Making longer flat exposures with ASI294MC Pro: ADU or histogram?

[Menno555]

Slowly but surely understanding more and more. And reading more.
One of the things I stumbled across is that the Zwo ASI294MC Pro produces better flats with longer exposure times then with shorter.
So far I did make my own manual flats with all the necessary settings and the histogram with a peak around 55 and the shortest exposure times. That translates into 0.1 to 0.7 seconds flat exposures, depending on the offset/brightness.
But a lot of 294 users are doing there flats with exposures of 3 to 5 seconds with an ADU of around 25000.

So, is there a way to "translate" SharpCap data into ADU? Or is that simply done through checking with FitsLiberator where the values are the ADU I presume? See screenshot below of a flat file (0.05 seconds, offset/brightness 30) where the mean is 36585 ... which is the ADU if I understand correctly?

Or will lowering the output of the flat panel (yes, I have one now ) and adding more cloth (whit t-shirt in my case) until I have a 4 second exposure at around 55 on the histogram do the same trick?

flat.jpg (227.93 KiB) Viewed 5017 times

[turfpit]

Menno

For flats, I use a mean ADU of ~30,000 for both CMOS and CCD, colour or mono cameras. This tends to result in a fairly centralised histogram. With a colour camera I use the luminance histogram but ensure that the R, G & B histograms are clear of the left and right hand sides.

With a CMOS camera I leave gain and offset (brightness/black_level) the same as for the lights and only adjust exposure using the centralised histogram technique. This invariably leaves the Mean ADU as displayed in FITS Liberator around 30,000. I always check a flat frame in FITS Liberator after capturing a set of flats.

With my CCD camera, Atik 314E, I only have an exposure control. Exposure is reduced until the histogram is around centre - histogram can be described as 50% saturated.

The exposures when capturing LRGB& Ha/OIII/SII flats will vary between filters.

Giving exposure values for flats is meaningless as there are such variations in cameras and other optics. Using the histogram and checking with a tool such as FITS Liberator gives a systematic and repeatable method for deciding the exposure value.

I always use bias frames for processing my flats in DSS/SiriL/APP. I have never had any issues with removing vignetting (5000x3000 field) or sensor or other optics dirt.

For CMOS cameras the above relies on having a good setting for offset/black_level/brightness (QHY/Altair/ZWO). For a good setting for offset I work to having, for a given gain, a bias frame histogram which touches neither left or right axis - invariably this ends up as a centralised histogram for the equipment I own.

The above works for me across a range of equipment but, as usual, YMMV. It took me at least 2 years to get to here.

Flat, Altair 183C:

colour-flat.JPG (59.66 KiB) Viewed 5007 times

Flat, Atik 314E:

mono-flat.JPG (58.79 KiB) Viewed 5007 times

Colour CMOS camera flat in FITS Liberator, Mean ADU = 30,447:

colour-camera-flat-frame-in-FITS-Liberator.JPG (145.83 KiB) Viewed 5007 times

Mono CCD camera flat in FITS Liberator, Mean ADU = 29,542:

flat-frame-in-FITS-Liberator.JPG (123.14 KiB) Viewed 5007 times

Flat with Ha filter in FITS LIberator, Mean ADU = 28,196:

flat-with-Ha-filter.JPG (143.1 KiB) Viewed 5007 times

Dave

[Menno555]

Thanks Dave! Very helpful!!
I was on the right track then
No bias frames here but I do use dark flats.
As for offset/brightness, for now I use either the value the Smart Histogram Brain is suggesting (together with all the other settings) or the ASI294MC Pro default of 30.

Plus, and maybe also thoughts on that, when capturing (dark)flats, I set the White Balance on Auto. This way the color channels are always well clear from either side of the histogram plus no one color is dominant then.

Thanks again,
Menno

[Menno555]

I did make flats and darkflats with this other methode now. ADU set to 28000-ish and exposure time of the (dark)flats is now 4 seconds.
Is it possible because of this that there is less stretching to be done? I tested it on a set I am busy with and I had to strech less to com on the same brightness.

[BlackWikkett]

Hi All,

I've been using the ASI294MC Pro for quite a while now, almost exclusively with SharpCap and live stacking. I've done extensive testing and have found a quick repeatable process for creating flats via the SC flats tool. I make sure to have the smart histogram on and aim to get the simulate luminance (white) channel at 50% on the horizontal scale without getting any other color channel over 100%. I make 50 mono with bias subs to allow the tool to create the master flat. For illumination i'm using a 8.5 x 11 inch Amazon tracing tablet (at lowest illumination settings) through a double sided T-shirt stretched over a 8 inch embroiderer ring.

Settings I'm using for ASI294 on RC6 at 1030 FL with IDAS D1 filter
Temp -10
Brightness 0
Red / Blue White Balance 50 / 50
exposure: 190 mm
gain 300

SC-flat-example.PNG (263.92 KiB) Viewed 4968 times

I've tried using lower gain settings and longer exposures for the subs to get to the same histogram. This will work from what I remember but takes a lot longer to make the flats. I've not found any benefit to doing the longer plus 1 second exposures for flats. Well at least not for SharpCap and live stacking.

-Chris

[Menno555]

It all depends on used equipment and skies
I know that if I use Gain 300 over here in Bortle 8, my exposure times will be very short. Sofar the Unity Gain worked best here.
And no Live stacking here. It works now with the 294 but it's way less fun then using and fiddling with programs like DeepSkyStacker. But that's just a personal taste I guess.
Also, Brightness 0 gives very weird results here.Black levels are way too low then to work with during processing.

[donstim]

I also have not had any problems with short exposure flats at gains from 121 to 250 for Live Stacking EAA in Bortle 6 skies with the 294 MC Pro. I use a varying brightness level depending on gain -- from 4 at 121 gain to 13 at 250 (and 23 if I use 300 gain). Flats are taken with a t-shirt wrapped around a light table at its least bright setting with white channel around 50% and color channels between 20-30 and 70-80%.

[Menno555]

I did some more testing with short and long exposure (dark)flats and I keep on having the same results: I have to do less stretching with longer exposure (dark)flats after stacking. And the flats are less "strong", more fluid.
See screenshot below. On the left a short exposure flat (0.4 seconds), on the right a long exposure flat (4.2 seconds) with all the same settings. These are just a single FITS from a 100 series, so not stacked.
Both made with a LED-panel where I had to use a brighter setting for the short exposure and a lot lower brightness setting for the long exposure flat.

flats.jpg (266.28 KiB) Viewed 4930 times