White Balance

Works fine for me when choosing a pure white, but gray should also be fine, if it's a true gray (i.e. the RGB values are equal in reality)... Remember that a gray card may not be produced in color neutral fashion unless it's marketed as such... Conceptually and mathematically there should be no difference between "pure white" and "pure gray." After all, a fully pure white value of 255.255.255 is not valid for setting white balance, thus anything you can set by is really "gray."

... Also, be sure not to confuse the white setting under Exposure with the settings under White Balance. The one under Exposure is just a "tint" setting and won't have the same effect; effects of the two are quite different.

Really? I've been trying this all night and when I select a white the colors go absolutely wacko (this is for the eyedropper in white balance). I'm assuming the original poster is commenting on this also. If I could get the white balance to work, it would be a big step forward for me. As it is I'm not sure Aperture is useable for me without it.

I've seen this on Canon 5D and 1D Mark II files. What kind of file did it work for you on?

Works fine for me when choosing a pure white, but
gray should also be fine, if it's a true gray (i.e.
the RGB values are equal in reality)... Remember that
a gray card may not be produced in color neutral
fashion unless it's marketed as such... Conceptually
and mathematically there should be no difference
between "pure white" and "pure gray." After all, a
fully pure white value of 255.255.255 is not valid
for setting white balance, thus anything you can set
by is really "gray."

I'm working with 95 percent 1D Mark II and 5 percent 20D.

Since Aperture does not have an RGB dropper (aka pixel meter) to check the value, I suggest exporting one of the "whacko" photos to Photoshop and using the dropper to measure the values and see if the RGB are equal. If they are not equal, then Aperture is doing something weird. If they are equal, I can see several possibilities:

1. Could be your monitor profile, and you previously compensated for it somewhere else. (Are you using hardware calibration? Doing so would eliminate bad profile as a possbility, assuming the unit is in good working order.)

2. Could be localized lighting issue. Example: Under mixed flash and tungsten lighting, if different subject areas are getting hit at different angles by different lights, the gray card becomes largely useless, requiring a "manual" decision.

3. WhiBal gray is imperfect, and you previously compensated for it somewhere else.

The pixel meter dropper thing will be the key to figuring out where the issue is here 🙂

3. WhiBal gray is imperfect, and you previously
compensated for it somewhere else.

the whibal grey card is one of the best ways to get perfect white balance. works perfect and flawless in every raw converter, from nikon capture to adobe camera raw and capture one.

just pick up the light grey card with the eyedropper = perfect. what is aperture doing wrong?

My point isn't to impugn Whibal or not. Asserting that it's perfect or imperfect doesn't help. I'm just stating the possibilities... Like I said before, the proof will be someone who is affected measuring the results in Photoshop with the RGB pixel meter.

Thistledown,

that is a big difference! Can you take the aperture wb adjusted image, open it in PS and check the RGB values there? I'm curious to see if they are anything close to even.

If not, that's a nasty bug.

Devon

I don't have the latest Photoshop C2 so I can't convert in there, but I can in Capture One LE:

Original image: temperature 3700 tint -5 RGB 255, 252, 240
Capture One LE adjusted: temperature 2650 tint -14 RGB 252, 252, 252
Aperture adjusted: temperature 4851 tint +5 RGB 254, 254, 216

This thing is broken. This is inexcusable in a $500 product. If there is something special needed by their white balance it needs to be documented and there is nothing about it. I'm getting more and more unhappy with this product. Broken white balance and exceptionally poor noise reduction. Barely acceptable performance on a top of the line PowerBook.

Apple fix the white balance ASAP!

I don't have the latest Photoshop C2 so I can't convert in there, but I can in Capture One LE:

Original image: temperature 3700 tint -5 RGB 255, 252, 240
Capture One LE adjusted: temperature 2650 tint -14 RGB 252, 252, 252
Aperture adjusted: temperature 4851 tint +5 RGB 254, 254, 216

This white balance is broken. This is inexcusable in a $500 product. If there is something special needed by their white balance it needs to be documented and there is nothing about it. I'm getting more and more unhappy with this product. Broken white balance and exceptionally poor noise reduction. Barely acceptable performance on a top of the line PowerBook.

Apple fix the white balance ASAP!

I just ran the test and got interesting results... which seem to clearly indicate that WhiBal is the issue.

1. I imported your Original image into Aperture, and I set white balance via the white card.
2. Exported.
3. Opened in Photoshop CS. The readout is a perfect 208, 208,208 in the chosen area, though the surface of the card does vary a bit in quality (which is not Aperture's fault), so some spots are more like 204,205,207 (well within expectation I'd say).

Then I opened the Original in Photoshop without any Aperture interference:

1. Before doing anything, the middle gray card reads as 122,144,158.
2. Setting the middle gray dropper in Levels to the middle gray card, it then goes to 142,141,142.

Now here's the "shocker."

With the middle gray card perfect, I measure the White and other gray, and they read as 199, 207, 214 and 148, 148, 149 - respectively.

I think that's clear evidence that either the WhiBal white card lacks color neutrality or both grays do. They can't all be correct -- if they were, setting one would make all three correct. While each gray card exhibits up to 3 points variance across its surface, the white card shows up to a 15 point gap all over. Conversely, if I set the white level off the white card, it will make the grays appear to be untrue by up to 19 points in my measurement.

Now if I go back to Aperture:

If I set via the middle gray card, the white looks terrible -- as it must, because the WhiBal cards are not uniform.

So in my experience, I absolutely can't blame this on Aperture. Can anyone corroborate this?


G5 dual 2.0

Here, WhiBal's own chart seems to support my finding that white card is off:

http://www.whibal.com/products/whibal/productswhibalcharts.html

Notice that this shows the white card having less red and more blue, which is what I found above. I also wathced the video manual, and the proprietor seems to imply that the gray cards are higher quality or more important than the white, which is more like a just-in-case card.

G5 dual 2.0