Related to my post in 66118, processing the HWS dither test data from 66077, it appears this could be a reasonable way to locate the centre of the test mass on the HWS images. Previously we have been running a long ring heater test over a few hours when the IFO is down, and calcuting the center the test mass to be the centre of measured OPD in the HWS images.

Instead I have recalculated the HWS wavefonts during a dither test without removing the tip/tilt parts of the OPD. If you do this and average the HWS images during a pitch/yaw dither test, and simply overlay the two images you get a crosshair shown here for ITMX and ITMY. The center of the crosshair should also be the centre of the test mass. In the above images I've just used the numpy.min() search on the array to estimate the center, but a quadratic fit to these profiles would probably give better (and sub pixel) accuracy.

For comparison here are the results from the recent ITMX_ringheater_test & ITMY_ringheater_test which seem to roughly agree. The benefits of this dither method are that we don't have to wait for the IFO to dethermalise and rethermalise for the ring heater tests, and doing the dithers only take a few minutes.