Testing the camera setup in version 2.1.03

Finally I have had the opportunity to test the new camera setup in LB. I have read a lot about it here in the forum, on our long holiday trip, negative (primarily) and very few positive feedback. It was tingly in my fingers to be able to test it myself.

To make it reasonably “neutral”, I deleted my camera settings from my LB, to be able to start from scratch, just like a new user.

Camera is not found on my Mac, I tried a lot…, but no, LB will not find any other camera than the one built into the computer.

It seems strange, because facetime has no problems switching from one camera to the LB camera.

I can follow “new” users who are disappointed already here when their camera is not found.

To get further in the text, I imported an old camera setup, deleted all calibrations and was able to start the tutorial from the setup itself (the camera itself was recognized immediately with the existing profile)

I tried, as suggested, to calibrate the lens with the targets as described in the tutorial with an iPad, made 18 images but only got 75% hit points. Here I cheated a little, because I know little about the calibration process and redid the test with an engraved version of the test board. New users often don’t want to know that 75% recognition is not enough and only later in the process will be confronted with a bad result.

I reached 93% with the engraved board, not bad but not my record either :wink: I took 22 pictures by the way.

Then I used the tutorial for workspace alignment in auto mode and it was done in zero time.

How precise is the result???

On my 600x400 mm work area I have a deviation from -0.4mm to +0.6mm. Normally I would redo the calibration if it is over +/- 0.5mm but to “simulate” a new user, I just continued.

I do the final test with a target that I freely place in the middle and the 4 corners of my work area. Here I place the target freehand and try to hit as precisely as possible.

I did not use average deviation but the result from the center test, X+0.5 Y+0.1 .

I think the result is “good enough” if you consider that it is a relatively large work area and no fine adjustments have been made.

What makes the biggest difference here with me on my Mac is that it is now also possible to manipulate the image itself, exposure, contrast, gamma… It raises the precision possibilities to another level.

To fine-tune the calibration, I do a more detailed test with at least 12 test fields. Then I can find out where I have what kind of deviations. I can also use that for an average deviation calculation or I can divide my work area into different quality fields for more precise use of the camera.

Conclusion:

I am divided, if I were a beginner I would not be able to get past step 1 - camera finding. Possibly “I” could get help to move on from the forum or LB directly. :wink: :+1:

When the camera is included (found), the rest of the calibration procedure is easy and straightforward to do with a very acceptable result. The deviation is below 0.5mm and I am sure that I can get the tolerance down again to below 0.2mm (with a little extra work)

It is a big step forward that the developers have made - Congratulations and many thanks for this fine tool. :clap:

PS. one small annoying thing, I still keep getting “unmotivated” connection problems - irritating.
(on a LAN connection)

Thanks for the detailed writeup @bernd.dk! I consider it very useful. I have my new camera, so I am going to be one of those “new users” as soon as I get a round TUIT. :joy:

Hey guys…
I had tried calibration also with the new ver LB…
with the larger bed 1300 x 900 it was a no-go
I usually have to down scale the work area fir the camera …ugh it is and 85Deg 8mp

I just ordered a 16mp 100Deg, hoping it will find the tags a bit better…
Not sure I like the new calibration process, BUT really like Bernds idea of engraved patterns…

I seem to have an issue with either too much lighting, or not enough..
Should prove interesting…

Chuck I’m afraid 100 degrees is not enough for your large machine bed, 85 degrees fits my 400 x 600mm with camera mounted in the lid. On the other hand I don’t know your camera mounting height… I hope it works out.

This has been my biggest problem on MacOS/LB as well. I bought an external app that could manipulate exposure and all the other adjustment options, but with LightBurn’s latest version it is finally included for Mac’s as well. Try turning off autoexposure and adjusting the light manually, then you can fine-tune contrast and gamma if necessary.

Right the 85Deg gets about 1100 x 850, so I am good with anything close to 1000 x 800 for center cutting… don’t really care about full bed view… but it is handy for cutting center items…
Maybe the 100 deg will work out a bit better… the lid raises up high enough to get a wide picture… maybe… the camera gets here monday… we’ll see …

Sounds ok. If it works for you you can center your material and benefit from the camera system.
I had to rework/engrave an existing item today, made 4 control marks for fine adjustment and was spot on with my work, no one will be able to see or guess that it was engraved in 2 different sessions with days in between.

You guys keep talking. I plan on doing mine tomorrow and every hint is welcomed!

Mike, you can keep us informed and if you run into any problems, we’ll take it from there.

I know you will. You guys leave no man (or woman) behind :heart_exclamation:

Just designed a solid flip down camera mount. I know you guys say it has to be rock solid and FIXED in position, but my setup does not allow a rigid mounting. I used to be a maintenance man, so I do have a chance of succeeding. Of course, I will photo and report my journey. :joy:

My Monport 130 has side lighting as most lasers do now, and I plan on putting the led strips on a brightness level control or at least ON/OFF .
it is too bright on the four corners when aligning the cameras, the white level smears the edges…

I have had an issue with hinges breaking on my laser.
it plays hell with camera alignment…
also just replaced the gass struts on the lid, these were 130lb rated, changed it to 80lbs
much better… hope it stops the side pressure and twisting…
sure miss the,ole days when stuff was made to last… cheap pot metal, too many impurities…

not a good pic, but you can see the camera (yellow 3d printed holder) on top
it is a good distance from the bed… that doesnt help with the focus…

If possible, I would recommend placing the “workspace alignment template” statically on the machine’s work bed. If there are perfect material alignment stops that form a 90 degree angle on the X/Y axes, the template (which of course must also be right-angled) can be reused later without losing its precise orientation.

In the pictures you can see my solution for the daily material alignment. I use 3mm MDF which is individually adjusted and numbered and stored as templates, when they are too worn I cut some new ones. In the middle are standard spacers.

I have been using it for a long time and just to be safe, I controlled it again today. I followed the instructions in lens and workspace alignment, wrote down the deviation and compared them with an older template. The difference in the deviations is minimal (+/- 0.1 mm) and can easily be due to different lighting conditions. (between the two calibration plates)

When recalibrating, remember to reset the correction factors.

I have also compared auto and manual workspace alignment. Apart from the fact that it is lightning fast in auto mode, there is no visible quality difference. If there was a possibility for further zoom and fine-tuning of the target with the arrow keys with the existing LB steps for that, it could possibly change the game. But as I said, with a deviation of +/- 0.4 total over 721mm in the diagonal… you have to consider whether you need additional precision.

I find the deviation by engraving a small target graphic, fine-tune the camera deviation and engrave the values ​​in the field. I reset the values ​​every time I test.

Finally, you can use the arithmetic mean to find a value that you can use for the entire area or stick to a specific area with the smallest deviation. (in almost the entire middle section of approx. 300x300, my deviation is, for example, less than 0.2mm and requires no further adjustment)

that is really good …:+1:

You’re fighting on two fronts with those conditions :wink:
Hope you find some industrial hinges that can handle the task, or you’ll have problems with inaccurate lid opening/camera angle. And believe me, 1 degree deviation and you can’t use the camera seriously anymore, especially with this distance.

Hi @bernd.dk, in the find New Camera Sources page you will only see cameras which are not already selected as source elsewhere for other camera setups. If you try setup the camera from scratch when there are no other prexisting camera setups for your device, do you still have this issue?

Pay attention to my exposure, to avoid what you describe I sometimes go extremely low.

Boy I,know this too well…
I dont use the camera for anything other than general positioning,
but hoping the new camera with the lower gas’s strut pressure will help a bit…
love how you get the tags to light up… I’ll pay attention to the camera setting you have… :+1:

This is good to know.

For the precision I need, this will probably do.

Have the camera mounted, just need something to keep it out of the way when I flip it up. I hope what you see in the image makes sense. That enclosure drops down and is not nearly as heavy as it looks.

If you can get a repeatable, consistent position of your camera it should work. Your box looks stable enough and with the 2 metal rods (threaded rods?) it should stay the same height. Can you handle the sideways movement?

I look forward to your results

It’s not easy with the big homemade enclosures of our lasers.
I mounted my first cameras this way, it worked ok too. But the area wasn’t that big either.

It is fairly rigid, and stable unless I give it a decent push sideways. I suppose the important thing is that it always returns to the same place when it stops.

But I guess time will tell. :grimacing:

This camera’s auto-exposure seems to work better with cardboard covering the honeycomb, so the entire surface is more-or-less the same brightness.

The front-right corner stil has trouble, for reasons I do not understand, but overall it works well.

I stuck COB LED strips around the perimeter: