Here I’ve attempted to show how both the NTDA and NMDA are susceptible to vertical data smearing of velocity data. Cursor square locations refer to the position of the overlapping NTDA/NMDA detections. The 0.5 cut is to the left and the 2.4 degree cut is on the right. This is a tough problem, with perhaps Spectrum width providing a sanity check for the NMDA at least

-marfafront