Issues with short exposures and flats

[metastable]

Hey Brian,

I'm not sure why that is. I just use SharpCap's Capture Dark/Flat tools. I guess Darks use the Output Format of the camera? I haven't really changed that setting since I started using SharpCap. I'll test it out when I get a chance.

Alejandro

[admin]

Hi,

an over-correcting flat is very likely to be related to issues with incorrect dark subtraction - either :

1) When making the flat, the darkflat or bias was brighter than it should be

or

2) When capturing lights, the dark frame is darker in the background than it should be (taken at a lower offset/black level is often the case).

The second is easiest to understand - if the dark frame is too dark, once subtracted it leaves the result with bigger pixel values than would be expected based on the light falling on the sensor - maybe all the pixel values after subtraction are approx 20ADU too high as an example. The flat frame brightens the image in the darker areas when it is applied, but it will apply to the actual image data *and* the extra 20 ADU that are present too. In the darker areas of the flat, those 20 ADU get boosted more than in the lighter areas, so the image gets brightened too much in the darker areas of the flat. When stretched you see this as an inverted correction.

The first basically has the same effect, but happens when creating the flat frame (the darker areas end up darker than they should be due to excessive subtraction, so get more correction applied).

cheers,

Robin

[turfpit]

Alejandro

Downloaded the files and the outstanding item seems to be the difference in gain settings between the light (=0) and dark (=100) frames.

As a general guideline I set the exposure, gain & brightness for the light frames, keep these exact settings for the dark frames and adjust only the exposure for the flat frame.

All frame types were captured as RAW16 and are all the same size. The flat looks ok with a mean ADU=30,000 and no histograms touching the vertical axes.

• Light & flat are FITS but the dark is TIFF format. There must be a reason for this?

• The light histogram is up against the left hand vertical axis, faint data being lost here and cannot be recovered in post-processing. This can be adjusted via gain and/or brightness if the exposure=15s is to be used.

• The lights were captured at gain=0 (the dark is at gain=100). It is normal for dark capture settings to exactly match light frame settings.

light-1.JPG

light-2.JPG

darks.JPG

flats.JPG

flats-2.JPG

With this histogram, the peak can clearly be seen separated from the left hand vertical axis and there is a gap before the histogram starts. This will ensure faint data is not lost.

histogram.JPG

Dave

[oopfan]

Dave,

Bingo!

Brian

[metastable]

Hey everyone,

Okay wow! I'm learning a lot right now. I'm not sure why the gain would be different between the lights and darks. I don't adjust it between captures. I'll need to make sure next session to double check that it stays the same. Thanks for all the detailed explanations from everyone. I have a lot of things to look into on the next session to try to nail this problem down!

Much thanks,
Alejandro

[oopfan]

Alejandro,

Also something to watch for is cooler temperature. Your Lights were taken at 8 degrees Celsius but your Darks were taken at 0.1. Did you capture the Lights and Darks in the same session? It looks to me like you may have used the Darks from a library that you captured some time ago.

Brian

[oopfan]

Alejandro,

Sorry, Dave just informed that I was looking at the Focuser Temperature. Never mind.

Brian

[timh]

I had a similar effect some time ago that turned out to be due to the offset (brightness) value at which the darks were taken not being exactly the same as for the lights -- so another possibility perhaps?


Scrub that ..just noticed brightness was the same

[metastable]

Hey Everyone,

I finally have some decent weather and thanks to everyone's help it looks like I've solved the issues I was having. I've made sure that my darks, flats, and lights all have the matching settings (besides exposure for flats), and I've adjusted the offset(brightness) to detach the histogram from the left and things are looking good!

Thanks,
Alejandro

[turfpit]

Alejandro

Good to hear you have sorted out your problem. Systematic capturing of calibration frames and an understanding of the histogram are big steps in producing decent images.

Dave