This is a comparison between the output of GDS strain and Strain as predicted by Pcal, both in magnitude and phase. The data utilized here is obtained from SLM Tool and some known corrections are applied before calculating the ratios.

1. Data Retrieval: The appropriate channels, listed in the link below, are read from the frames, and 60s FFT is done to obtain amplitude, uncertainty and phase at the calibration line frequencies. This FFTed data is stored as a txt file at location in the link below.
   • Data: https://ldas-jobs.ligo-wa.caltech.edu/~gmendell/pcalmon/daily-pcalmonNavigation.html
   • Channel List: https://ldas-jobs.ligo-wa.caltech.edu/~gmendell/pcalmon/daily/pcalmon_2015_12_11_06_04_43_PST_Fri/data/H-logfile-pcalmonH1Data-1133791500-10800.txt
   • SLM Tool description: https://dcc.ligo.org/LIGO-P1300120
   •  
2. Data Analysis: The GDS-CALIB_STRAIN produces ‘Strain' so no further processing is needed. However for Pcal signals (photodiode readouts), some post-processing is needed to convert it into ‘Strain'.
   • Following Corrections are applied:
     • inverse of Digital and Analog Anti Aliasing filter.
   • • pcal signal is whitened so needs dewhitening.
     • correct for IOP time delay
     • correct for time delay coming from model difference. (there should be a  61 us delay between Pcal and GDS which was realized at a later date so the data analyzed here did not need this correction). 
     • divide by arm length to convert displacement to strain
   • All these corrections are done using the matlab script xpcalcorrectionL1/H1. 
   • After the corrections, the ratio between GDS and Pcal is taken at all calibration line frequencies.
   • The scripts used to produce all these ratio plots are located here: https://svn.ligo.caltech.edu/svn/aligocalibration/trunk/Projects/PhotonCalibrator/scripts/SLMTool/
   • Each set of plots contains
     • Time Series of all kappas calculated using the equation in T1500377 and the data from SLM Tool . (Page 1)
     • Time series of relevant kappas (kappa_tst, kappa_c and f_c) and their histogram. (Page 2)
     • Plot of GDS/Pcal, both magnitude and phase ( including time series and histogram) at 4 calibration line frequencies. (Page 3-6).
3. Results for H1: 
   • Representative 5 days were picked between Sept 11 and Oct 21 ( 09/18, 09/23, 10/05, 10/11, 10/14, 10/18).
   • The ratio of magnitude is closer to 1, within 5%, and any discrepancy is a direct consequence of change in actuation function reflected  in kappa_tst.
   • The phase shows some discrepancy until 19 Sep, which corresponds to 13 us, because GDS was not corrected for arm-delay. 
   • https://ldas-jobs.ligo-wa.caltech.edu/~sudarshan.karki/GDSPcalCompare/H1_SLMTOOL_Analysis.pdf

All these analysis were  done in C00 data and I am sure these discrepancy in phase will disappear on C01 data. I will analyze and confirm it once the SLM data for C01 is available.