Is the exposure time scaling required?

[simonrw]

I was running donuts on some data with a (wildly) varying exposure time, and got identical performance between altering the exposure time and not.

with:

without:

(upper is X, lower is Y. Exposure times vary from 2s to 30s. Blue points are 2s). Differences are within floating point error.

[jmccormac01]

Richard add this feature in the NGTS implementation. I never had it in the original test code. I copied it over without testing as I assumed it was shown to be useful. I will think about removing this when repackaging the code in the new year.