Introducing Snapmark (September 2018)

This sounds AWESOME!!

1 Like

:smile: And it’s so fast you’re going to squeal. (All the Print & Cut files I’ve got work great with minimal adjustments…and those are mainly trying to get the marks to fit on the page along with the design.)

4 Likes

I could tell by your write-up!!! So exciting!!

1 Like

@dan That seems to be a great feature! Would love to try it out and give some Feedback! Hope my unit is on the good side of your Data-Noise-Map :wink:

1 Like

I’ve used Snapmarks to cut hundreds of tags for cookies and the alignment was spot on!

I also use it for personalizing items that I make batch and then can add a boat number and such, if desired.

I know have a stack of cardboard jigs for various items.

Love the feature!

17 Likes

Looks like a great feature. Given the current inaccuracy of visual alignment, it would be very disappointing if this was not made available to all uses (including Basic) free of charge.
:grinning::thumbsup:

15 Likes

Maybe a new crumb tray with a couple of these… what were they called again?.. “fiducials” printed on it could be made up? Then would it be possible to affix a permanent jig consisting of two perpendicular rails to this new, experimental crumb tray so those rare rectangular materials could be cut with sub-mm accuracy?

Whoa, it just hit me! OMG, the Glowforge could detect this tray automagically, maybe even each time the Glowforge is powered up or somethin’, and then this special crumb tray would be ready to use by default!

11 Likes

HEY! #RabbitHoleWarning PLEASE! Jeez! We have jobs that require not futzing our productivity!

8 Likes

Ok. So I’m going to be the one to ask the stupid question. You have a camera in the lid. You print the fiducial on paper. Input thickness. The lid camera system takes a picture of the paper identify the fiducial, then to align cuts to the fiducials on page. Multiple people say this works perfectly.

For those that don’t have this option yet, how is this different than the system that doesn’t work now? I can set the thickness, allow the camera to take a picture, then manually move my file to spot the exact spot I want to cut (zooming in as far as needed), and be 1/4" plus off especially in the corners.

Is the laser head now moving to the approximate position of the fiducial and making several small corrections (similar to homing the head with the GF symbol) then taking a better picture directly over the top of the fiducial to position correctly?

3 Likes

After the lid fisheye does its limited thing, the head comes out with its camera and does its thing. If all goes well, the pre-scored or printed image aligns with the file you have loaded in the GFUI and merges everything. When it works, ie. the files are right and you set the printed or scored material reasonably close to what you see in the GFUI, it works like a dream. I’ve never been able to measure any offset from what I had programmed.

7 Likes

it uses the head camera for more accurate placement since it only needs to look for the one image. The same way 3 registration circles work on other systems from trotec and epilog. They just use 3 to create a triangle.

3 Likes

Well… why is the description of what the Glowforge doing actually doing so poor? This is the first time someone actually said that the CAMERA in the head was being used to increase accuracy. Seems using a whole new feature of the machine would be worthy of a mention.

That said, why can’t they enable this new “feature” of using the head camera to take a picture without the fiducial? It goes like this:

  1. You enter material thickness
  2. Lid takes picture of bed
  3. Allow the user to drop a set of electronic “fiducial pins” on the screen near where you care about alignment of your drawing.
  4. The head goes out and takes scans of those “fiducial pin” spots, increasing the fidelity of these locations, creating an accurate high quality picture “square” of these spots, and then places the replaced “accurate” image square on the screen in these “fiducial pin” locations.
  5. You then move your electronic file to align visually with these more accurately scanned locations on your screen.
  6. Print

No special extra work required printing paper markers, kept vertical, special svg files with embedded fiducials?

You would just use visual cues (magnets, hand made mark, edge of material, etc) to position as you do now.

For those that want the extra accuracy of the printed fiducials and special files, I have no problem with them using this new feature. In fact, I love that they have something that is more accurate.

Feature request: please add using the camera IN THE HEAD to obtain a more accurate cut than what I have now…

7 Likes

Your process seems nice, but I wonder what happens when your material is too uniform or when there’s some kind of repeated pattern in the area of interest.

Sounds like you might be able to do most of what you want by printing out a bunch of snapmark pairs and taping them to your material with appropriate care.

(One other thing I’ve noticed, or thought I’ve noticed, is that once you’ve done the snap thing, you can still move stuff around in the GFUI. So, for example, when I was engraving pencils, I zoomed in and moved the text to align with the jig-cut lines after uploading the file.)

If the material is uniform or repeating… that’s the beauty of it… you don’t need anything special, just a dot or mark of your choosing on a piece of tape, the magnets you placed on top of the part, or whatever. You are just getting an accurate scan that you can accurately position in areas where you care about alignment.

I can’t tell you how many times now I have lasered over my magnets on the project I’m working on now. Hundreds of little foam parts and trying to get cuts in the right spot is a pain. I would mark all those locations to make sure I didn’t run them over with the cut file!

1 Like

Ah, now I understand. From your first description, I thought you were saying the camera would just take a picture of the material as it was, without any added markings.

I don’t think anybody outside of Glowforge can explain why they don’t do what they don’t do. I can think of lots of features that seem possible with the hardware we have today and that I’m surprised we haven’t seen. I also know how hard it can be to prioritize a product backlog.

In an effort to resist the temptation to respond to a positive announcement with a complaint, I prefer to think of Snapmark as a sign of good things to come. We’re getting a thing we didn’t have before. This is true regardless of whether or not other problems remain.

28 Likes

Amen

giphy

15 Likes

All I know is that now I’m going to be obsessively checking the GFUI for a magnet symbol.

10 Likes

Can’t say for sure, but if we use their fiducials it is a known thing to them. The size & format are defined. That simplifies the recognition and then the corresponding corrections they need to make to position everything. The software can look for something that seems to be the fiducial and then correct its view of it based on what the fidiucial should look like precisely, check again to see if it does and then apply that correction to everything it “sees”. If the correction doesn’t result in a precise fiducial view then it’s either not the fiducial expected or there’s a different problem so it can’t apply a bed sized correction.

Using something other than their fiducials means you’d have to have some way of telling the software what yours was and describe it mathematically exactly so the software could find it and calculate the correction needed. In a drawn project (vs photo, etc) it would be straightforward to create a self-defined fiducial mark and then indicate to the software via some kind of point & click in the GFUI where it was so it could send the camera there and calculate what it sees vs what the SVG description is and make the correction calculations. But that would require a whole bunch of other software processing that isn’t required if you use one that’s predefined by GF. So the Snapmark implementation is likely the result

6 Likes