Overview
Although we would need precise calibration, the sensitivity seemed suddenly improved on Mar. 30 morning JST in the silent run (24608), we have no idea why it did. To study the reason, I just compared some relevant channels. No idea I have, but the coherence between DARM and IMMT1 trans, at least, disappeared at 117 Hz and 163 Hz. Anyway we would need more statistics, or more silent run samples.
Details
Fig. 1 shows DARM and IMMT1 trans spectra on the following times:
DARM | IMMT1 trans | |
Mar. 21 00:00:00 UTC | green | brown |
Mar. 28 23:10:00 UTC | magenta | light blue |
Mar. 29 23:35:00 UTC | red | blue |
At least at 117 Hz and 163 Hz, the coherence shows drastical change on Mar. 29 UTC run, which corresponds to the morning silent run on Mar. 30 JST (24608; OLG 24601). On the other hand, there is still (but somewhat smaller...) bumps both in DARM and IMMT1 trans.
Fig. 2 shows time series data of the relevant channels like in 24588. While IMMT1 trans (blue) got some shock like in 24588, such shock was not seen in MCL (brown curve), which is not like the past. Fig. 3 shows the corresponding spectrogram of IMMT1 trans like in 24557; the time 0 is identical as Fig. 2. As already discussed, even no zero bumps can be seen, but maybe smaller.
Recap
I can say the bump in IMMT1 trans would get small but still there. The drastic change would be the reduced coherence. So maybe occasional alignment improvement would cut the coupling path to DARM somehow. I have no idea what is this path; beam jitter to BS alignment? or electronic coupling? Hope more samples of run with improved sensitivity...