The calibration team has posted version 3 ("final"!) calibrations for 
all three interferometers in the calibration web page, in the S2 page:

http://blue.ligo-wa.caltech.edu/engrun/Calib_Home/

Notice that there are two different epochs for H2; each epoch has 
different reference actuation A(f) and sensing functions C(f), but the 
same reference open loop gain G(f). This is due to a suspension 
diagonalization done in the run.

The changes in the response functions for H1 and H2 with respect to the 
latest "version 2" calibration are less than 10% in magnitude and 5 
degrees in phase, and change with frequency with a mean of 1 for 
magnitude and zero degrees for phase. The changes for L1 with respect to 
v2 make the response function in strain/counts about 15% higher  in 
average, mostly due to a difference in the actuation function used in 
the original model and in the new (correct) model. Changes in phase are, 
like in LHO, less than 5 degrees. Notice that this will make sensitivity 
estimates about 15% *worse* than with the v2 calibration.

The time series for alpha coefficients for all three detectors are less 
than 1% different than in version 2. The statistical errors of these 
coefficients change in time, becoming smaller when  the calibration 
lines were increased in H1 and L1.

--L1Alpha.Beta.txt: Statistical errors in alpha: 3.4% before 731714478, 
<0.7% afterwards.
--H1AlphaBeta.txt Statistical errors in alpha: 9.8% before 730793022, 
2.9% afterwards.
--H2AlphaBeta.txt Statistical errors in alpha: 3.5%

Some groups interested in calibrations over time scales longer than a 
minute (more likely pulsar and stochastic) can benefit from averaging 
the coefficients over an appropriate intervals of several minutes; we 
are leaving this choice up to the search groups, starting from the 
series we have produced.

We will keep making consistency checks, investigations on better methods 
to estimate alphas, calibrations versus time, etc; but we will apply 
what we learn only for better error estimates in S2 calibration and for 
better calibrations in S3; we will not likely change the "final" 
calibrations posted now.

We have *not* yet verified the calibrations during injection times (when 
calibration lines had different amplitudes), we will posting this 
information soon.

Let us know any questions you may have.

Gabriela (for the calibration team)