18% Gray Card

[KentDub]

When you use a 18% gray card, it is supposed to be a perfect mid-tone, correct? Using spot metering and having the camera say my exposure is correct, the RGB numbers (with all sliders in ACR zerod out), are around 100. I have to move the exposure slider to 0.8 to bring them to 128. For most images that I have used using the gray card, I can only adjust the exposure slider to 0.35 before clipping begins.

So - is my camera trying to protect the highlights (Canon 5dmkII), or is 18% gray card supposed to be lower than a true midtone (RGB 128,128,128)?

[_GUI_]

Forget about getting 128 using a 18% gray card, there are too many factors involved to make the resulting value different to 128:

1. First of all a camera sensor deals with 3 colour channels, R, G and B, that get very different exposure values in the RAW data and need to be white balanced to get aligned. So you cannot expect a precise resulting aligned value for them 3
2. Second every camera, based on its particular design, allows a different amount of headroom from camera metering (your 18% gray card) to saturation. That means the same neutral RAW processing on every different camera will provide a different final value
3. Third, even if 1 and 2 were not a problem, the RAW data value produced by camera metering is not the important one to check the 18% gray point is OK, but the value that this RAW data achieves in the JPEG processing were a tonal curve has been applied rearranging captured levels. So it shouldn't be in the RAW development where you should look for the 128 value but in the camera JPEG

The good news is that actually it's not necessary to achieve a 128 in a neutral development of the captured RAW data since you can rearrange it to your needs with any level tool (curves, levels, brightness, exposure,....).

Regards.

[KentDub]

Thanks GUI for your responce. I did some further google searching and found a wonderful site that had a page specifically about what I was trying to do - set up accurate defaults for ACR (Equivilant to your point #3). I posted the the question assuming proper white ballancing (#1).

It is a very technical article, but is definatly inline with what you are saying. Specifically, assuming a correct exposure in a calibration environment, the brightness slider should be adjusted so that your 18% gray becomes 128 (127.5 actually). From there, further calibration can take place. This also provides a much more camera-specific default than the out of the box settings ACR gives you (+50 brightness, +25 contrast), and explains a lot why all of my shots were especially dark when I tried a full zerod out development. The brightness needs to be adjusted because of what you said (#2), each camera has different capabilities for capturing dynamic range. The brightness slider restores the midpoint, but compresses the highlights (rather than clip them).

For anyone who is interested in how to determin camera-specific ACR defaults with a color checker target, see this site: http://www.rags-int-inc.com/PhotoTec...crCalibration/. It is very technical and differs from the calibration method described in the book 'Skin' by Lee Varis. Reading through it, it seems to provide more in depth details and allows for more camera-specific defaults than just the calibration tab of ACR.

Thanks.

[Bill44]

When using a grey card in sunlight make sure you are reading the true card and not glare which will affect your reading.
The Whibal card I use has a black and a white section on it plus some writing. If you cant clearly see the black/white/writing then you are seeing glare and have to turn the card until you get the correct position.

[Colin Southern]

A mid tone is simply 1/2 way between the reflectance of a black and white object (2 stops down from white, 2 stops up from black) - it doesn't co-incide with an exposure that's 1/2 way through a sensors dynamic range though.

In summary, it's significant for normal reflective scenes, but not for scenes with backlighting or specular reflections.

[eNo]

And if this weren't confusing enough, Thom Hogan claims that Nikon DSLRs implement a middle gray = 12%.

[Colin Southern]

Originally Posted by eNo

And if this weren't confusing enough, Thom Hogan claims that Nikon DSLRs implement a middle gray = 12%.

The latest I heard was that manufacturers were implimenting an ISO (as in International Standards Organisation) value of 13%, but who the heck knows (I might add "who the heck cares as well"!) - I don't know the formula, but I've read that the difference between 18% and 13% is about 1/3 of a stop. The other thing to consider is that we're really only talking about the mid-point of a reflective scene - which bares no resemblence to anything I normally shoot. If we only ever shot reflective scenes and only had cameras with 4 stops of dynamic range it would be vital, but that's just not the case.