Lid Camera Calibration Beta

This is all happening while you are trying to run the Calibration? Two things to make sure of…
the material needs to be flat (no warp in it) and the material has to completely cover the front of the crumb tray. If the material isn’t pulled far enough forward, the calibration attempt will fail.

1 Like

Yeah, it stops at calibration. I have the Medium draft board (PG) flat on the cutting area. I have the material right up against where the honey comb stops. I’m going to try again, this time with the material right up against where the tray ends, hopefully that works.

Thanks for your help!

Good luck!

1 Like

No luck, still got the “The process was interrupted, so no changes were made to your calibration.” error message. I’m burning through my Draftboards and having to resort to re-covering it with low tack tape…

I also noticed that my camera seems to be tilted. The material is squared up to the edge and I am seeing a significant angle.

Couple of things to check…make sure the lid is closed completely. And check that the tray is sitting down evenly in the dimples.

That canting displacement though might be something that support needs to take a look at.

There’s about a mm a play on the left side, but nothing significant. One more run after making sure the lid camera is seated right, I saw a post that the camera was coming off the lid.

I’ve also reset the wifi connection on the Glowforge.

EDIT:

Sorry if my post wasn’t clear earlier, the camera isn’t loose to the point that I can turn it, I was referring to this post:

Still ran into the calibration failed after this last attempt.

I’m gonna put this issue to bed for now. Thanks for all your help @Jules

The camera shouldn’t be loose…if you can turn it, that is definitely a problem that support needs to look at.

Sorry if my post wasn’t clear earlier, the camera isn’t loose to the point that I can turn it, I was referring to this post:

Very Blurry Lid Camera Problems and Support

Received Glowforge basic today in UK. All looks intact, though hole in bottom of box and two handles missing. Setup went well, got to tutorial ruler print and the camera, including refresh shows below, extremely blurry image. Please advise. [blurry-camera] edit: yup, crumb tray in place, in its “little dimples”. Appreciate the suggestion, always good to check the obvious Three of us here to check process. [glowf-bed] edit: Honestly I hadnt cleaned the lens, simply cos I didnt …

Still ran into the calibration failed after this last attempt.

Okee doke. As soon as someone from support sees your post they might be able to have you try something else. :slightly_smiling_face:

There is another camera on the bottom of the head, next to the lens opening. Have you tried cleaning the two transparent “windows” there? This has solved the “clean the lid camera” problem for several people. (@kevinmcvey, maybe that error message needs to be tweaked a bit?)

3 Likes

My image was pretty warped, too, before I did the calibration.

You’re doing the right thing by re-masking and reusing your material - no need to use a new sheet for each try.

3 Likes

Ahhh I didn’t clean those. I’ll definitely do those two when I get home. I did, remove the head and made sure that the ribbon cable is fully seated.

I appreciate everyone’s help on this. I’ll make sure to write up a troubleshooting guide once I get it up and calibrated.

1 Like

So any updates on my problem? If I satrt another thread, they close it saying to post in this HUGE, unruley beast of a thread. My calibration is terrible…

I’m a little confused about ‘set focus’. Calibrating improved my overall accuracy a lot, but it’s still less accurate in some spots than others – center is perfect, corners are a little off. Even when I use set focus and click it in the center of the material.

Unless I use set focus directly in the center of each little target, in which case it’s oh-my-gosh unbelievably accurate. Like, look at this one from the bottom right (where my machine tends to be the least accurate).

bottomright

And this is the least accurate one still! The rest are so scarily accurate that it doesn’t even look like I cut anything afterwards. Like this one from the center. Holy precision, Batman!

center

But… I don’t understand set focus. How I did these was to set focus for one, individually, and then print them each one at a time. Is that how it works? Or can I put five things in the layout and set focus on each of them and them print them all at once. Does that make any sense? Does clicking set focus in a second spot give the glowforge more information… or does it just override and replace the last measurement it took?

This is not a big deal either way – just curious if you can take multiple measurements or not.

If you only take one measurement at a time, is it better to take it from the center or from the least accurate part of the bed?

The calibration does fix much of the being off, but only in sections. It’s like its making multiple maps(of each GF mark) during calibration and picks which dewarp map to use based on where you want it to focus. But, complete guess.
But the overall accuracy is the same, just we had to do much guessing to figure out how the warped camera would place the images before, now it’ll correct the warp in a spot, place your image, then move to a new spot. It’ll change the focus viewpoint but the accuracy doesn’t change.

That’s the correct way to get the improved calibration where you want it. What the new process did was create a depth map of your machine’s bed relative to the head. That’s why using flat stock was so important. If there was any warp, the map of your machine will have that warp reflected in it so if you had a high corner, the machine calc will think that part of the bed was high.

@rpegg did a really nice write-up of how to use the Set Focus for best results.

Set focus measures the approximate spot you tell it to go to.

Not every machine is perfectly flat, not every crumb tray is perfectly flat, not every material (especially) is perfectly flat and created at exact tolerances.

The set focus helps to rule out variables from all of those and measures at the spot you want to print it.

Have you tried re-running it? Just mask over the same board, you don’t have to use a new one. Make sure it’s flat, and all the way to the front of the honeycomb. Clean the lid camera and the camera on the bottom of the head (not the lens – well, you can clean that too, but also clean the little windows on each side of it – one of those is the head camera) and try it again. I’m guessing there was something off the first time – some warp in the material, or it wasn’t all the way to the front, or too much external light for it to focus accurately, or the lid wasn’t seated right, or there was a fragment in one of the dimples causing the honeycomb tray to be canted, etc. etc. etc.

After another run, and while it is much better than before, it is far from perfect. I am between 125 and 250mil offset now. If I only cut a small object, it is closer since I can focus there, but if I want to do something bigger, the warp is real. Here each of the first 5 targets I did in the corners and middle, they were close since I did them one at a time, and focused on each. Not perfect, but pretty close. Then I did 4 copies of the targets, and you can see that the errors go towards the center. Not sure I can trust material with it only accurate in the dead center, so I am going to still be cutting up cardboard for locating every single print…
Here is the bed with the four targets at once:


And the upper Left one from that image zoomed in:

1 Like

Set Focus only works on one spot at a time, so the way you did it first is correct – focus, score one target, focus, score next target, etc. Doing all four at once will still cause some offset, since there’s sure to be a little variance across the surface of the board.

Your results look pretty awesome to me!

2 Likes