DBE: Selecting sampling points near nebulosity

[Farzad_k]

Hi,

When the image contains a lot of nebulosity how does one go about selecting background samples that represent light pollution without winding up telling the system the colors of the nebulosity are also the background?


Thanks a lot

Farzad

[msmythers]

Farzad

My suggestion would be to look at reference images taken from sites with little to no LP. Then compare those images with yours and decide where to place your samples.



Mike

[Niall Saunders]

Hi Farzad,

Yes, Mike is basically correct. If you are working on any image where most of the data is 'foreground nebulosity', then there might be very little in the way of 'background' for the DBE Process to work with.

If you can find areas of the image that you know are indeed 'background', just pop a DBE marker into them (avoiding stars, as per normal) - then see what DBE gives you, not just as a final image, but also as the background image PI intends to use to correct the original. Quite often, a thorough investigation of the extracted background can tell you a lot about what you are actually working with,

Hope this helps.

[Geoff]

I find that in these cases ABE often works very well.

[Farzad_k]

Thanks everyone.

So I am hearing that we should indeed be careful of where we select a sample point and if we are unsure of where good data may exist we should look at similar images taken at dark sites to know where the good data exists. I can also see the good data myself, just wasn't sure how to deal with separating the background from foreground. I am assuming that light pollution related gradients are uniform across the image and selecting locations where good data doesn't exist helps also with areas that we are not sampling because of mixture of good and unwanted data.

I will give ABE a try also.


Thanks