I got you now. That’s what I had assumed initially but the part I didn’t get is that you’re saying that the image you see, no matter how distorted, isn’t representative of the progress being made.
What is the line between “standard lens” and “fisheye lens”? All cameras have some distortion, typically more with wider FOVs. I have the 8mp 120w.
If it needed to be Standard but I selected Fisheye, what consequence would that have?
Lenses that compensate for barrel distortion are “standard”, and lenses that do not are “fisheye”. The math used to compensate for the distortion is different.
If the lens is technically fisheye, but has a very narrow FOV, like a 60 degree, it likely doesn’t matter, because the barrel distortion will be small. If the camera has significant barrel distortion and you use “standard” mode, it won’t be corrected as well, and you’ll lose a bit of precision. How much so I can’t really say.
1 Like
Just a heads up.
It did work with the provided compensation and I was able to to get good accuracy with the alignment.
However the laser is quite large and the alignment test is rather small in comparison, which leads to a few centimetres of inaccuracy in the border areas.
Is there a way to make a bigger alignment calibration and directly calibrate to a bigger area or even the whole working area of my laser?
Engraving 4 squares in all 4 corners and manually adjusting the camera until it overlaps would work too, but I couldn’t find controls for this either.
The shift option is not enough, since rotation, zoom or camera section can not be changed.
Any hint on that?
There’s an option to scale the alignment markers directly within the alignment tool. You actually want to make this as big as possible for the reason you describe.
I would like to scale this differently. I have a 1610 machine and the way this is commonly made, to limit the lid’s mass it’s hinged a bit forward so the last 1/3rd or 1/4th of the bed isn’t visible. The alignment markers ideally need to be able to spread more horizontally than vertically to get the best span.
But I got “ok” results by scaling 350%.
In a separate Topic Oz mentioned that the alignment mechanism requires a square layout, at least in its current form. So expanding horizontally wouldn’t be an option for now.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.