Corrected typo for C01 filter production. Option for Blackman window function.

The main purpose of this MR is to correct a typo in a variable name in the DCS filter generation code. It appears that this section of the code has not been thoroughly tested for a while. Secondly, but less importantly, I added the option to use a Blackman window for FIR filters.

requested review from @louis.dartez

assigned to @aaron-viets

note lint CI failure.

ideally the fix and the blackman window changes would have been separate commits, but this is probably ok.

We should definitely add tests for this code. Might as well add them now.

added 1 commit

• 3760cf27 - make lint happy

Compare with previous version

added 1 commit

• 7796136a - fir.py: fix for response corr

Compare with previous version

After fixing the response correction noted in the last commit, the comparison between the response function applied by the filters and the modeled response function looks as expected:

mentioned in merge request Calibration/ifo/L1!12 (merged)

mentioned in merge request Calibration/ifo/H1!10 (merged)

added 1 commit

• c72ef36e - fir.py: Corrected condition for when to apply response corr

Compare with previous version

added 1 commit

• 12756e4c - fir.py: Added cutoff frequency for response correction

Compare with previous version

added 1 commit

• a69e6134 - Undoing unintentional change

Compare with previous version

One more tunable parameter added in the second-to-last commit allows us to apply the full response function accurately (including inverse sensing and actuation) up to a certain cutoff frequency. Since there is a very sharp spectral feature just above 4 kHz from the suspension dynamics that we do not understand and cannot resolve very well, I set the cutoff frequency to 4 kHz. The plots below show a comparison between the response function applied by these filters and the new proposed model around the time of S240925. A significant discrepancy in the 'GDS' data is expected, since the majority of the model in this calibration mode was applied in the front end (a model we now think is inaccurate). I don't know whether this exact curve is the expected one. Perhaps the best check would be to compare the 'DCS' calibrated data with the monitoring lines.

added 1 commit

• 763de30d - remove print statement

Compare with previous version

@louis.dartez Since you are listed as the reviewer, what is preventing this from being merged? We need this for C01 production, unless we just use this development branch to make the filters we need.

thanks for the ping @aaron-viets. the MR is merged

approved this merge request

resolved all threads

mentioned in commit c3c07caf

merged