Tested Gets a Glowforge - Second Pre-Release Unit in the Wild


Hrmm dat offset on the trace & engrave.

I wonder how many other units are out there… Glad to see them out there.
Jelly that I didn’t get picked to get one. Oh well.


Awesome, although it looks like the trace functionality on the above video is off as it gets away from center.


Not to bolster false hope, but I really think that they are just now rolling them out. @marmak3261 was just notified this week and then had one in hand within a day. Given your other tools and experience, here’s to hoping!

Yeah, I noticed that, too. There was mention near the end that Glowforge told them the trace capability will improve when the start using the second camera on the print head to do the tracing.

1 Like

For what it’s worth, it looks like a scaling problem rather than an offset problem. On the lines in the center, increasingly farther off toward the edges. Still not a good thing, but a different not-good thing.


Anyone notice the offset of the outline cut on their sample ‘Savage Cave’? Is that for kerf adjustment? Or off screen manipulation that we might not have seen?

EDIT: Submitted this too late and missed the above comments. I’d agree that it’s probably a scaling problem from only using one camera.

What’s happening with the engraving about 59 seconds in, the engraving looks like it’s cut 20 lines but the head only moves 2 lines, is that just a trick of video editing?

its just an image processing error. I saw the same thing at maker faire.

They arent able to determine depth with the lid cam, so they dont in reality know how close the drawing is to the camera, so until they get the depth sensor working these things wont engrave the right size.

1 Like

its a time lapse, just so happens to be at the right frame rate to make it look like the head is moving smoothly. that engrave probably took 15-20 mins


I continue to be blown away by the quality of the proofgrade materials I’ve seen, in person and on here.


Looks like parallax error. I guess they have to make an assumption about the thickness of the object in the bed (although you’d think they could get the information from the proofgrade bar code.) Should be fixed once they start using the camera and rangefinder built into the print head.

1 Like

Well I got to hand it to Tested for a good and honest review of the pre-release unit.


So excited that Adam has been consulting with the glowforge team. I have a hugeee respect for that man. :blush:


If it is parallax error, they could also (potentially) fix by printing a registration pattern on the masking material. But yeah, probably better ways.

1 Like

they do have one! the proofgrade QR code. if they are all the same size they could definitely use that as a scaling device.


He’s got a lot of fans at Glowforge HQ. :slight_smile:

We do have the height sensor working; the errors are because we haven’t finished the software model. That model has to accommodate:

  • Tiny manufacturing variances that affect the placement of the camera by a few thousandths of an inch or a degree
  • A lid which may have a tiny particle stuck in it, leaving it at a minutely different angle than before
  • Materials whose height varies across the surface
    Right now the code just assumes everything is the same, and that introduces errors that magnify across the material. We’ll have improvements for that coming.

Good suggestion! Those are temporary, though, until we get our UV ink printing system working.


The only problem with that is it forces the customer to only use the Proofgrade materials. Will probably use some Proofgrade but in the end that would be a deal breaker for most people.


Relatedly, while I understand this would be low priority, I’d love to be able to print my own QR codes for personal use with locally milled materials.

I’m fine with just keeping a log, too, so this is really not that important, but maybe a few weeks / months post-launch.