Can’t say that I have ever seen the “hills” in any of my projects like you have here, but I also have not “stretched” a gradient out over as long of an area. My tests were basically about an inch long at the longest, others were 1/8" at the shortest, and using the whole power range. This effectively condenses the powerband into a smaller area and might make such “hill” defects unnoticeable.
What I have seen in my work is that as the power increases and the laser cuts deeper, it is not linear, but it is a very shallow acceleration curve. However, multiple passes also amplify the non-linear power curve. In other words, say you run two passes, the shallow end will get a little deeper but the deep end will get REALLY deep.
Clear acrylic is a good material to help illustrate because you can look into the cut edge of the material and see how the engraving changes depth.
