Hi all, astronomer!
Playing with Nikon D810
a.
Nikon did it
wrong again.
No "fair", unprocessed RAW (NEF) again.
They use some algorithm to scale/offset data based on signal from "dummy" (for bias) and "optically black" (for dark) pixel columns.
With this mean value for any dark exposures (with reasonable ISO and duration up to 900 seconds) is close to 600ADU...
Left "wing" of pseudo-Gauss histogram distribution is cut and filled with some value (equal to scaled bias?).
Histograms of master bias (x128) and master dark (900sec, x15) for ISO800 are shown.
Nikon D810a ISO800 Master Dark & Bias Histograms by
Yuriy Toropin, on Flickr
Another point to mention - there is no real HOT (saturated) pixels on the master dark at all. The maximum value in CFA is ~5000.
Looks like they could additionally filter out saturated hot pixels.
Or - may be they specially select only the perfect sensors for D810a, without hot pixels at all?! Thanks, Nikon!
I've tried Mode 3 (switch ON Long Exp Noise Reduction mode, then turn camera off during Job NR). Surprisingly, I've got only 1 (one) frame with "nice" histogram out of 6 tries.
Something goes there also.
***
Anyway, dark current enable tool is needed for the Nikon's D810a real "astrocamera".
astronomer, is it doable? Can I somehow help you with this?
(Own D750, have D810a on testing from Nikon local representative office)