Taking flats, all new to me...

[LunarView]

7 minutes ago, carastro said:

I don't think the colour paper will matter.  

Carole 

No, it’s not the colour of the paper, that is white, it’s the colour of the light coming through the white paper...!! The paper seems to make the light a pink glow...from the flats panel..

[LunarView]

Hold on am I being stupid here, my camera is mono, so the flat will be mono, so the colour of the light should also not make a difference on a mono camera..? Should it, or am I missing something..?

But will it  be an issue on a OSC camera like my old DSLR..?

[Oddsocks]

7 minutes ago, LunarView said:

so the flat will be mono, so the colour of the light should also not make a difference on a mono camera..?

For ease of post processing with both mono and OSC/DSLR cameras the light source should be as close to white balanced as possible.

The reason being that you want the flats exposure time to be consistent across all the filters used. If the light source is predominately red when you shoot flats through each of the three RGB filters you will find the exposure times are not consistent, red would be the shortest, blue the longest. If the times were very different and depending on the camera you might then need to shoot additional darks to match each flat. The situation becomes more extreme when you move into narrow band imaging where you might find the flats exposure for OIII is ten times longer than that for Ha when using a light source that produces most of its light in the red band.

With OSC/DSLR images, calibration routines for some image processing programs calibrate the light frame with the flat without debayering first, calibration is performed on the raw data. If the colour spectrum of the light source for the flat is not white balanced, then after calibration and debayering you may be left with a colour cast in the background of the calibrated image and incorrect stellar population colours that have to be corrected in post processing. Not all calibration programs work the same way though, some apply a so called Box-Car filter to the raw Bayered colour flat to average each 2x2 Bayer group before it is used for calibration and this removes any colour imbalance, turning the flat into a mono frame. You would need to read up on the image processing program you are using for calibration to determine which methods it uses and then adapt your flats procedure if necessary to get the best out of the program.

While not a deal breaker it is much simpler to aim for minimum post processing adjustments to white balance by using a white balanced source.

 

[LunarView]

4 minutes ago, Oddsocks said:

For ease of post processing with both mono and OSC/DSLR cameras the light source should be as close to white balanced as possible.

The reason being that you want the flats exposure time to be consistent across all the filters used. If the light source is predominately red when you shoot flats through each of the three RGB filters you will find the exposure times are not consistent, red would be the shortest, blue the longest. If the times were very different and depending on the camera you might then need to shoot additional darks to match each flat. The situation becomes more extreme when you move into narrow band imaging where you might find the flats exposure for OIII is ten times longer than that for Ha when using a light source that produces most of its light in the red band.

With OSC/DSLR images, calibration routines for some image processing programs calibrate the light frame with the flat without debayering first, calibration is performed on the raw data. If the colour spectrum of the light source for the flat is not white balanced, then after calibration and debayering you may be left with a colour cast in the background of the calibrated image and incorrect stellar population colours that have to be corrected in post processing. Not all calibration programs work the same way though, some apply a so called Box-Car filter to the raw Bayered colour flat to average each 2x2 Bayer group before it is used for calibration and this removes any colour imbalance, turning the flat into a mono frame. You would need to read up on the image processing program you are using for calibration to determine which methods it uses and then adapt your flats procedure if necessary to get the best out of the program.

While not a deal breaker it is much simpler to aim for minimum post processing adjustments to white balance by using a white balanced source.

 

Many thanks for the explanation...

I have just downloaded the trial of APP so am going to give that a go, not sure on the way it works with stacking though....yet.. 

[symmetal]

With a mono camera having a colour tinge on the light source doesn't really matter. It just means a bit more faffing around. If you want similar exposure times for each colour you'll have to add/remove sheets of paper to get the exposure histogram peak near the middle depending on what colour flat you're taking. In your case keep your acrylic sheet in place and take a blue flat, adding some extra paper sheets if necessary to get the 2s  or so exposure with the histogram peak near the middle. For your green flat you may need to add another sheet or two of paper while the red flat may need even more sheets to maintain the histogram peak near the middle.

Alan

[symmetal]

Just to add, you don't need to be too critical in adding sheets of paper to get the histogram peak near the middle. You're only aiming to keep the edges of the histogram 'bump' for each colour away from the left and right ends of the histogram display where it will get clipped and lead to calibration errors. If your 'bump' is quite narrow meaning there is little vignetting you could have the 'bump' peak anywhere between 25% to 75% or so histogram range and it would be fine. When calibrating using flats only the difference values between the darkest and brightest parts of the flat are used and not the absolute values.

Alan

[LunarView]

23 minutes ago, symmetal said:

Just to add, you don't need to be too critical in adding sheets of paper to get the histogram peak near the middle. You're only aiming to keep the edges of the histogram 'bump' for each colour away from the left and right ends of the histogram display where it will get clipped and lead to calibration errors. If your 'bump' is quite narrow meaning there is little vignetting you could have the 'bump' peak anywhere between 25% to 75% or so histogram range and it would be fine. When calibrating using flats only the difference values between the darkest and brightest parts of the flat are used and not the absolute values.

Alan

Many thanks for your help, makes abit more sense to me now... ??

[Pompey Monkey]

23 hours ago, carastro said:

I don't think the colour paper will matter.  

Carole 

Certainly not enough to notice. I've taken flats using a black tshirt over the scope before and they came out fine ?

[carastro]

Quote

 If you want similar exposure times for each colour you'll have to add/remove sheets of paper to get the exposure histogram peak near the middle depending on what colour flat you're taking. 

Generally I find it easier to adjust the length of the exposure than faff around with trying out different numbers of sheets of paper.  Narrowband will take much longer exposures though, so might be worth removing some sheets of paper for those.   

Carole