Yeah, the horse isn’t getting any less dead, but I had an idea I wanted to share anyway.
We’ve all seen each other’s test targets showing various degrees of drift, which varies across the work area. It would be nice if, for a given material thickness, we could run through a manual calibration process, aligning expected and actual results for multiple operations.
It could work a lot of ways, but imagine something like this…
- Put in a piece of Proofgrade, or your own material with a height measurement.
- Start calibration mode.
- The Glowforge does a low-power cut, marking a crosshair in the upper left corner of your material.
- The app shows you a zoomed in image of that area with draggable artwork in the cut position–just like we can see right now.
- The user drags/rotates/nudges the artwork onto the actual cut mark and clicks NEXT PLEASE!
- The process is repeated at several other places on the material–as many as it takes to get the data needed.
- (You could cover the item with copier paper so you didn’t consume it in calibration.)
- Math happens, and a custom de-warp map for this material thickness is recorded in The Cloud.
- Future camera placement for this material thickness then automagically works better.
- (Seems like having one really good de-warp map may be able to improve results for other material thicknesses too, but I dunno.)
I realize this is a very un-Apple-like process, and also possibly very un Glowforge-like. But this is a professional-grade tool, and I think it is reasonable to at least offer the option of advanced procedures which can improve results.
Ever used a Wacom Cintiq digitizer? They have a similar multi-point calibration procedure built into the drivers. You can do just a few points, or you can dive in and do 100 points if it matters to you. That’s a multi-thousand-dollar professional creative tool, too.
Right now we do often have to take extra steps to get our placement right. Optionally front-loading some of that work, to reduce the ongoing need for extra work, sounds like a win to me.
Maybe the above concept is unworkable for some reason I don’t yet understand; hopefully, there are still improvements to be made, though.
Just my two cents!
OK, now back to my project…