Page History
...
Tip | ||
---|---|---|
| ||
If you make a plot of the measured timetool delay it is quite noisy (e.g. the laser delay vs. the measured filter position, TTSPEC:FLTPOS, while keeping the TT stage stationary). There are many outliers. This can be greatly cleaned up by filtering based on the TT peak. I (TJ, <tjlane@slac.stanford.edu>) have found the following "vetos" on the data to work well, though they are quite conservative (throw away a decent amount of data). That said they should greatly clean up the TT response and has let me proceed in a "blind" fashion: Require: tt_amp(evt) amp [TTSPEC:AMPL] > 0.05 AND 50.0 < tt_fwhm [TTSPEC:FLTPOSFWHM]< 300.0 This was selected based on one experiment (cxii2415, run 65) and cross-validated against the current experiment (cxij8816, run 69). |
...