Paul C. Buff, Inc. Technical Forum

Setup:
Camera: D810
ISO: 100
Shutter: 1/250
Aperture: f6.2
White Balance: Flash
Output: NEF RAW
Software: ViewNX2/Lightroom
Light: Einstein 640 / Color Mode
Light Modifier: Omni with diffusion sheet

I want to get some input as to what's causing a different white balance once I import the photo to ViewNX2 or Lightroom.

When I import the photo and view in ViewNX2, I'm getting a white balance of 6461. It's worst when viewing the data in Lightroom 5, where I get a 7100K white balance. This seems to be far from the 5600K setting on the Einstein 640.

In order for me to get the desired white balance, I have been using Expodisc.

Any idea possible cause and how to mitigate the problem such that?

Probably two things: Software differs in Kelvin readings, and your modifier setup the second issue.

I ended up buying a Sekonic C-500 color temp meter (They have a newer one out now too.) and found everything in modifiers I have alters the color temp. Then I discovered the software doesn't agree either. Even Nikon's Capture NX-2 doesn't report Kelvin but uses a slope of the Red and Blue lines with a 1.00 being perfect (but you can't use it on the D810 as software is old.). The 3rd party software makers don't have proprietary info on the camera RAW so they take a WAG as to what it should be. Quite maddening seeing a 1000K difference between them as well as a tint color difference, but that is normal.

I even have a couple of beauty dishes and seems even the older one has oxidized and changed color a lot over the new polished one that looks like a polished silver mirror. Then come the diffusers and they can go really blue in color and some of mine were 1,500K higher. Some fabrics have brightners/whiteners in them and cause a bluish Kelvin too. Seems more of a problem with the polyester fabrics too. Some others seemed to fade to a yellowish over a decade of use and actually became better to use than new ones.

I had issues trying to balance them against daylight and could see normal skin color on daylight sides of faces, and a blue on the flash fill side. Nothing helped, short of gelling the lamps to balance, or using the Control Points in CaptureNX-2 to warm the flash side.

What I ended up doing is dying my own rayon diffuser fabrics down to about 5,250K where Nikon seems to place their daylight Kelvin at. A very, very weak RIT gold/yellow dye solution that takes well to rayon (Polyester doesn't take dye well at all as it seems melded into the poly when made.). Trial and error, but look at older Vivitar 283/285 series flash unit's diffuser color and you get an idea as to how the color tint should look to get the cold flash warmer.

Mack

White balance is a tricky thing to define. First and foremost, 5600K is the color temperature as the light leaves the tube. There is then the color shift introduced by the diffusion dome, which is about 300K, leaving the color temperature at about 5300K. Then the modifiers can then add a shift, silver may add a little in one direction (maybe 100K or so), and white fabrics will drop temperature about 200K. This is if you have a white fabric that has not been treated with optical brightners, that "make whites whiter". These can shift the color up by 1000K to the blue.

Then you have factors within the camera. Each camera will record color a little different, and each lens can have a different shift (I have seen this even with the same model lens). If there are any filters on the lens, that can shift color as well.

I have no test data on the neutrality of an Expos Disc, but I will assume it is fine. However, you are measuring the light from the source, but not necessarily the light falling on your subject.

If there is a large window that can also contribute light and color, and may or may not be accounted for by the Expo Disc. Large surfaces of color (like walls, ceilings, carpet) can also affect the color hitting the subject.

Then you have different software programs that define color differently. I found a significant shift in defined temperature between Adobe Camera Raw and iPhoto, and neither of these agree with the color meter used. I forget the actual numbers, but the spread was pretty close to 800-1000K.

Ultimately, it matters little what the actual color temperature is (as long as each light is close to the others, which is the importance of the COLOR mode) or how the software defines it, as long as it is neutralized by the Expo Disc. And really, if all color temperature was straightforward and predictable, the need for the Expo Disc would not be present.

Thank you for the insight. It's very helpful. Just want to make sure that my Einstein does not need any repair.

Leo

Nothing sounds off. Another thing to consider is a low flash output and a high model lamp output can also cause a color shift in the images, due to the high contribution of the model lamp.

I normally set my modeling lamp same as or lower than the flash output. I do set my main light around -4f ('16).

That should not cause any problems, then.

Amen Mack,

I've been through every possible color meter, white balance card, camera and lenses, modifiers, etc.

They are all different and can vary by 1000K with the same light and position and power. Watch out for the Sekonic color meter (as well as all others I've bought) - mine can easily give a 300K+ different reading if you move it a few inches.

I use a dark neutral grey room, a WhiBal card, Adobe Bridge and Nikon D300S as my standard. In the same setup, my D7000 will yield a 400K higher reading - both with Nikon lenses.

So if I want a quick ballpark I should use:

5600 Einstein in color mode
-300 diffuser dome
+100 Omni reflector silver
-200 Diffusion sock
------
5200K

Is that correct?
And of course shoot in raw, adjust one shot and sync in Lightroom.

Finally, since the diffuser dome is standard why don't you just spec the Einstein at 5300?