Obtaining a Sky Flat
Below we show a series of images taken as sunrise approached on 20010630. D(time) is the time the exposure was taken after the first frame; sunrise occurred at D(time) = 12 minutes. Note also that the telescope was AT ZENITH.
At the bottom of the page there is a plot of the count rate per three-second exposure, in which the sharp increase in counts just after sunrise can be clearly seen. Note: these data were taken in NDSTARE with NORMAL Gain. Hi-Gain will give twice as many counts per photon!
Run | Image | D(time) in minutes | Mean Count (in 3 sec) |
329 | ![]() | 0 | 0 |
330 | ![]() | 1 | 2 |
331 | ![]() | 3 | 2 |
332 | ![]() | 5 | 3 |
333 | ![]() | 7 | 3 |
334 | ![]() | 10 | 3 |
**335 | ![]() | 12 (sunrise)** | 3 |
336 | ![]() | 15 | 3.5 |
337 | ![]() | 17 | 5 |
338 | ![]() | 19 | 7 |
339 | ![]() | 21 | 10 |
340 | ![]() | 22 | 12 |
341 | ![]() | 23 | 12 |
342 | ![]() | 25 | 16 |
Sky counts through the F-P with an exposure time of 3 seconds (F-P set for H2 1-0 S(1) imaging; NDStare + Normal readout). The x-axis records the time in minutes; note that Sunrise occurred at D(time) = 12 minutes. The y-axis is the counts measured in 3 seconds. Consequently, a 30 second exposure time would have given 10-times as many counts on the array. | ![]() |
And Finally…
A recent observer had the following experience while taking a sky-flat at sunrise: “We did our skyflat yesterday morning with the telescope 3 hrs over pointing towards the sunrise and got just the right counts (1000-2000) in 60 sec by starting the sequence right at sunrise (this was by accident as we were planning to do it 12 minutes after sunrise but just as well we didn’t; just noticed that your webpage does actually say with telescope at zentih). Tried it just now with telescope 3hrs to west at sunset. We started it a few minutes before sunset but it was too late really (300-1000) counts. With telescope in that position need to start it 5 mins before sunset I reckon.”