QR Code issues

[Brian Lamb]

I had a customer on one of my medical alert bracelets that wanted a lot of info on the bracelet, couldn't fit anywhere near all the text so I offered a QR code. I've done a few before but not this much info and they seemed to work. Any suggestions? Scanning the code enlarged off the computer screen works fine, after engraving the phone finds "nothing".

393755

[Kev Williams]

My guess is too much QR code to fit too, I'm assuming Cermark? In looking at the pic it appears the digital 'pong' dots are overlapping one another. A fiber could probably pull that off that small, but a C02 & Cermark is asking a lot :)

[Brian Lamb]

Evidently, yes, Cermark and a CO2. I had done one before, and it worked, but had used a free QR code generator instead of using the built-in QR generator in Corel. Do all QR code generators create the same code? No variations? I seem to remember it being bigger blocks and spaces, this one does look like a blob all run together....

[Kev Williams]

The more info, the smaller the blocks-- it all takes up space :)

[Mike Null]

I think it might be contrast. Scanners need sufficient contrast to read the code. Not all substrates will work.

[John Lifer]

Not the contrast. I've done some on pretty junky surfaces, mainly with my fiber, but some with cermark. It is way too much information and as you see, the dots are not distinct. I'd either trash that one, or sand it off and cut information down by half. and try again. Engrave on a scrap first. I put my company name, web address, actual address and phone on mine, and it is pretty complex. Really made for maybe 36-48 characters. Not paragraphs.

[Kev Williams]

While drinking my late morning coffee I thought I'd mess around a bit--

I took a 1" x 3" SS ID plate and see how small I could make readable with what equipment I have. Also, because part of the issue here is 'too many words', I used a font called Steelfish, it's extremely compressed yet very readable. Whenever someone wants too many words, it's my go-to font :)

top left is what I was able to pull off with my 30w GCC and Cermark. My LS900's CPU is in the shop, and would've done a noticeably better job because it produced a much smaller beam spot than the GCC. The letter heights range from .090" on top to .050" on the 5th row. Below the 5th Cermark row I did rows 3,4 and 5 with the fiber. Big difference! However, the 3 rows done by the fiber revealed a 'wobble' issue I need to look at. So I changed up the hatch routine a bit and ran the bottom 2 lines again, plus a 3 row which is the same as the 2nd row, only I thinned out the lettering by .001" using the contour tool. The results:

the whole plate for visual ref-
393789

Note the Cermark isn't too bad, even the tiny line is mostly readable, and the top 2 lines are very readable, and with this font you can cram a lot of words into a pretty small space! Now on the 2nd fiber run on the right, thinning up the lettering on the 3rd line increased the readablilty quite a bit- The easy way to do this in bulk is to go online and find a similar font that's thinner ;)
393791

closeup
393790
-the bottom line contains 35 characters in a .048" tall x .635" long space...

My issue, the fiber text from the left section:
393792
--I have a seriously crazy 'lazy' in line 2 going on, and a few smaller bobbles in places. Changing the hatch routine helped a ton, but the G in DOG in the 1st closeup pic still has a bit of wobble to it. Time to start pulling cables and hitting them with the electronics spray cleaner...

Anyway, fiber is the way to go for really small stuff. Or a 355nm UV laser for REEEEEAAALLLLY small stuff! Kinda thinkin' they're a bit spendy though...

[Brian Lamb]

OK, I did some more testing, at least on the computer. My QR code is set at .600" square, bracelet is only .750" wide and curved on the edges. The scan won't read at 100% or 150%, but when I change it to 200% it will read, which means the QR code would have to be 1.2" square.

Thats at the best contrast possible, white screen black marks. As Mike said, stainless reflection might cause even more issues. Not sure it's worth wasting any Cermark or stainless to test further.

Kevin, the point of the bracelets is in most cases a medical alert, so making the text as small as possible isn't really a good idea in an emergency situation. I stick with Arial 9pt as the smallest I will go, otherwise folks with even slightly bad eyesight won't be able to read it without a magnifying glass.

[Kev Williams]

I realize people will need to read the thing, I was just showing how much smaller a fiber is capable of.

The largest lettering on my test plate is 9pt, and my 64 year old eyes can read it easily.

As to the Steelfish font vs. Arial; where you can fit 100 Arial characters horizontally, you can fit approx. 220 Steelfish characters. And FWIW I increased the character spacing by 12% to help prevent the letters from running together.

--both lines are the same height--

393834

So while you may not get QR code to work with Cermark, a compressed font might get you there...

[Bruce Clumpner]

Late to the thread, but it looks like it's over-exposed or ran too hot?... however that works with Cermark? The loss of detail could be causing your issue. (note the loss of detail in the caduceus)


I had a customer on one of my medical alert bracelets that wanted a lot of info on the bracelet, couldn't fit anywhere near all the text so I offered a QR code. I've done a few before but not this much info and they seemed to work. Any suggestions? Scanning the code enlarged off the computer screen works fine, after engraving the phone finds "nothing".

393755

[John Lifer]

The issue is more the phone and it's camera than the engraving. There has to be enough 'white' are for the camera to be able to differentiate between the black and background. Less information gives you more white space
And by saying too much it is too much for a SMALL engraving.
I didn't make myself clear and Brian proves this by increasing it's size. You CAN make a small QR code, I have mine running on a 10mm square on anneal on stainless cups. (.39 inches) Reads with my three phones easy.
I did one as test using a 'frost' setting on plain aluminum at 16mm square and it is enough contrast to read also. Better than white on a shiny anodize part.

[Kev Williams]

yeah...it's the engraving ;) -- has to do with 'tool width', in this case, the laser beam...

I created this QR code, it says something like 'this is a qr code test'- put it into corel, blew it up big, and did a Trace. Even being just a photo, the edges and corners are extremely sharp and straight. I resized it to .600 high per Brian's size. At that size I had Corel measure a single dot, and even tho these single dots are probably larger than Brians, they measure only .026" square, pretty easy enough for a fiber, but with a C02 beam spot being around. .006", that's adding 25% more to the perimeter of everything!

This is the QR verbatim at .600"--
393897

This is the results after adding .003" to the perimeter of everything and radiused the corners .003" too, to replicate the results using a .006" wide laser beam...
393898

there's a couple of holes I'm missing, but you get the idea, with all the overlaps the result will be gibberish to a QR reader!

Now, if you were to take the time to inward-offset everything .003" before lasering, you'd get pretty close to the original when done... very time consuming, but possible... :)

[Darren Wilson]

Hi all,

My understanding of QR codes is they require a quite area/zone, to help them scanner read them propery. Reflective surfaces do not provide the quiet area required. I have found when marking reflective surfaces I have to add the quiet area to get reliable scans. Add that to that what Kev showed above and this is probably why they are not scanning. Marks below were made with fiber...
394085

[Kev Williams]

I love the oxymoronic irony of how your quiet space looks exactly like white noise :D --the fiber did a great job!