Proper flat frames with DSLR?

[Larry 1969]

Hello all!
I recently got one of those led light boards to start taking flats with.
The first time I used it I didn't have camera connected to my computer so there was no way to monitor them. They turned out too bright with an exposure time of 1/4000...
This time I have my camera connected and APT running a flat plan. I dimmed the light board quite a bit and I'm using the white T-shirt method and my exposure time is now 1/125...
Attached is a screen shot of APT with the histogram. This histogram is similar to my light frames from last night. Are these good flats?

Thanks!

Larry

[UlteriorModem]

Looks good to me. If you want put one in dropbox or something so we can have a good close look.

But at first glance it looks okay

[Larry 1969]

Thanks Tom!
I certainly can upload one but what exactly would you be looking for?
I thought you would judge the frame by the exposure time and histogram.

Larry

[UlteriorModem]

Well we can look to see if any dust motes, vigenetting, etc. that should be in the image. Also inspect it for shutter shadows, light leaks, that sort of thing.

Cant really see that sort of thing with a snapshot of a screen display

[Stuart]

I can pretty much guarantee that at 1/4000 and 1/125 of a second your flats are too dark. Don't measure your flats with the Backyard EOS histogram. Get another program that can measure the ADU of the flat--you're looking for 40% of your max ADU.

[Juno16]

AP Lab works great to measure flats ADU.

[Larry 1969]

Thanks guys! I'm not sure what ADU is but...
Here's a dropbox link to one of them.
https://www.dropbox.com/s/loydga9f53xsv ... F.CR2?dl=0

I can easily delete these and take more once I know what I'm doing.

Larry

[Larry 1969]

AP Lab works great to measure flats ADU.

Thanks!
A search for AP lab didn't yield any useful results. Can you give me a link?

Larry

[Juno16]

I’m out shopping with my wife, but try this.

http://lars-frogner.github.io/Astrophot ... index.html

[Larry 1969]

I’m out shopping with my wife, but try this.

http://lars-frogner.github.io/Astrophot ... index.html

Just got back from shopping with my wife.
Thanks Jim!

Larry

[JayTee]

Hi Larry,

Also follow Steve333's overview of AP lab. Here's the link.
viewtopic.php?f=22&t=6284

Cheers,
JT

[Stuart]

ADU is Analog to Digital Units.
When a photon hits a photosite, depending on your quantum efficiency an electron will be produced and will be placed in the photosite. The number of electrons that each photosite can hold is called the Full Well Depth.

Once the shutter closes, and the well dumps its electrons, it gets converted into digital signal by an analog to digital converter. The output of the converter will be a number, called the ADU. Each camera has a bit depth by which the maximum ADU is expressed and because of how they are manufactured will be in some factorial of 2. So a regular CCD camera has a bit depth of 16 which is 2^16 with a max of 65,536. DSLRs generally have a bit depth of 14 which is 2^14 to a bit depth of 16,384. Ultimately it doesn't really matter as programs will rescale to whatever bit depth you want.

Flats should be taken so that about half the well depth is full of electrons. So assuming a 16 bit scale, with a max of 65,536, you should aim for a mean flat ADU of about 20,000-35,000. Anywhere in there is fine.

Many programs will measure your ADU on the fly--SGP does this and will tell you immediately without the need to export the file to another program. Backyard EOS (to my knowledge) does not and instead people rely on the histogram. For whatever reason this ALWAYS leads to flats that are way too dark.

Once you figure out how to measure a flat I recommend that you measure the mean ADU of a bias frame, and the mean ADU of a saturated frame. Your goal is about half way in between.

[Larry 1969]

ADU is Analog to Digital Units.
When a photon hits a photosite, depending on your quantum efficiency an electron will be produced and will be placed in the photosite. The number of electrons that each photosite can hold is called the Full Well Depth.

Once the shutter closes, and the well dumps its electrons, it gets converted into digital signal by an analog to digital converter. The output of the converter will be a number, called the ADU. Each camera has a bit depth by which the maximum ADU is expressed and because of how they are manufactured will be in some factorial of 2. So a regular CCD camera has a bit depth of 16 which is 2^16 with a max of 65,536. DSLRs generally have a bit depth of 14 which is 2^14 to a bit depth of 16,384. Ultimately it doesn't really matter as programs will rescale to whatever bit depth you want.

Flats should be taken so that about half the well depth is full of electrons. So assuming a 16 bit scale, with a max of 65,536, you should aim for a mean flat ADU of about 20,000-35,000. Anywhere in there is fine.

Many programs will measure your ADU on the fly--SGP does this and will tell you immediately without the need to export the file to another program. Backyard EOS (to my knowledge) does not and instead people rely on the histogram. For whatever reason this ALWAYS leads to flats that are way too dark.

Once you figure out how to measure a flat I recommend that you measure the mean ADU of a bias frame, and the mean ADU of a saturated frame. Your goal is about half way in between.

Thanks Stuart!
That's a lot to process. I'll get on it!

Larry

[JayTee]

I use an EL panel for my flats. I processed 3 of my flat frames (with 3 different exposure times 1/30, 1/15, 1/8) through AP Lab. It gave me the ADU level for each of these flat frames. All three did not measure up to half of my ADU full well depth. But I could see how much more signal each longer exposure gave. It was then easy to extrapolate a new exposure time which, as it turned out, put me right in the ballpark for the correct ADU reading for my flats (1/2 second). So as long as I use this camera through this telescope using my EL panel at the same illumination level I know the correct exposure time to use for my flats.

Cheers,
JT

[Stuart]

Hey,
I extracted the lightness and measured the mean ADU. It's 5324. Way too dark. It should be 25,000-30,000

[Larry 1969]

Thanks!
I'll work on it tomorrow.
I shot M44 and M86 last night and I had a tough time processing them. I suspected the flat frames. I'm shooting them again tonight and will try to get the flats sorted and reprocess it all.

Larry