The CFHT Adaptive Optics Bonnette :
The main difficulty comes from the fact that the interferometer that is used to do this measurement is very sensitive (as it should be!) and is easily affected by vibrations. A vibration damping set-up was tested and modified ad-hoc, and every pump, cooler, compressor, etc. in the telescope had to be turned off.
A first glance at the newly measured influence functions doesn't show much difference from the old ones (i.e. no mirror aging). However the data reduction is tricky because the interferometer was not at the exact same position as five years ago, and so some rescaling has to be performed to compare the two data sets. This might actually hide some scaling difference.
More detailed analysis reveals some differences between the old and new influence funtion measurements: when taking the laplacian (double derivative) of the phase maps, to reveal the area of curvature, the new influence functions show less signal (which is not consistent with this measurement), and it is also less clearly defined. Furthermore, using the new influence functions it is not possible to remove all the static aberrations of the deformable mirror, the residual of which has a cos(theta/12) pattern. This explains why some PUEO images have a six-fold pattern on the first Airy ring. Is this effect new? If so what can be producing it?
A major re-alignement of AOB is currently in discussion. This will be the right opportunity to also test the deformable mirror with a better access to it.