Complete results from this DQ shift may be found at here. 

• 7 - 8 Jan were characterized by a total of three long locks with a combined duty cycle of 56%. The first lock showed excess noise in the 8 - 20 Hz band which seemed to correspond to rung up bounce and roll modes. There were some ETMY saturations (range drop glitches) and the glitch rate appeared typical. 

• One issue on those days was the appearance of a line at 1500 Hz which could be the second harmonic of the violin mode. The first harmonic was present but not visible in the h(t) spectrograms. A comparison of both lines from 1 Jan and 7 Jan shows some differences including an increase in the amplitude scale by a factor of 10. 

• On 9 Jan, just before 10:00 UTC, another RF45 episode began and dominated the properties of h(t) until the final lock on 10 Jan. The initial glitch was well documented in the alogs. The loudest BBH triggers on 9Jan had newSNR > 9 but, from the omegascans, were clearly a consequence of RF45-induced periodic glitching reflected in h(t). BNS triggers were not available on 9 Jan. 

• The first lock on 10 Jan seemed to have excess low frequency noise from bounce and roll modes rather than RF45. This was determined from the SUS bounce and roll mode spectrogram, the absence of strong low frequency noise in the h(t) Rayleigh plot, and a fairly nominal glitch rate. With RF45 active, the Rayleigh plot shows strong low frequency features and the glitch rate is greatly elevated.