Non linear data image

[Francy75]

Hi all,

I was wondering if exists some technique or tool to convert a non linear image into a linear one; my digital camera takes pictures that include white balance and they are extremely bright becuase of light pollution and so my attempts to process my image using StarTools are not producing nice results. Does exist any turn around about non linear images issue?

Thanks in advance for any help!
Francesco

[admin]

Hi Francesco,

Can you tell us the model of your camera?
StarTools can reverse some of the stretching in an attempt to make the data linear again - when opening your data choose 'Modified and not linear' and then 'Attempt to reverse stretch, activate tracking'.
Often though, the stretch is not usually the biggest problem, it's the file format that is used; JPEG destroys a lot of faint data that is important in astrophotography. Many (all) digital cameras further apply sharpening, noise reduction and and other 'enhancements' that further destroy detail.

[Francy75]

Hi Ivo and thanks a lot for you reply!

I've an olympus pen e-pl5 and my camera can take pictures with raw file format; in the following link you can find the picture I took (it's a single picture of M-31 galaxy using 8 minutes exposure time with sharping and noise reduction applied by the camera):

http://www.francesco-puccini.it/temp/M-31.jpg (it's a jpg file just to upload it in a more web friendly format but I always use raw images for processing my images)

while in the following link you can find my attempt to convert it into a linear image using dcraw (dcraw -w -q 0 -t 0 -o 0 -4 -T):

http://www.francesco-puccini.it/temp/M-31_dcraw.jpg

Do you still advice me to use 'Modified and not linear' as importing option? Does StarTools work well with long time exposures that produce very bright pictures (at least for me because I live in a area with much light pollution) or do you think StarTools can work better with multiple pictures having shorter exposures?

Thanks a lot
Francesco

[admin]

Hi Francesco,

Yes, StarTools should definitely be able to work with this.
And indeed, stacking many pictures is the way to go!
The only trouble with long exposure times is that you'll start blowing out (e.g. white clipping) very bright features such as stars, galaxy cores and some brighter nebulae.
It's a bit trial and error.
How long was your exposure time this time?
If you can do 120 seconds that'll be great, but light pollution may cause prohibit that...

[Francy75]

Hi Ivo, thanks a lot for your very useful informations.

I took this picture with a 8 minutes exposure time and the galaxy was quite near the horizon, so definitely not the best conditions ...but I think 2 minutes in an higher position in the sky should not cause much light pollution problems in my case.

I would have just other 2 questions for you .... the dcraw conversion to a linear image gives me the option to import the image as "Linear, was Bayered, is not white balanced" or do you still advice me to choose "Modified and not linear" and "Attempt to reverse the stretch, Active tracking" option?

I honestly don't have much experience with multiple pictures and so my doubt is how many pictures should I take to make the same details of one single long exposure one; taking the picture I showed to you and reference (8 minutes, ISO 800), how many "2 minutes pictures" should I take to catch the same details and should I increase the ISO?

Thanks again!!!
Francesco

[admin]

Sorry for the late response.

I would have just other 2 questions for you .... the dcraw conversion to a linear image gives me the option to import the image as "Linear, was Bayered, is not white balanced" or do you still advice me to choose "Modified and not linear" and "Attempt to reverse the stretch, Active tracking" option?

No, if your image is linear, use one of the two linear options. Also if you use dcraw, add the following options to keep the data uncolorbalanced (better!)

Code: Select all

-r 1 1 1 1

I honestly don't have much experience with multiple pictures and so my doubt is how many pictures should I take to make the same details of one single long exposure one; taking the picture I showed to you and reference (8 minutes, ISO 800), how many "2 minutes pictures" should I take to catch the same details and should I increase the ISO?

Roughly you'd have to take 8/2=4 pictures, but in practice a little more to compensate for the fixed read noise which is now multiplied by 4x as well.
If you can take 3 minutes that's even better. Depending on your camera, ISO 800 - 1600 will be about right.

Hope this helps!

[Francy75]

No worries for the delay .... and thanks a lot for the informations, they are very useful!!!!

Francesco