We had to pause observing at 21:40 local time to refill LN2, because its level was decreasing faster than expected and the pressure in the vessel was greater than normal. In this procedure, I did not hit "stop" exposure but "pause" followed by "resume" — to hit "stop" after resuming. As a result, the system was unhappy and it was not clear if Vsub was off at this time. Therefore, we had to restart SISPI. We took a few 10-second exposures because focus was very soft after restarting. Nevertheless, I noticed that the system was more stable and we did *not* experience any of the problems as in the previous nights (such as hexapod interlock or incomplete readout) that would require to restart SISPI several times during the night.

We experienced some images with elliptical (coma?) PSF in the r filter (150 sec exposures). This happened for the images of one fields in NW position (images 1133945–949). The largest ellipticity measured was ~0.36, with several images with ellipticity >0.2. We suspected that this could also be caused by wind because we were pointing ~20 from wind direction, although wind speed was only around 3.0–3.5 m/s. We slewed to a different field (SW) and ellipticity improved immediately to 0.04 (images 950 and 951). We were back to the NW field and reduced guider exptime to 150 ms. Ellipticity was back to 0.09 and 0.07 for images 952 and 953. We then raised the lower shutter and closed the lateral openings for image 954 but ellipticity was still 0.06. We increased guider exptime back to 600 ms and got ellipticity 0.07 (image 955). Finally, we opened the dome trying to replicate the conditions when the problem appeared, considering that wind speed and direction were similar to image 945. However, ellipticity was 0.00, 0.02 and 0.03 for images 956–958. We do not know what caused this problem, but it did not appear for the rest of the night.