Using one minute of free swinging cavity (not very well aligned) from gps 951663630, I run a fit algorithm to all resonant peaks, extimating finesse and speed. The first three plots show the typical results (circles are real data and the line is the fit). The model can reproduce surprinsingly well the data.
The 4th plot shows that there is a correlation between the reconstructed speed and the reconstructed finesse. One expects that higher speeds give a larger number of ringings and therefore a better estimation of finesse. A new measurement will be tried forcing higher speed of the mirror. Negative speeds should be an error of the fit, which should be forced to positive values only.
The 5th plot compare the absolute value of the speed reconstructed with the fit algorithm with the speed reconstructed from the time interval between consecutive resonant peaks. There is a quite good agreement.
Conclusions: this fit algorithm works quite well and we should try again the measurement with better alignment of the cavity and kicking the mirror to have larger speeds.
The reconstructed finesse (considering only high speed) is around 150 as expected.
