Stefan, Sudarshan, Evan

After recovering low-noise locking, we noticed something was wrong with the calibration: the region from 10 to 30 Hz was too low, and the region from 50 to 200 Hz was too high.

We looked at the calibration lines, we took a DARM OLTF, we compared SDF settings, we inspected foton coefficients, but we could not figure out what was wrong. [Actually, the EY L3 lock gain in the CAL CS model was 1.0 instead of 1.25, but that wasn't the root of the problem.]

In the end it turned out that FM1 of CAL-CS_DARM_FE_ETMY_L1_LOCK_L has a pole at 0 Hz, and the filter had charged up, causing the CAL CS version of the L1 control signal to show a fake, white-ish noise floor above 10 Hz.

We turned the filter off and then back on again, which fixed the problem. However, we can already see that the filter has started to charge up again. So we've replaced the 0 Hz pole with a 0.001 Hz pole.

Sudarshan also found that the CAL CS version of the whitened DARM control signal had a fake 1/f shape between 8 and 80 Hz, which was due to insufficient whitening. So now there is an extra pair of 1 Hz zeros and 100 Hz poles in CAL-CS_DARM_DELTAL_CTRL_SUM_WHITEN.