[Matt, Jenne, Elenna]
Matt and I tried running a new jitter training on CALIB STRAIN CLEAN, and noticed that we could likely subtract a bit more jitter noise from CLEAN, and also probably reinject less noise, especially around the power lines. We then reran a training on NOLINES to prepare to apply the new cleaning today.
Jenne then walked me through how to generate the script which would update the cleaning parameters. I copied over the old observe.snap file, in case I made a mistake and need to revert to the old coefficients.
Once we were in observing, I ran Jenne's template which checks how the cleaning is performing. It definitely looks like it is doing a good job. I accepted the new parameters. Once we are locked for a little longer, I will generate some comparison plots to the old cleaning parameters.
Here is a ratio comparison plot, showing the ratio of CLEAN / NOLINES for an observing time last night and today. We're still thermalizing, so this may change slightly. Matt and I adjusted the plot colors and alpha so the old cleaning is on the bottom and you can see where we are no longer injecting extra noise, and also where the cleaning has improved slightly, especially for the peaks at a few hundreds of Hz.
