Print and Cut testing!

It’s a bit tedious to set up, but Print and Cut is quite effective!
This process does NOT use an alignment camera, nor does it require a “formal” fiducial- you create a circle (only for CAD purposes, Tool layer makes sense) and position is center on the first reference point on the job- something you can actually see on your printed media. It may be a crosshair you drew outside the printed image, the corner of the printed image, or just any distinct corner near the edge of the image.
At the same time you position the laser head’s red dot over that point on the printed stock. Click “Set First Target Position” and it locks coordinates of BOTH the selected point on LB’s design workspace and the head position on the laser. The workspace target and laser head target cannot be entered at different times, it must lock both together with the same click.

Then pick a second point on the design workspace and on the printed stock. Of course this works more accurately the further apart these 2 points are.

It will rotate and scale your design to align to those points. The system switches to Absolute Coordinates, as the User Origin would only conflict with the selection already made.

In this case, the graphics are from a laser printed paper sheet glued onto plywood. Within LB I imported the same image and used Trace Image to create cutout vectors. At this point there is no known relationship in scale between the printed material and the LB object, but that’s ok it will auto-scale it

Alternately, you could use Gimp, Photoshop, etc to do a more sophisticated Magic Wand etc method to do your selection. However, LB ultimately needs vectors, and at that point I don’t see how to save the selection outline as vectors. The photo editing tools can instead just make everything a transparent outside the object you want to cut out and that will Trace inside LB easily and with precision.

This can be used for:

  1. Cutting out printed media as shown here. SUPER useful for making game pieces. This was printed paper glued to plywood. The image could also be ink transferred or direct printed. Or you have something already printed, like a poster, and want to cut things out of it. You’d need to scan or photograph it as best as possible and work on the scanned image if you don’t have the source bitmap.
  2. Aligning a physical jig placed on the bed. e.g. a jig with 8 form-fitting cavities for 8 blank combs of the same size and shape for engraving. The jig will just need to have 2 alignment marks burned on it when it’s made. The jig is mag-clamped onto the bed so it won’t move, but exactly where it is and its angle is not important as P&C will align with the alignment marks when you do the “set targets” step. That only needs to be done once when the jig is mounted, then you can not only swap combs but edit the design in LB and the new content will also align with the target points. In this case you might choose to “Align Output to Target (no scaling)” to keep the scale exactly the same from run to run.
  3. Aligning a single item to engrave/raster/cut on, like a phone or laptop.
  4. Recovery of a split or broken job. Most likely case would be starting the laser job and unfortunately being unable to finish the job at that time for some reason. Traditionally we just resign to throwing away the stock and start over. But with P&C it’s possible to bring the partially done work back later, clamp it down, do the P&C alignment, and continue running the job where it was left off. It’s not accurate enough to continue a raster or recut additional passes over the same line that was cut the first time.

There is no need to do this per final item, but it does seem you’d need to do it per print. That is, if you slap down a 4ft x 2ft piece of printed plywood, you could cut 300 game tokens out of that. But are you going to have a single printout that big? I used a desktop laser printer so a print is limited to 8.5"x11", minus the margins. Gluing multiple sheets onto the 4’x2’ plywood would defeat the alignment concept here, as they are not accurate in their spatial relationship to one another and cannot be cut accurately with precision alignment marks on different pages. You could glue 10 color printouts, each with their own alignment marks, to the sheet but you’d need to realign P&C for each printout. Thus there is increasing value for larger format printouts that you can glue or transfer to laserable material as one piece.

After 3 successful jobs, I tried a third and ran into a mysterious offset that is probably a bug. Reran the P&C alignment and got the same result. Not sure, still investigating. It is likely specific to some weird part of the design or the way I used P&C.


Just to show how easy and effective this process is, here’s a visual walk through.

  1. Bevo was printed on paper and I dug up a scrap piece of 3mm ply.

  2. Bevo was glued on the ply with 3M 45 adhesive spray.

  3. Placed my stock deliberately askew on the bed and in LightBurn walked through the Print and Cut Wizard. (This video explains the Print and Cut process in the context of recovering an aborted cut.)

  4. Scored the image to show how amazingly precise the vectors and the image sync up.

  5. Cut results are just what you’d hope for … perfect.

This does significantly raise the value of a large-format inkjet printer, because this process would have to be repeated for each 8.5"x11" printer sheet, and tiling multiple sheets with the margins trimmed and manually aligned would largely defeat the point of auto alignment.

A large printer could be used to make larger things, or cut many smaller printed parts on one sheet, all done with one alignment task. The value increases up to 1.6m width of Tarkin.

We have a large printer in coworking, needs work AFAIK, but it’s actually not all that much bigger than a regular printer.

One interesting thing- someone did hack a large inkjet printer to print directly on wood:

That is NOT replacing the controller or hacking the firmware, something probably no one will ever do, the details of the pinout and electrical drive currents alone to the printhead would be enormously difficult to nail down, but that’s just a fraction of the task, replicating all the priming, cleaning, calibration, etc seem insurmountable. But they didn’t do any of that- they just replaced the stepper-driven roller for the paper with a stepper-driven linear feed table. All they had to do is adjust the steps to be the correct linear feed per step, and got the print surface at the correct distance from the inkjet head. The firmware had no idea it was feeding anything different.

There is a print shop very close to Asmbly that does direct print to wood, using a large format uv printer. It’s how I get my cornhole boards.