HaOiii extracted images appear to be "low bit"

Software Siril
21

It’s weird yes!

I looked for zero pixels in bias frames (directly from NINA Stats):
High Full Well Mode: ON | HCG | G_100 O_100 | Min: 22 (1x) → 0
High Full Well Mode: ON | HCG | G_100 O_180 | Min: 99 (1x) → 0
High Full Well Mode: OFF | HCG | G_100 O_100 | Min: 335 (0x) → >300
High Full Well Mode: OFF | HCG | G_100 O_180 | Min: 24 (1x) → 0

With last settings my bias stats are:
B&W layer: Mean: 181.4, Median: 181.0, Sigma: 6.0, Min: 18.0, Max: 338.0, bgnoise: 4.7

I should probably go for a little higher offset just to be save I guess (especially since I don’t loose that much dynamic range and don’t really need it either).

I also looked at randomly picked single and stacked exposures of my data sets including calibration frames and lights and these are the results:

High Full Well Mode: ON
High Gain Mode: ON
Gain: 100
Offset: 100
Temp: -10°C

Bias:
single: B&W layer: Mean: 101.0, Median: 101.0, Sigma: 3.4, Min: 20.0, Max: 181.0, bgnoise: 3.0
master: B&W layer: Mean: 101.0, Median: 101.0, Sigma: 0.8, Min: 82.0, Max: 156.0, bgnoise: 0.7

Dark (300s):
single: B&W layer: Mean: 102.2, Median: 102.0, Sigma: 22.5, Min: 17.0, Max: 50863.0, bgnoise: 3.3
master: B&W layer: Mean: 3.4, Median: 3.0, Sigma: 22.5, Min: 0.0, Max: 50701.0, bgnoise: 1.0

Dark (600s):
single: B&W layer: Mean: 102.7, Median: 102.0, Sigma: 42.7, Min: 25.0, Max: 51049.0, bgnoise: 3.4
master: B&W layer: Mean: 3.9, Median: 3.0, Sigma: 41.9, Min: 0.0, Max: 50908.0, bgnoise: 1.1

Flat:
single: B&W layer: Mean: 23247.0, Median: 23091.5, Sigma: 7454.1, Min: 12887.0, Max: 34097.0, bgnoise: 10140.7
master: B&W layer: Mean: 22926.2, Median: 22960.0, Sigma: 8306.2, Min: 8976.0, Max: 33515.0, bgnoise: 11114.2

Dark Flat (matching the flat):
single: B&W layer: Mean: 101.1, Median: 101.0, Sigma: 4.0, Min: 20.0, Max: 9605.0, bgnoise: 3.0
master: B&W layer: Mean: 101.0, Median: 101.0, Sigma: 2.0, Min: 75.0, Max: 8868.0, bgnoise: 0.7

Light (600s):
B&W layer: Mean: 131.8, Median: 129.0, Sigma: 115.1, Min: 43.0, Max: 52146.0, bgnoise: 11.9

Light (300s):
B&W layer: Mean: 119.2, Median: 117.0, Sigma: 105.2, Min: 39.0, Max: 51069.0, bgnoise: 9.1

I think I will try different process combinations. Maybe I get better results then.

22

“High Full Well Mode” lowers the gain (as expressed as e/ADU). So gain 100 with HFW+HCG will require longer subs vs HCG with HFW off.

In terms of overall gain at the setting gain=100, LCG+HFW < LCG < HCG+HFW < HCG

HFW mode does not change anything about the pixels or the electronics, it just changes the gain range. To me, it only makes sense to use HFW with LCG - as a way to allow longer subs on really bright targets or with really fast optics.

23

Thanks for the clarification, makes sense.
I will try with HCG, Gain 100, Offset 250, HFW off and hope to get better results then :slight_smile: Thank you!