Hi,
I had my first use of my ZWO ASI2600MC last night. As it was almost full moon this was a practice to iron out any problems - this is my first dedicated astro camera, I’ve been using DSLRs up until now.
The only problem I have seems to be with my flats - they’re under correcting.
Yes, I know that’s a view with Histogram mode on. I’m using it to emphasis the problem and check the result as I’ve tried changes.
I followed the recommended nominal 32k ADU mean using an LED flat panel. All captures were taken with KStars/Ekos/Indi at the same gain (100), offset (50), and temperature (-10).
All Siril stacking has been done manually to the tutorials using a synthetic bias of 500.
I suspect the problem might be due to my LED panel being too bright so that I had to use a very short exposure for the flats - any opinions? Also if that is the problem, is there any way to rescue the flats that I have?
One of each light, dark & flat frames can be found here.
Hello, I don’t understand what can be happening. On a single image it’s not as bad as on your picture, but still slightly visible and it often gets worse by stacking.
The flat and light colors seem much alike, deficient in the red strangely. Are there some white balance settings applied on the camera?
I’m just wondering if 500 is the correct value, usually it’s a factor of a power of 2, and here the dark has a median of 502 for 30 seconds, that seems very close, but not impossible…
I’ve set the colour balance to 50/50 in the indi driver. ZWO defaults it to 52/95 which is much too blue (from what I’ve read on CN & SGL).
I did the whole take different offset bias frames, export and plot business as described at https://siril.org/tutorials/synthetic-biases/ just to confirm the value to myself even though that tutorial happens to mention:
…for a 16b camera, there is no way to tell. For instance, for a ZWO 2600MC, the multiplier is 10.
For offsets of 30,40,50,60,70 I got median values of 302,403,502,602,702. I did try using a value of 502 as well as the stated 500 as I wasn’t sure of the reasoning behind the following (also from the synthetic biases tutorial, my emphasis):
If I add a linear trendline, forcing the intercept to be 0, I get this plot
Using 500 or 502 made no visual difference. I also tried entering the offset as both a dynamic value ( =10*$OFFSET) & static values (=500, =502) in case there was an error going on there. No difference.
To try to investigate I also tried manipulating the brightness/contrast of the stacked flat in GIMP to test the effect. I found that virtually any (like a 0.005 multiplier) reduction in brightness or increase in contrast resulted in extreme over correction of the lights.
