A little help? (with RGB ratios for exposures)

[Tim]

After a bit of digging, it seems quite a few imagers are using a ratio of 1:1.2:1.6 for RGB exposures, and i'd like to give that a try.

I'm not very good at maths, it just doesn't make any sense to my brain, so could someone confirm that I have worked this out right please

If I wanted as a base exposure time, say 600 seconds;

Red = 600s

Green = 720s (600 x 1.2)

Blue = 960s (600 x 1.6)

Is that right?

If it is, then ideally I would want the Green exposure to be the 600 seconds, but I cant work out how to get the Red exposure time, unless you divide 600 by 1.2 ???

That would then give

Red = 500s (600 / 1.2)

Green = 600s

Blue = 800s (500 x 1.6)

Have I got this right? Sorry to sound like a right thicko, I just don't have an aptitude for numbers! And is it really that much of a difference in exposure??? A whole 5 minutes longer on blue than red? It doesn't sound right :/

Thank you

Tim

[Earl]

is not g2v calibration the best way?

[Tim]

Only sometimes apparently........I'm after a quick fix easy method

[Tim]

There's this, but look at all the numbers, it's just depressing :/

http://astrophoton.com/tips/color_balance.pdf

[steppenwolf]

Why not take a daytime image of a colourful scene using the same exposure for each channel and then produce an RGB image to determine the colour cast (if any!?!). If, say the image is too red then copy and paste the red channel to a new file in Photoshop, reduce its opacity by, say 10%, flatten it and use this new red in your original RGB composite. A little hit and miss but once you have a realistically coloured daytime image, you will know what correction factor(s) to apply. The daytime image will have been illuminated by a G2v star ........

[rfdesigner]

The 1:1.2:1.6 is for the 383 + Baader filters. the 460 will behave a little differently

Also it depends what the altitude the image was taken at, the lower the altitude the more the blue is attenuated, so the more you need to multiply it by to get back to white.

I've been trying to use B-V values and measure the Blue/Green chanels to determine what I've got and so get an accurate measure.. sadly the B-V values in Stellarium should come with a VERY big health warning as they are taken from the Nomad database, and as such that is a database of databases, with photometric values for any one star comming from more than one dataset (i.e. blue from A, visual from B ) it's a reasonable measure of Blue or Visual, but when you take the B-V difference you really need both measurements from the same set. I found I had stars that were supposed to be 'red' being bluer than some of the supposed 'blue' stars.

B-V is really the way to go but you need good reference data.

Derek

[rfdesigner]

Oh yes, another way..

You could try imaging the moon. in which case you should get 0.85B:1.00G:1.15R weightings. (yup.. the moon is slightly pink!.. just been through a calibration of a moon image)

see last graph from

http://earth.esa.int.../papers/dobber/

Derek

[Tim]

Hmmm, there's a thought, I could actually do that very easily in register Steve which allows you to weight the colours. Only problem is getting a daytime target from my OBS, I don't think I can!

[Helen]

http://www.deepskycolors.com/tools.html might this help?

Helen

[Gina]

"My bwain hurts Bwian!"

[Tim]

It might Helen Pixinsight actually does those things too, but the talk of coefficients does indeed make my bwain hurt, just as Gina suggests

Thanks all