I'm glad I know it now, but seriously annoyed at Panasonic. Custom WB is supposed to make things better and easier, not worse.

@shian - "Daylight does not render accurate colors on the GH2, whereas Tungsten does."

If I understand your testing procedure correctly, it sounds like you simulated daylight with a 3200K tungsten lamp through a CTB filter. In my experience, this combination produces only an approximation of daylight, and I wouldn't be surprised to see the kind of color shift you documented. When shooting in tungsten, I've gotten more consistent results balancing the camera to the native temperature of the light source, rather than attempting to use filters to alter the light.

Of course, the GH2 has a well-known greenish tint, regardless of how you set the color temperature. In daylight, I've found the following settings produce a better color balance than the default:

GH2 Daylight Settings

Film Mode Smooth

Contrast -2

Sharpness -2

Saturation -1

Noise Reduce -2

WB Sunlight

A-4

M-2

@LPowell if you peep the thread, you'll see that I did the test in both direct and indirect sunlight with the same results, its definitely something in the wavelength of sunlight.

Haven't done any official tests, though almost always prefer the color of the 'cinema' film mode when shooting with daylight. Maybe not as gradable as Smooth, though with contrast dialed down, I rarely find lack of DR to be an issue. Where it is, I use some diffusion between sun and subject and/or bounce some light into the shadows. Only real issue I had with cinema preset is with blue skies and gradient banding, though was only when shooting stock or moderate patches and haven't encountered it since installing the Flow Motion patch. Would be curious to see if the daylight vs tungsten discrepancy still exists when using the cinema preset or some of the others for that matter.

@shian This is great, this has been on my list of things to do, using the color checker to create a series of LUTs. Just need to get you to switch to Premiere Pro now :) I always find "near a window" to be very challenging for CC--I thought the result look solid. It's interesting that the contrast on my monitor looks just a a tad low on the squares, but it looks just fine on the actual scene. I guess my next question is basically, does one avoid any in-cam color balance at all, like the green correction many have been using, once you have a set of LUTs.

@shian, liking your work here. Have you looked at the 18% square with the RGB Histogram and compared the daylight and tungsten, I would be curious to see if you can see any difference in the colour balance, between R, G and B they should all be the same value (level) on the 18% square. You say it was exposed so 18% grey was on 50 IRE, curious to know have you tried setting the exposure on your camera with you meter and see where daylight and tungsten actually fall regards their IRE level, I would think they may be different.

Thanks lpowell, I had not thought about the image data flow. I will check a few of the settings to see if there is any noise added, but if the noise is the same I'll try to do the adjustments in the camera for the reason you describe.

@shian Wow. It doesn't look like the MBR plug-in did a good job on that frame at all! It looks way over-saturated. It's also pushed the luma and made it clip too, and as it's not a 32-bit plugin the clipped values can't be recovered afterwards either.

@PhilC they are off by about 2-3 IRE. Tungsten hits right at 50%. Daylight is right between 2-3 IRE over.

@Aria I'm not sure the difference is something to be worried about. There will never be a case where you have to intercut color temps within the same scene where the framing and subject would be identical enough for anyone to see a noticeable imbalance. It just makes getting an accurate reproduction of skin tone, and a few other colors much more difficult. My testing has not yet shown that Daylight renders unusable color, just that it is not accurate with what is considered a color standard in the video world.