The white balance

[Tiago Ferreira]

I always use 50/50 because it works, it's easier to make darks (it's a round value) but mostly because i once read Robin making reference to the power of the ZWO 294MC being at it's high when using the 50/50. And i trust Robin.
I don't understand the deepest technical nuances of a camera but what i interpreted has to be with the pixels and the distribution within them that works better with the 50/50 ALTHOUGH boosting the green tint and therefore, increasing those values will return a natural color on the screen.
From that, i understood Robin preference for that 50/50 as a way to use the full power and later, in the post process, adjust to more natural colors.
My dilemma is that i'm centered on EAA, therefore, live stacking is my main purpose. And i get the greenish tint BUT then i adjust the histogram to get the natural colors. So the darks and the live stack work at 50/50.
Robin, can you explain in a simple way the advantages i have keeping using the 50/50 live stacking with the ZWO 294MC or do you recommend to change the way i'm doing things with the white balance?
Thanks

[admin]

Hi,

in most cameras, the way that white balance adjustment is done is by taking the pixel values that come off the camera and then boosting or reducing them by some factor. For instance if you turned up 'blue balance' from 50 to 90, it might multiply all the blue pixel values by 1.8. If you turn green down from 50 to 20 then maybe it multiplies all green pixel values by 0.4.

My view is that using this sort of white balance adjustment risks throwing away image data in a quest to make the on screen picture look 'good'.

For instance, assume 8 bit mode for a moment, if you have turned up blue as suggested above and every blue pixel is being multiplied by 1.8, then every pixel value from 142 upwards will get turned into a 255 - maximum brightness by the adjustment. Any detail in that range (perhaps contributing to the colours of brighter stars) is lost. Similarly if you are reducing the brightness of a channel, multiplying by 0.4 means that pixel values of 70, 71 and 72 from a green pixel will all give a value of 28 after the adjustment. Again, detail is lost, and once it is lost, it cannot be recreated.

This is the reason for my often repeated suggestion to keep the colour balance settings at the 'neutral' level - 50/50 for ZWO and to get the right balance in the final image after stacking. That way, all of the data that the camera produces is fed into the stacking process, and any adjustments for colour balance can be made, tweaked, undone, redone and improved as often as you like when processing the final stacked image.

cheers,

Robin