I have created a depth map but when I try to process it creates these weird spikes that are even deeper than the target depth. Does anyone know how to fix this?
I was hoping somebody more knowledgable would reply. Can you post your input file or a link?
I haven’t done any CNC, but for a 3D print, I would suspect some mesh inconsistencies. Since you’re working from a picture, that doesn’t really work. What was the source of the picture? Something you built, or a snapshot with a camera? It’s possible that the camera has some dead pixels, or the photo was taken with a high ISO that produced a lot of noise.
Hello, I finally found out what the problem is. The software converts the color values to depth black is the lowest value and white is the highest value and it sets black to the target depth. However when we add transparency, a more transparent value has a lower numeric value. So a semi transparent black pixel has a value lower than a normal black pixel, causing it to be lower than the target depth.
That sounds like a poor ARGB → Greyscale conversion algorithm. And as the alpha is (generally) the most significant bits in the 32bit integer, setting alpha would make those spots much “deeper”.
Did you use Luban to convert the image, or other software? If Luban, that’s a pretty easy bug to report.
I used Luban, do you have a link where I can report bugs?
I tried to reproduce it for the bug report, but I have a new version and it seems to be fixed now.