attach your image in the software, giving you better visualization of where the output will actually be. Many of these systems will also allow tracing of the outer perimeter of the job to be done to show the full material utilization which is very helpful.

The third is one we here are all familiar with, the camera based system. There are a few variations out there, including one that takes multiple photos of the bed and stitches them together (poorly), and some open source versions that use fisheye lenses similar to that of the glowforge. Being newer, most of these are still works in progress and provide semi-accurate placement, which is great for getting the best out of scrap materials etc.

The last and most accurate is fiducial/registration point based alignment using cameras placed on the laser head. A number of manufacturers have started using these to give incredibly accurate alignment for jobs. Special marks are placed on the workpiece which allow the camera to determine material position and rotation, and automatically adjust the job to match those values ensuring an incredibly accurate cut.


Many people, myself included, expect lasers to be incredibly accurate in every aspect. If something can engrave less than the width of a human hair, we should be able to easily cut a hair right down the middle right? Not quite.

While it’s easy to accurately plot laser output, one of the most difficult tasks is to have your material/object in the right place while that is happening. This was one of the major driving factors behind the camera systems that have been showing up in lasers recently, especially since the glowforge landed on the scene. As it is still technology in its infancy, it still has a good ways to go until its going to be able to turn 1 hair into 2 (split longways at least), or seamlessly match up 2 engraves without a gap or overlap.

Some users report up to 1/4" or more offset in the placement of their design on their glowforge to the actual output. This is a long way from the promise of easy, accurate placement we were promised at the beginning. It has frustrated users to the point where alignment posts are commonplace, crazy workarounds have been invented, and some users have even decided to switch to another laser brand because of it (theyre in for a shock once they get their new lasers! more on that in the next section).

While it is frustrating, I do think glowforge will ultimately keep their promise! Snapmarks has been released in beta, and has shown to be incredibly accurate, giving users the ability to accomplish what has, up until this point, not been possible, and there is more to come from that front im sure.

Surprising Comparisons

For many laser users, the glowforge is the first laser experience they’ve had. When things haven’t gone as expected, the general urge is to try something else. As I am constantly learning about the tools I use and what is available, I have tried many lasers, and have seen many others go down the same path. There are many lasers out there that offer superior speed, more power, and a handful of other features that the glowforge does not have. Surprisingly, when it comes to alignment, they ALL have problems… even the big 3 (Trotec, Epilog, Universal).

Having used snapmarks for a while now, Ive grown quite reliant on it, and use it often, especially for accurately placing engraves on pre-made objects. It has been great! Seeing that the big 3 also offer camera/fiducial based alignment systems (usually costing $5000+), one would assume that theirs must be bigger and better than a laser that costs half as much as one of those cameras. … and heres the kicker: that is a false assumption.

The ability for any of those laser systems to properly align an image to be engraved, to fiducial markers, does not exist. Beyond that, at this point in time, none of those systems even allow you to manipulate the rotation, width, or height of images to be engraved in their UI. The only thing the incredibly expensive camera systems do at this point, is to allow a vector cut to be performed with orientation corrected via the fiducials. To quote one of the technicians: ‘the algorithms are incredibly complex for the cut, and unimaginable/impossible for raster engraving images’. We all know this isnt the case, because the glowforge can do it, but everything is relative.

I think its fairly easy to take for granted the seemingly basic functionality the glowforge affords us. I know I have to a degree, and Ive been pretty shocked to see the state of other much more expensive laser systems out there recently. To think that this small company who seems behind in their promises, is actually so far ahead of the competition in so many areas is crazy right? Especially when many of them have been in the game for 20 years plus. I can tell you one thing, theyre all scrambling to catch up.

The Current Reality on Alignment

As of now, if you have snapmarks enabled, you have one of the most, if not THE most, accurate alignment systems available in laser engraving systems today. As crazy as it seems, it’s the truth. The only company that could offer something similar was Universal, and only on their XLS line ($100k+), and it came with a massive number of caveats related to distortion and skewed images. No thank you.

Where Things Are Going, and Could Go

I think the glowforge has disrupted the laser industry quite a bit with their innovations. Yes, they still have some of those launch campaign checkboxes to tick, but know that no one else in the laser industry is anywhere close to solving the things glowforge is working on. I like what Ive seen so far, and Im looking forward to seeing where they take things in the future. The things they are working on currently are showing a lot of promise, and the flexibility they’ve designed into this first gen machine can allow for even more options should they choose to do so.

Snapmarks opens up the ability for lid-camera auto calibration in order to fix optical placement accuracy of the lid cam. This could easily be accomplished simultaneously with the launch of their new invisible QR code based materials, where the entirety of the masking sheet is covered with UV QR codes. By using the lid camera to see each QR code, and the head camera double checking the placement of those, de-distortion maps can be generated to perfectly render the bed image exactly where it should be, meaning no more misplacement.

The swappable heads can allow for new features to be added, such as a head with a red dot to be built in (or they could use the current depth detection red dot) to allow for tracing of cut/engrave areas on the materials for double checking that your material is really where it needs to be. (before they fix the optical placement via snapmarks that is).

Real time depth sensing can allow for them to detect edges of materials, along with their height, allowing them to build a 2.5d model of your material, accurately positioned on the bed, providing the ability to perfectly engrave a flip sided image on non-symmetrical objects.

Newer heads can be released to enhance these capabilities and keep your machine up-to-date and viable much longer than any competitor machines.

If they keep it up, I think the future is bright for glowforge.


Thank you for this thoughtful essay. My only hands on laser experience is the Glowforge. I have, however, watched my projects being produced on an Epilog Helix and the alignment issue is an ongoing challenge even with this excellent industrial machine. I have never regretted waiting for the Glowforge and have enjoyed learning how to best utilize its capabilities. I am lucky in that I have Snapmarks and have found them to be incredibly useful. I am excited to share in the continued improvements being made by the Glowforge team, and if there are new print heads being developed I am sure I will find a way to use them.

I appreciate the hard work and patience of the many people on this forum, like you, who have so generously shared your expertise so that novices such as me can benefit from your experience.


Completely agree. I’ve also used a Full Spectrum and I do wish we had the laser dot as well, since it’s handy for eyeballing, especially on the edges (and since it’s basically already available). That said, if I had to pick, I prefer the camera. The idea of both is very appealing!

I don’t have Snapmarks yet, but from what I’ve seen, they are definitely a game-changer.


Excellent write up. I also think the future for us is bright with Glowforge. I don’t have snap marks as of yet but have not had any alignment issues that I could not use a simple work around to overcome. There is always more than one way to accomplish something The search function on this forum is one of the best tools we have at our disposal.


Up until just now I had thought the decision to make the head be attached by magnets was foolish, but I never thought about the ability that affords us to easily change/upgrade the head in the future.


From way back in the day…

The Expansion Port

In one of our pre-release unboxing videos, a customer remarked how strange it was that the laser head isn’t permanently attached. Not only is it removable, it just snaps into place with magnets. That’s not an accident.

The head isn’t a permanent part of your Glowforge; it’s a Glowforge accessory. In fact, it’s just the first of many accessories we have planned. From the start, we designed the laser arm of the Glowforge to be flexible. It has connectors to supply both 30 watts of power and high speed serial data. This means in the future, you may be able to use the expansion port to exchange the laser head for anything from an inkjet printer cartridge to a 3D extruder, creating multi-pass multi-media projects seamlessly.


i’ve used only two, the GF and the Universal PLS. i would love to merge the two systems together for one even better interface.

i will agree that snapmarks really tilt the alignment side for certain things to the GF. especially preprinted materials. if you preprint digitally, it’s really hard for your printer to properly register the print in a consistent, exact location on the page. could be slightly crooked or a little left or right of expectations. on the Universal, i’m lining up the corner of the sheet to the 0:0 corner and using a ruler to measure to crop marks and lining that up in the UCP based on my ruler measurements. if the print isn’t perfectly square, it can be quite a pain to be exacting. with snapmarks, i just have to wait for the alignment (which, to be fair, isn’t as snappy as the name). but when it hits, i’m within a very small tolerance, both location and rotation. with the universal, i have extra prints and am prepared to have throwaways to go along with the inconsistencies of the prints.

now, once i get the jig set up on the universal, it’s quicker because i don’t have to resnap every sheet. but i have variations per sheet at times, so more waste. if i’m doing a lot, i might prefer the universal as the setup time gets spread out. if i’m doing a few, the GF is much faster and more accurate.


If we could get the snapmark “lookin’ good” time down to about 30-45 seconds. :heart_eyes: either way, 2 minutes (which is what it takes on my system) is worth it for knowing that my print is going to be dang near perfect.

Printing is a significant cost for me and when I screw up (pre-Snapmark), it really sucks.


i totally agree that it’s worth it compared to screwing up.

thing is, a jig on the universal actually makes more sense if you have to keep repeating the same thing. snapmarks take too long to make sense to do 30 die cut cover sheets. otherwise you’re looking at taking at least an extra hour in “snapmarking” vs, say, 15 mins of adjusting to get the first one right and then placing the rest in the same spot. knock it to 30 seconds and we’re talking about similar timeframes.

1 Like

I agree. Fixed jigs are perfect for reproduction. The Snapmarks work for me because each file is a little different, more than likely, and I don’t have to worry about a perfect trim. I just trim the 24” paper down to just outside the snapmarks and glue it up on the board, rather than having to trim the print perfectly on each side and then glue it up perfectly.

So we need a way for Snapmarks to hold alignment until disabled. That way, you could make a jig, snap it at the beginning of a run, and not have to do it again until you reinsert the jig.


the issue with jigs is that i need to create registration crop marks to try to rim outside the real crop marks accurately and then line it up to the jig. it’s an extra cutting step for me (to adjust for rotational inaccuracy of the printer paper path). there are some things it’s not a big deal for. i fudged it with some of my puzzles. but other things, like proposal covers with die cutting, i need the rotation and location to be just right. snapmarks will do that, but slower to repeat.

so do i want to cut everything an extra time for registration marks or do i want to wait 20+ times for snapmarks? sigh

1 Like

What would be ideal - if I were running the same puzzle file over and over - is Snapmarks in conjunction with a jig/repeat mode.

I think it would go like this:
Cut out all of my backer boards, with snapmarks scored on one of the boards

Trim and glue up (yes, have to be exact, but it’s not too hard)

Put the Snapmark jig in place, lock with hold downs, and Snap

Fit in 1 puzzle to the cutout

Print - and job is saved

Insert next puzzle, and hit Repeat Print


Great post, I’m going to move it to Tips and Tricks so it doesn’t get lost in the noise in Everything Else!

I never thought about a laser … Saw the Glowforge campaign … Had to have it!

Really appreciate your comments concerning not only the Glowforge, but also other lasers. Thinking I have the perfect machine for me!


This is a fantastic write up with some very positive vibes. There still so much technically that I don’t know about a laser and probably never will, so it’s marvelous to be able to rely on others who DO know about that stuff. Thank you so much for taking the time to put that all together to share with us.


Ach I totally agree. Heck, on something which really matters e.g. placement of an image on a chopping board etc. I spend WAY more than 2 minutes manually checking and rechecking alignment to make sure it’ll be right. Even then, it’s a wing and a small prayer. If I could use Snapmarks and just hit the button and be confident, yeh I’d be very happy with that!

Thanks for the thoughtful write up on the state of laser cutter alignment.
I, as well as many others, have groused about the lack of the promised pass
through features. Nevertheless, I have been able to make things that would have been
impossible with other tools and do not regret the purchase. I would like to have snapmarks enabled on my machine as that would increase my productivity and allow more things to be made. I guess many of us are waiting for snapmarks.


That is definitely part of the issue. As so many rave about Snapmarks, those of us without can’t help but feel a bit disgruntled. I support GF as a company, I understand the beta process, and so on, but none of that changes the fact that right now, my laser doesn’t have the best alignment in the industry. There are workarounds and it’s not making me lose sleep at night, but I’m also not quite ready to pop the champagne cork yet, either.


I really appreciate the bigger picture look from someone who has used multiple laser types. Since it’s unlikely I’ll ever purchase one of those more expensive lasers… I have no perspective other than measuring the GF off their own benchmarks. The few times alignment has really been an issue for me, it’s needed a good cleaning.

1 Like