ARCTIC Speed Tests, July 2023
conducted by Russet McMillan
The image overhead for ARCTIC also includes a time for writing the image to disk. This can be a significant fraction of a second, and it may vary at different times of day because the computers at APO are supposed to minimize background tasks at night. Russet tested the full overhead by comparing start times for a sequence of 5s images (this means you should subtract 5s from the times listed below to get the overhead, and the dead times below are around 2-3s). She did these tests - using several different setups - in the afternoon, and the results might be slightly better at night. The last test may be representative of nighttime computer loading, and it had the most consistent overhead times.
Users who are considering non-standard binning should consider the following details:
Test results:
fast read, quad mode, full frame, 4×4 binning, 5s exposures
mean time between exposure starts: 7.55s, std deviation: 0.04s
7.61, 7.53, 7.58, 7.65, 7.57, 7.54, 7.51, 7.53, 7.54, 7.51, 7.53
fast read, LL mode, 4×4 binning, subframe 200×200 binned pixels (~90 arcsec FOV), 5s exposures
mean time between exposure starts: 7.10s, std deviation: 0.05s
7.10, 7.08, 7.07, 7.07, 7.07, 7.08, 7.07, 7.24, 7.07, 7.10
fast read, quad mode, full frame, 3×3 binning, 5s exposures
mean time between exposure starts: 8.04s, std deviation: 0.04s
8.06, 8.11, 8.11, 8.04, 7.99, 7.99, 8.01, 8.02, 8.04, 8.05
fast read, LL mode, 3×3 binning, subframe 300×300 binned pixels (~100 arcsec POV), 5s exposure
mean time between exposure starts: 7.49s, std deviation: 0.01s
7.52, 7.48, 7.48, 7.49, 7.47, 7.49, 7.49, 7.49, 7.50, 7.49