For years now, during the development of the needle cutter, I have always recommended 10-15 perforations per millimeter for good cuts in DTFB. Therefore, I personally would recommend slowing to 850 mm/min to get the 10 perfs/mm at 8500 rpm cutter speed. I honestly don't know where Edward came up with 2000 mm/min and 8500 rpm... that should only give 4-5 perfs/mm and looks like maybe that's what you are seeing. I've not made a big deal of it because, for reasons unknown to me, Edward has been cutting foam with surprisingly good results and building airplanes... i.e. so it's hard to argue with his results. But if I was seeing what you are seeing, I would slow down and see what happens. It won't cost you anything except a little time to check it out...