Vent: alignment :(

Are you saying that different browsers yield different results on the Glowforge even when you are using the same design?

Safari is my primary browser.

There is/was a bug where zooming in on the image caused the preview to shift on Safari. There is/was also a bug where you couldn’t reorder operations by dragging them around in the list on the left side of the window. I don’t know whether either of them has been fixed. (It sure would be nice to have a comprehensive set of release notes for the GF UI, including user-visible bug fixes.)

3 Likes

It doesn’t look like the head camera is ever going to get used for alignment. I wonder if the resolution is too poor.

1 Like

I don’t recall reading that anywhere. Do you happen to remember when they said that?

1 Like

The only reference I found to the head camera is that it is one of two cameras used to align the head and materials. I don’t think I’ve ever seen a reference to us having an interface to it so we can utilize the head camera through the GUI.

Maybe it’s just not a thing.

I tried asking about the head camera being used for registration over here… A response was given.

I haven’t read all the comments here so forgive me if it’s been mentioned. Yes, the alignment issue seriously needs to be fixed, but don’t be afraid when you are doing multiple operations on the same uploaded file. (Such as wanting to make another engraving pass.) The alignment looks off after the first operation, but in my experience the files will still align and cut correctly in subsequent passes. (Obviously you can’t move anything on the bed.) Just wanted to make sure you knew that as I wasted a few projects before I realized it :slight_smile:

4 Likes

No because you can see it is off in the camera image. Obviously resolution decreases in the corners but it should still be good enough to place you design on the material to within a camera pixel.

I was referring to the resolution of the head camera, not the lid camera.

Again, any Official mention of the head camera being utilized by the operator for cut/engrave alignment is either intentionally avoided, or not spoken of.

Either, they can’t, or won’t.

Maybe it’s something they can implement later if they can’t get the lid camera to work properly, unlike not putting limit switches in to correct wandering head alignment issues.

Sorry.

The head camera has much better optical resolution over a smaller area. It may have less pixels in total but certainly the claimed resolution was better than the lid camera in the corners:

Able to view one square inch with resolution of 0.002” (0.05mm)

However, like a lot of things, this has been removed from the technical specs.

1 Like

Well, that ought to be workable. I wonder why it seems like they don’t seem to be planning on using it.

I think it is planned for pass through alignment and doubled sided cutting but it would be a bit clumsy positioning your artwork using a 1" peep-hole view unless it happened to be smaller than 1".

If they present hi res images from the head camera, it would not need to be limited like that. They could scan an area of interest, stitching together multiple images to give a full screen view of the magnified area.

True but that requires some very clever stitching. You only have to look at the Muse to see how to do it badly. Not only do you need to de-warp it perfectly so edge pixels line up exactly, you also have to match the brightness and contrast due to lighting changes.

Anything that has any 3D depth would look different from different view points, so you would need a full depth scan and build a 3D model to make it look right.

2 Likes

I’d love to be able to zoom into a 1" square to find the origin. If the head camera worked like an optical edge finder you could use it to establish corners, center points, change the angle, pretty-much anything you could want. It would be a lot like zooming in to position your work, but… you know… actually accurate. You could get an overview of the bed using the lid camera, click on the part you want to zoom in on and the machine would move the head camera to that position for fine tuning.

I still hold out hope that this is in the pipeline.

… is still in the spec list. Even if it doesn’t go down to 50 microns, it would still be a big improvement.

4 Likes

Does the GF have a visible laser pointer on the head? Imagine a manual registration mode where you drive the head (and visible dot) with buttons, park the dot somewhere on your material and tell the machine, “this is the upper right corner of the artwork.”

Now you are not relying on a preview: you have TOLD the machine where to go in trustworthy terms that it understands, the location of the head.

To confirm the cut plan, the GF could do a quick physical preview, with the visible laser dot simply moving around the edges of the artwork bounding box… or even around the outer border of the artwork, if it is a closed shape.

My vinyl cutter works kind of like this and it makes it extremely easy to accurately cut things out of odd scraps. It doesn’t do a motion preview, but you drive the head around and push a button to say “origin is here” and then Bob’s your uncle.

I love the idea of the camera, it is one of the reasons I pre-ordered… but if placement via the camera doesn’t become automagical, we are going to need a Plan B. People are already resorting to things like doing a quick score on the masking material to verify placement. We’re already taking extra steps. If those extra steps are going to be a way of life, then the UI should find a way to make them as easy as possible, however that may happen. I am sure there are many solutions.

2 Likes

I can glue a laser onto the head that gets me better than 1/4 inch accuracy. Maybe I can 3D print up a kit.

1 Like

But it’s ludicrous to need to do so.

2 Likes

Especially as the head already has a red laser pointer in it.

1 Like