Fitting with waveform templates

Having low-pile-up data with clear response function, one can build a phenomenological approximation for the individual detectors.

The waveform function f(t, A) for single response on amplitude A can be obtained as a 3D surface in a number of ways, including their combinations:

• Piecewise slicing by A, where every slice (for approximate A value which is not necessarily has energy units) is interpolated with polynomials
• True 3D surface
• A single template transformed with affinity transform (movement, scaling, possibly shear to some extent)

Physical quantities (time, max amp at MPV, charge) can be obtained with same Moyal parameterisation as used currently for analytic fitting.

Although this can of data potentially will introduce a great deal of additional calibration data, it should resolve reflected signal distortions and mitigate some peculiar features introduced by shaper.