A very preliminary simulation of the overshoot of the shaper at 100 ns can be obtained with the function:
Double_t shootfit(Double_t* x, Double_t* fitpar) {
}

where:
x[0] is time in ns (in the overshoot region x[0] > _delta)
yy is shaper output in V


const Double_t shootexponent=-3;

_delta = 300+(0)*qin; // qin < 12 fC
_delta = 300+(0)*12-(8.5)*12+(8.5)*qin; // qin > 12 fC


fitpar[0] = (-0.35)*qin; // qin < 12 fC
fitpar[0] = (-0.35)*12-(-0.13)*12+(-0.13)*qin; // qin > 12 fC