[Aavso-photometry] Flat Stuff

[Richard Miles]

> Richard Miles wrote:
> > I also use 2 or 3 different exposures.  So taking a dawn set, the first
> > set are with say 2-sec exposures, and I image in the range 20-50%
> > full-scale. Then I move on to say 0.5 sec, finish that set before moving
> > onto the last at 0.2 sec.  The shutter on the camera is electronic and
> > is in effect <<0.001 sec so shutter speed is not an issue.

Arne Henden wrote:
> If I were you, I would test this.  I have yet to see an electronic
> shutter that was perfect.  Our uniblitz shutters, for example, start
> to leave an objectionable vignetting pattern for any exposure shorter
> than 3 sec.  Flats are the most important calibration you will do,
> and any error in the flats impact all science frames.  Test them in
> many ways.

Arne - thanks for making the point about shutter vignetting, it's a key 
limitation with flats depending on the CCD camera.  The fast readout of USB 
2.0 interface cameras does help to increase the rate at which images can be 
taken and saved to disk.  I use the SXV-H9 camera and image readout/save 
times are between 1-4 sec depending on binning.  My finished flats usually 
comprise an average of 100 frames.

Like you I never take things for granted, so last October on a particularly 
clear night I did some V photometry of Gamma Cas (V=1.8) and took the 
opportunity to check exposure by taking a total of 390 frames at 11 exposure 
times ranging from 0.03 sec to 2 sec.  With scintillation it meant taking a 
large number of frames especially at the fast end of the exposure scale. 
Using fainter field stars (to ensure no saturation at 2 sec) I did the 
photometry (instrumental mags) and found that the exposure was linear with a 
standard error across the range of 0.3% but to my surprise I found that 
(using AstroArt for image capture - it's similar with Maxim DL) the 
'shortest' exposure allowed by the software was 143+/-3 msec.  So setting 
the exposure to say 0.1 sec or 0.03 sec you just end up with an actual 
exposure time of 0.143 sec!  This shortcoming in the software driver has 
been recognised and AstroArt have very recently introduced an update - this 
is also being changed in Maxim DL in their current update (4.07).

So, being able to achieve uniform illumination of the CCD with exposures 
down to 0.15 sec makes twilight flats really straightforward.  The main 
advantage of the short exposure time is that the flats are made whilst the 
sky is relatively bright - so the stars rarely show up in the images and 
there is no need to apply median filtering or the like.

Cheers,
Richard