Is Vision Encoder worth? Is there any big difference if you use it?

Just got my Vision Plate and I would say… its not worth it. Honestly, seems like mine got worse. I’ll try to recalibrate it again, but yeah… everything I’ve measure is actually worse than un-calibrated. We are talking about very small differences so its not the end of the world if you do a calibration and its not as good, but it is a little bit of a slap in the face.

All numbers an average of 3 measurements near the same position.

Uncalibrated:
200mm Square - Measured 199.86mm
150mm Hole - Measured 149.95mm
20mm Hole - Measured 19.75mm
5mm wide Slot - Measured 4.95mm

Calibrated:
200mm Square - Measured 199.58mm
150mm Hole - Measured 149.62mm
20mm Hole - Measured 19.73mm
5mm wide Slot - Measured 4.92mm

Also note, this is not calibrated to the filament. Base profile to base profile.

5 Likes

Same here but I’ve only printed a 20mm cube and my H2D is of course like all of us new (less than a month old). It will certainly need that we leave the H2D without additional calibration for several months and then redo the vision calibration to compare. I don’t think we will see any difference on a brand new printer which was calibrated from production.

I have before and after calibration on a 20mm square with a mitutoyo caliper, values ranging from 19.96 to 19.98. If I clean my jaws or push a little bit more or less, I can get on and off 0.01/0.02 so I’d say pretty much well calibrated from start.

It was on PLA basic btw, standard filament profile

It’s probably just a matter of time before someone posts a high-res image of the encoder plate to print at home. $115 is just dumb.

2 Likes

What will you use to calibrate your home (paper) printer that is surely less accurate than the H2D is by default?

5 Likes

You’re going to wind up spending more than 115 dollars trying to recreate it yourself, unless you already have specialized equipment that can print details that fine and accurately.

2 Likes

The type of features/checks you’re referring to are not what this vision encoder is intended to solve.

2 Likes

Curious if this plate was really just for factory calibration, and figured they’d sell it as a feature.

Given the machine is brand new, I’m not finding it does much. I guess time will tell if it comes in handy.

1 Like

Which features/checks are you referring too? Is it in regards to the inner and outer diameter holes from this comment I made?

1 Like

See here Understanding the visual encoder calibration results - #8 by kip

The vision encoder calibrates the motion system.
It does not compensate for material shrinkage and/or wrong profile settings.

6 Likes

Not asking it to compensate… just expecting the baseline, single filament, single file test to be more accurate and it wasn’t, after 3 identical tests.

When you do the same thing… before and after the calibration, and get worse results with the only change being the calibration, you have to accept the data. You don’t make adjustments to get the results you want.

What are your results with a very very stable CF filament? If you have some.

Haven’t tried any CF/GF filament. But don’t worry about shrinkage. Shrinkage will be a constant. The only requirement is environmental and procedural consistency with no significant variances.

I wonder how many people havnt realized there are 2 plastic covers for the plate. The 2nd cover is a super thin peel and barely noticable.

My educated guess says that the 2nd layer needs to stay on.

Nope. Its just a protective peel like the ones that come on most things. The 1st cover is just a storage sleeve

I couldn’t make that up out of your first post. I thought you peeled off more than one layer.

1 Like

Sorry. Just meant it arrives with 2 protective covers. The peel itself is super thin. I almost missed it.

1 Like

Dimentional acuracy is way better with CF filaments. PETG CF absolutely does not shrink as much as regular PETG. Thats one of its main benefits

But do you understand that you are suggesting that the Vision plate only works when you are printing CF filaments? That simply isn’t true. I think you are confusing things.

Shrinkage has no bearing on this test, just so long as the material and conditions are the same, shrinkage is a relative constant (meaning it will be the same in both sides of the test, before and after).

Changing to a CF filament filament will shrink less, but the variance will remain the same. I am only measuring the variance. This variance will be largely all X/Y motion, or what the printer thinks it is doing correctly. The fact that the machine has been very repeatable, just confirms, the numbers are the numbers.

Also with the numbers being very repeatable, I can (at any time), create a profile for the material and get the variance even tighter. But that is not what I’m measuring. I am measuring how well the Vision plate works to clean up the base inputs.

One positive that would come from using a CF filament, would be the lack of a Bambu applied shrinkage adjustment. With PLA, I’m pretty sure they have an automatic shrinkage account, but I’m not sure they do the same for other types of filament. So that would be ideal, and remove all automatic adjustments from the test. But honestly… I’ve seen enough to know what the Vision plate is good for. So, if you have a problem with accuracy, or your machine has gone through significant changes, the Vision plate can help you get back to a good place. But if your machine is working well, there just isn’t a lot in it.