How little can features be for the trace function to work?

Hi all,

I’ve just watched the Q&A of yesterday, and loved that @dan that he could create all kinds of things with the Glowforge while hating Photoshop. I’ve kind of tricked my non techie sister into buying a Glowforge for her cardmaking shop. She want to provide a service to her customers by preparing cards for embroidery by pre cutting the holes.

At this moment she has quite a poor workflow by scanning the booklets, mark each spot in a seperate layer and than engrave it as a raster using a Chinese laser cutter.

Would the trace modus see little dots drawn on paper, en be able to cut them as “a vector”? This would speed up her preparing flow and the cut speed significantly making it much more convenient and cost effective. Main concern I have is that the dots have to be that small (like a whole made with a needle) resulting in the trace feature not properly identifying them as a “feature”.

Looking forward to hear from you!


Actually you can probably do something a little easier with the Glowforge. Set up a vector for the holes you want, then resize them to fit exactly where you want to go on the cards using the built-in camera. Dan recently said in another thread (I don’t remember which one, or I’d link you) that you’ll be able to resize and move things around in the software, so this sounds like a prime example for that functionality.

1 Like

i edited the title of your post just a little for clarity.
I am curious about this as well.

A lot of what we have seen of the trace feature was done using sharpies. How fine a line will we be able to use? Ballpoint pen? 0.03 pigment liner?

The art for the catan pieces looks intricate, and we know that was done “traditionally”. Was it sized up or down before engraving? I think I heard @Dan mention one or the other, but I can’t remember where that was.

The point is I’m hoping my sister won’t have to vectorize anything (I could do it myself as mechanical engineer and hobby-designer, but as said my sister isn’t and she does not like to be dependent on me for projects). In my ideal scenario the workflow for my sister would be: take booklet, take a fine liner, trace the dots on a sheet of paper(the embroidery lines will mess up direct scanning), put the paper in Glowforge, open browser, click “cut outlines”, TADAA!

As I heard Dan talk about the magic, I regained hope this workflow could actually become true.

That’s exactly what I was talking about. Call it laserize or whatever if you must, but something is still vectorizing the drawing, then you could place a couple of sheets of stock in the laser, hit copy and paste and boom.

1 Like

Okay, then I misunderstood what you were saying. Likely due to the resizing which is not the purpose after she traced it by hand (I thought you meant she could make the vector larger and therefore easier and later rescale it for the Glowforge).

But glad to hear it will likely become possible.

Its supposed to cut in 1000dpi, lets hope the camera can meet that depth. This is actually a pretty good question… If I remember correctly dan posted something about the flying head camera used for high def close up scans to get more accurate photo info, im assuming for this exact purpose. As the camera on the lid is a wide angle lens it will have a higher pixel density in the middle of the bed than on the edges, so they would have to account for that, which is what the laser head cam is for.

I really hope they can meet the 1000dpi for the trace feature. It would make sense to match the capabilities of both abilities, but I guess for many applications this level of detail is not necessary.

1 Like

my quick and dirty test just showed me that a relatively new sharpie marker takes 625 dots to black out a 1-inch square (25 dots wide, 25 dots tall).

1 Like

Genius way of checking, however as far as I’m aware dpi is a single dimension measure(not square), so we know the trace feature will be at least 25dpi, but I am hoping it will be a bit higher ;), I guess 300 would be sufficient for most uses?

I think this is right – it’s how many dots (or pixels for ppi) fit along a one inch… line.

um… derp. I knew this. :flushed: thanks for correcting me!
Right, so dots per inch, not dots per square inch.

I tried again with a 0.03 pigment liner, and got 80 dots.

1 Like

Let’s just wait and see if someone from the staff or maybe our beloved beta tester can give some backed insights in actual obtainable resolution.

I can imagine there can be some kind of torture tests like the ones you have with 3D printers, where you print a specific pattern with your inkjet or laser printer with lines and dots of different sizes (which will show what minimal features the gf can trace and cut). This you could use to check whether your Glowforge is still calibrated correctly and optics are clean(which I know shouldn’t be a problem). This calibration on specific material could maybe even be used as a benchmark to give some kind of indication on the estimated condition of the laser tube.

This gets a bit of topic but came to my mind when I was talking about the torture test.

Is there already some kind of request for automated value determination for “non-Glowforge” materials? As the laser used to measure depth of the autofocus(said on yesterday’s Q&A) is likely sensitive to other wavelengths than the laser for cutting, a sheet of a specific material not hurt by the cutting laser (steel could be an example?) influencing the autofocus laser in a distinctive way, could be placed under the non-gf material. The autofocus laser should then be able to tell by running a specific shape(why not the Glowforge logo ;)) with ever increasing power, which speed/power ratio is needed to cut through the material. To make this feature even more awesome, it would be extremely cool if, when wanted, the gf would engrave a marking with the needed speed/power settings for the gf, so the next time you insert the same material in the gf it recognizes it and knows which values to use.

I guess this should be technically possible, and could become a one time purchase because we only need one sheet of this reference material. I guess this would really empower makers who might want to experiment with all kinds of materials.

So something from the Q&A yesterday that I’ve been wondering about and maybe you know better. If I trace on a piece, then have the GF raster it, can it save the image and reuse it on another piece. Without having to hand draw it again?

1 Like

I think Dan indicated that that should be the case, yes.

1 Like

yeah, that is how I understand it. I’m not sure if you would then be able to download that image to play with in other software. Like if you doodled something on some scrap for fun, and then after a week or a year you realized that you absolutely loved it and wanted to incorporate it into other art in photoshop, or print it on your 3d printer…

or… imagine if someone made something silly for their child, and shortly after passed away… and that something silly suddenly became a treasured artifact. Being able to make a replacement one would be pretty much incredible.

Not trying to be morbid, but was just reading an article about the digital legacy that my generation gets to enjoy and contribute to that earlier generations never even dreamed of, or at best might have had a fraction of in some physical form. Thousands of photos, Tweets, facebook likes, family interactions… Glowforge projects.


lol I just remembered, isnt that functionality demoed in the original video with plane invitations or something?


It would be very cool to be able to use the macro camera on the head to take close-ups of small features. We can rule out the lid camera from being able to get down to 1000DPI though, as that would require a 240-megapixel sensor (12,000 x 20,000 pixels).
edit: unless the camera could zoom

How high of a resolution can we assume the lid camera will be? I imagine 20-megapixels is realistic-ish. With a 3:5 aspect ratio that would be ~3500 x 5800 pixels (~289 dip).

Perhaps the density issue could be compensated for with aspheric lenses.

You definitely can scan, resize, and save. The lid camera is much lower resolution than the laser is capable of; it depends on the location on the bed, but I’d estimate 100 DPI. To work around that, draw larger & resize (that’s how Catan was done) or scan with a separate scanner.