M81 & M82 Flats

[Philter]

Hi. 

After buying a Sky-Watcher Evostar 80ED DS-Pro last year I finally managed to to get some images last night, although not as many as I would have liked.  After setting up, trying to polar align and battling with software I only managed to get 39 minutes on M81 and M82 in total.  I don't know much about processing yet but this is what I got out of Astro Pixel Processor and stretched (aggressively) with curves in Photoshop:

• SW Evostar 80ED
• SW 0.85 flattener/reducer
• Nikon D5300 (unmodded)
• IDAS D2 filter
• Lights - 52 x 45s (limited due to trailing)
• Bias - 100
• Flats - 50
• No darks
• Bortle 7

I can see a few bright spots that match up with dust bunnies on my subs that I thought flats would take care of. I was wondering if if there might be something wrong with my flats or the way I applied them in APP. They were taken using an EL panel and exposing so the camera histogram peak was in the centre.

 

This is what the masterflat from APP looks like. The most prominent dark sport seems to line up with the brighter spot in the integrated image.

Thanks

 

[geordie85]

I find that uncorrected dust bunnies are dark and over corrected dust bunnies are light so it may be something to do with how APP has applied them. 

[vlaiv]

This is to be expected with bias only calibration.

Flats are supposed to correct light signal only and if you use bias files for calibration you will not calibrate out dark signal - electrons that accumulate during exposure. On cooled sensors this signal is very low, but it can be significant on regular DSLR.

You can take matching darks, but even then it's not guaranteed that you will have good calibration - you don't have set temperature control - so temperature of your lights and darks will depend on ambient temperature that can change.

[Philter]

So the flats might not be able to remove the the dust spots correctly without dark frames, is that right?

[ollypenrice]

What you're getting seems to be over-correction which can sometimes happen if you don't calibrate your flats with darks-for-flats. These are just darks taken at the same settings and temperature as your flats. Give a multi-second delay between flats and darks for flats in order to avoid heating the chip up too much.

Olly

[vlaiv]

1 hour ago, ollypenrice said:

What you're getting seems to be over-correction which can sometimes happen if you don't calibrate your flats with darks-for-flats. These are just darks taken at the same settings and temperature as your flats. Give a multi-second delay between flats and darks for flats in order to avoid heating the chip up too much.

Olly

Not sure if over correction can come from missing flat darks.

Here is rationale: calibrated = light / flat

Over correction means calibrated is higher in value (brighter) than it should be. From above simple equation, its easy to see that it can happen if light value is higher than it should be, or flat value is lower than it should be.

Only way that flat value can be lower than it should be is if one "over corrects" with darks or in case of saturation (here it would not be the case) - when values get clipped on max value.

Under correction can certainly come from missing flat darks, as in that case calibrated value is lower than it should be and that means - light value is lower than it should be (hard to happen unless you over correct with darks - darks on higher temperature than lights, or mismatch in exposure), or flat being higher in value than it should be - meaning dark current or bias not subtracted.

[ollypenrice]

26 minutes ago, vlaiv said:

Not sure if over correction can come from missing flat darks.

Here is rationale: calibrated = light / flat

Over correction means calibrated is higher in value (brighter) than it should be. From above simple equation, its easy to see that it can happen if light value is higher than it should be, or flat value is lower than it should be.

Only way that flat value can be lower than it should be is if one "over corrects" with darks or in case of saturation (here it would not be the case) - when values get clipped on max value.

Under correction can certainly come from missing flat darks, as in that case calibrated value is lower than it should be and that means - light value is lower than it should be (hard to happen unless you over correct with darks - darks on higher temperature than lights, or mismatch in exposure), or flat being higher in value than it should be - meaning dark current or bias not subtracted.

This is something I read a long time ago and took at face value. Perhaps I shouldn't have done so.

You, I and a lot of others participiated in a very long thread on over-correcting flats in 2016.

In my case it only ever affected one camera and I never got to the bottom of it. The camera was on long term loan and has now gone back, taking the problem with it!

Olly

[Philter]

Looks like I've got some reading to do!

I'm now wondering if my flats are underexposed as I used the in camera histogram to judge my exposure and I've just been reading that it displays for the 8-bit JPEG data and not the raw data.

I've attached a single flat file, is anyone able to take and look and see if it's exposed correctly? 

Thanks.

M81_FLAT_Tv12s_1600iso_20190225-22h52m19s251ms.NEF

[alacant]

Hi. The nef looks very bright. Here's a side by side linear and histogram with -FWIW what I'd expect it to be- one of my canon cr2 flat frames. HTH.