EX1 white balance revisited

[Piotr Wozniacki]

I must say I'm still uncomfortable about the difference - sometimes striking - between the dialed-in (preset) value of color temperature, and that measured with the AWB button (against a white card). I know the theory behind this difference (one is making for green-magenta, while the other is not) - but seeing how dramatic the difference can be (with the measured WB looking much "better", i.e. without any magenta cast), I've been wondering:

- is there a way to "reset the preset", so that it's more like the true WB as measured off the white card?

Just take a look at those 2 grabs - the difference is striking, and the measured (left) is correct.

Or is my EX1 faulty? Or is the dreaded IR contamination in play somehow differently with both settings (the speakers' grills are black correctly in the left pic)...

[Alister Chapman]

There are many factors that will affect the colour of the preset white including the matrix settings, actual light source, reflections, polarisation etc. The manual white will almost always give a more accurate result as it measures the actual colour levels and compensate to produce a true white image. The preset is useful for giving a fixed setting that is consistent (even if wrong) that makes grading simpler or for situations where you can't do a manual white.
You could dial in an offset to the preset or change it altogether but there will still be many situations where it will be wrong, that's why you have the manual option.

Looks like you are using the standard matrix. Try cinema matrix to get rid of the red cast.

[Piotr Wozniacki]

Thanks for the reply, Alister. I'm aware there would be differences, but never saw them as great as in the pics above! Mind you - everything else was absolutely the same when taking them: the lighting, PP used (hence the same matrix and gamma), exposure...

As to the matrix, you're correct - I'm using Standard in this particular PP. Frankly, I like it the most (of course, when my WB is correct) - Cinema is just a bit lacking in reds, while Hisat has them over-saturated to my taste. After almost a year of experimenting, recently I started using Standard at +35.

[Daniel Epstein]

Looking at the jpegs it seems like there is some daylight coming into the room which the Auto White adjustment would compensate for while the preset would not as it is expecting a calibrated 3200 source. Very few lights actually produce 3200K exactly and mixed light situations can confuse the camera much more than your eye. Little variations in color temperature in this range show up clearly in color rendition. Also you should include a subject in the picture as you might find flesh tones look better under the preset even if the wall doesn't. Or maybe not. The point is you have a choice. On cameras with adjustable presets I like to dial in a number rather than AWB as sometimes Auto White over compensates for the mix of light. Also in the field it is important to know the monitor you are using to decide what color you are actually producing as the monitor can fool you as easily as the camera.

[Piotr Wozniacki]

Yes Daniel - good point about the mixed lighting in this particular scene probably being the main reason for such defferent results. With consistent lighting (e.g. outdoor on a sunny day), the measured 5600 K is almost exactly the same as the 5600 K preset.

[Leonard Levy]

Piotr,
I think you are stressing out over a nonexistent problem.

Your preset white balance is carefully designed camera setting meant to look white under 3200 lamps. It doesn't rely on some any camera circuitry to "read" the color of the light and make sure it is 3200.

When your white balance tells you the light is 3200 its because some wanky circuit in the camera is just trying to give you a number for comparison purposes. Its not an expensive color meter and doesn't need to be. Your expecting it to be far more precise than it is designed for.

I've watched these circuits evolve in cameras for about 20 years and they have been all over the map. For a while one of the Sony's was giving readings like 2200 where you would have expected 3200. Two cameras side by side would often give different readings.

Just get used to what your circuit is telling you.

Lenny Levy

[Piotr Wozniacki]

Originally Posted by Leonard Levy

Piotr,
I think you are stressing out over a nonexistent problem.

Lenny,

I totally agree; my original post in this thread (along with the striking grabs) should be a good indication of potential discrepancies that the novice EX users might come across.