Colour Temperature / White Balance for imaging NLC

[johnm]

Most digital cameras allow you to set the Colour Temperature / White Balance (CT/WB) manualy to a specific temperature in place of the normal Automatic White Balance (AWB) where the camera computes the colour temperature by analysing the image prior to saving it as a JPEG.

When imaging NLC it is not clear what the AWB algorithm will compute the colour temperature as and thus how it will balance the red & blue light.

Do people imaging NLC set the colour temperature manually and if so what colour temperature do they use ?

I can see an argument for setting the temperature high as the scene will be 'blue' but this will emphasise and red light and cut down the blue dynamic range.

I can also see an argument for using a low colour temperature as this will reduce the red response and reduce light pollution and and remaining sunset glow.

The other option is at least with Canon Cameras is to record RAW images and then apply the colour temperature correction when the images are processed. This of course means a lot larger image size but also has a hiher dynamic range than the 8 bit limit of jpeg images. Important if you want to apply tools like contrast enhancement.

I will be interested to hear your experiances.

John Murrell

[noctal]

Good question. Out of habit more than deliberate thought I have been using "Daylight". After all, the clouds are lit by sunlight. Have you thought about it more?

Regards,
Alister.

[johnm]

Alister et al

When (if??) I get to image some NLC I will experiment with the different settings !

I suppose the real answer is to shoot in RAW format and then the colour balance can be set when the images are examined but that takes up a lot more storage, makes the download from the camera even slower though now it runs while I am asleep that does not matter.

I will have a look at the exif data from some of the NLC images I took in the past - the EXIF data should show the temperature the camera decided to use. Another use for EXIFTool that I have mentioned in another thread is to extract the data from all the NLC images in a folder and write them to a text file. That makes analysing easy when you have mastered EXIFTool.

John

[Bill W]

Hi,
An interesting thought. If you look at the gazillions of digital NLC shots we now have they (to me) vary from a bluish colour at high elevations to distinctly yellow and low elevations.
The causes being scattering and absorption of particular wavelengths as the light comes to us via differing path lengths in the intervening atmosphere.
I have always set my camera to daylight but of course this could offer the chance of some experimentation as you suggest John.
In the excellent book on Colour and Light in the Outdoors by Minnaert(if you don't have a copy go and buy one!) he describes a remarkably simple instrument called a nigrometer, basically a cardboard tube.
Might be interesting to take different sections of a display and examine them in isolation (like looking at them through said tube) and seeing the differences required to bring the various parts to the same balance.
cheers,
Bill.

[johnm]

BillW et al,

Having had NLC visible for an hour yesterday morning I did a bit of experimenting with White Balance in my images.

The best results from the standard settings is to use 'Daylight' (5700K) on my Canon 70D as BillW has suggested

If you take RAW images you can adjust the white balance in your image processing software that way you can see the changes for each of the pre-sets as well as incrementing the temperature in 100K steps to see if that improves the image. I normally use Canon Digital Photo Professional as this has the same colour temperature adjustment as you get in the camera though you can 'tweak' the image colour as well.

I have updated the NLC imaging page on my website to reflect this information.

John