For me, signal processing offered absolutely nothing over image processing and my own reading of brightness values. It was incredible to see that Octave, Scipy, LagThresholdInfluence, and PeakDetection template smoothing programs not only ignored symmetry, but punished it. Ha. I am hoping that a “model plot” , a conscensus made from a dozen or so nice AFM images of surfactant protein D can be used to make an accurate assessment of the number of peaks along the trimers and hexamers in other images.