Darks

Evora has a dark current of about 4 counts/minute*. We're still learning about the dark current; for small exposure times using darks could introduce more noise than it removes. Consider that for aperture photometry the tradeoff is reducing the structure of the dark current versus adding the noise inherent in any dark image. From Phil Massey's CCD Reductions with IRAF: 

...you may want to take an equal length dark frame, subtract a bias frame from it, and decide if you are worried about how much dark current is left. . . . I usually take a few of these but never use them. Applications where dark current will matter are long-slit spectroscopy and surface brightness studies--cases where the background is not removed locally.

 In that case, A. Charles Pullen in The Zen of IRAF recommends making a master dark, but with a warning:

If you make a master dark per Berry and Burnell (2000) (25+ images at an exposure time at least 5 times your longest exposure) IRAF can use a single median combined image, the “master dark”, and apply it to all your frames. It will adjust the value of the dark subtraction based on the ratio of the exposure times. This is called a "scaled dark". Note, that you must subtract a master bias image from each of your darks prior to combination.

There are potential problems with scaled darks. You are trusting that your dark current is truly linear, that there are no light-leaks in the system, and that the temperature is really constant.

Fortunately, Evora's dark current is linear, at least up to an exposure time of 35 minutes. For this and other information about dark current on Evora, see the Jupyter notebook below. A typical dark from Evora can be found directly below it, as well.

*It was previously reported that a larger number of counts appeared in a vertical band between lines 200 and 500, but the most recent test was unable to confirm this.