Master flat feedback appreciated.

[rcheshire (Rowland)]

I'm interested in feedback on the master flat posted here.

Camera - 1000D, 12 bit A/D converter, maximum ADU 4095 - I've aimed at 2048, which I understand is a good target, though it came in at 2200. The camera doesn't have the sensitivity to improve on this. Some minor adjustments to panel luminance will provide more control. At this stage, I'm more interested in feedback on the overall appearance.

The sensor filter glass was recently cleaned.

Many thanks.

[Bassnut (Fred)]

It looks about right for a master flat, but a bit noisy and possibly clipped. How may subs, and what exposure times?.

There are no dust donuts at all, is it that clean, or way out of focus?.

[rcheshire (Rowland)]

30 subs - 1/30th second, iso800 - bias calibrated - Windsorized clipping - I cleaned the sensor filter glass. I count 3 dark circular disks left, centre and right across the top 1/4 of the image? I took time making sure focus was set to infinity.

I agree about the noise - other than iso I'm not sure of the source - dark calibration didn't show change.

[Octane (Humayun)]

Rowland,

Try ISO-100. Also, bias frames are unnecessary; the bias is included in the dark.

ISO-800 is just far too noisy for flat frames.

Ensure you also take the equivalent length dark frames to calibrate the flat lights.

Somewhere between 17 and 25 for both lights and darks will give you a good signal to noise ratio.

H

[rcheshire (Rowland)]

Thanks H. ISO - my mistake. I have been under the misapprehension that all frames including flats should be at the same iso. Thinking about the reason for flats, that is clearly wrong.

In relation to the bias, PixInsight subtracts the bias from the flat. Dark calibration is done where needed, usually longer flat exposures - it fits in with their particular solution, native to PI, that disregards a whole lot of other parameters. Otherwise, yes, I agree, the bias is in the dark.

[rcheshire (Rowland)]

The flat frames should be created with the ISO speed of the light frames. The temperature is not important.

That's where I read it. The DSS on-line manual, and a few other places. However, there is contradiction among authors here and there - Jim Solomons Cookbook.

I will create a new set of flats for the next batch of lights - the results even at ISO 800 are a big improvement on dark calibration alone. It will take a day to calibrate, register and integrate that lot again - maybe down the track a bit.