The Km and uniWB

[Duplo]

Now first off I guess this thread might be located in the wrong place, but as it is directly related to the K-m and its white balance presets, well I guess thsi is the place.

Some may argue that uniWb is post processing, but in reality it is about the road to an in camera histogram reflecting the actual RAW capture.
I take for granted that most of the RAW shooters out here understands the uniWB concept and why a linear tone curve is important to RAW shooters in general

A lengthier discussion on that subject will likely belong in a totally different forum anyway.

Whitebalance on the Km compared to the sensors native WB (uniWB):
Now first off, all of the presets in the K-m shows severe pushing of either the red or blue channel compared to the sensors channel sensitivity.

The red channel is generally pushed closed to one full stop regardless of what WB preset used only exception is the tungsten preset where the red channel actually is not pushed at all.
The blue channel though is pushed severely when the tungsten preset is used and most true to the sensor sensitivity using the Shade preset.


The table below shows, the how the individual cahnnels are pushed compared to the native channel sensitivity and the kelvin value assosiated:

Preset:		R		G	B		Kelvin
----------------------------------------------------------------------------------
Daylight	1.738281	1.00	1,363281	5400 (5)
Shade		2.027344	1.00	1.027344	7900 (4)
Cloudy		1.996094	1.00	1.187500	6550 (9)
Daylight FL	2.183594	1.00	1.250000	6850 (17)
Day/Wht FL	1.824219	1.00	1.417969	5400 (10)
White FL	1.644531	1.00	1.714844	4250 (9)
Tungsten	1.000000	1.00	2.835938	1800 (-5)
Flash		2.007812	1.00	1.226562	6350 (11)

I have looked at how much they can be tweaked too.
Using the flourscent white light preset with a G7 adjustment gives the following.

Preset:		R		G	B		Kelvin
----------------------------------------------------------------------------------
White FL (G7)	1.375000	1.00	1.382812	4400 (-10)

And that is the closest you will be able to get to uniWB with adjustments to one of the presets on the K-m.
But be aware that red and blue channel even with this preset are pushed a fair bit.
Now if the K-m was able to take a preset Kelvin temperature with an adjustment we would have an easy way to get there, but it is not so.
I will get back once I have figured the uniWB values out

Tone curves on the K-m:
I have had a quick look at some of the picture styles too and what pentax calls neutral is definitely not linear, so even neutral has a fair bit of intepretation going on, i am far from done mapping out the tonecurves and what tweaks that will get me closest to linear, but I am out of time for tonight.

Now all of the above may or may not be the same for the K10D and K200D, as they use the same sensor, but I have no way of knowing it for sure.

Hopefully the K-7 will be both easier to adjust and reflecting the RAW information a lot more than the K-m can be tweaked to.

Was it usefull?

[deejjjaaaa]

one related article (actually 2), might be interesting

White Balance in Digital Cameras: Problems | LibRaw

http://www.libraw.org/articles/magen...n-digicam.html