[MNCurt-K10D]

From a couple of posts here and there, I've gotten the impression that if I shoot an image of a standard 18% Gray card the resulting histogram should peak at the center if everything is working properly. Can anyone confirm this to be the case, in theory? I'm talking about standard, open-aperture metering with a DA or FA lens. I did a quick-and-dirty check with my FA 50/1.4 and FA 100/3.5, and while the 50/1.4 seemed to deliver centered histograms over a wide range of Av settings, the 100/3.5 seemed to wander around a bit more. Before going any further, I wanted to make sure that in fact a centered histogram is the corrected 'expected' result from a gray card image. I also assume that the metering mode (spot, center weighted or matrix) should not matter if the gray card completely fills the image - again, am I off base on that? Thanks for any input or guidance!

Curt

[SpecialK]

If you shoot a single color (doesn't matter which one) your histogram should be a narrow spike, not a bell curve. If it is meduim gray, it should fall about the center of the chart.

Some people say the "digital" meter is calibrated to 12% or some other number, not 18, but 18% gray has been the standard in film days and probably still is. It is the same exposure as blue sky pointed away from the sun.

[jeffkrol]

From a couple of posts here and there, I've gotten the impression that if I shoot an image of a standard 18% Gray card the resulting histogram should peak at the center if everything is working properly. Can anyone confirm this to be the case, in theory? I'm talking about standard, open-aperture metering with a DA or FA lens. I did a quick-and-dirty check with my FA 50/1.4 and FA 100/3.5, and while the 50/1.4 seemed to deliver centered histograms over a wide range of Av settings, the 100/3.5 seemed to wander around a bit more. Before going any further, I wanted to make sure that in fact a centered histogram is the corrected 'expected' result from a gray card image. I also assume that the metering mode (spot, center weighted or matrix) should not matter if the gray card completely fills the image - again, am I off base on that? Thanks for any input or guidance!

Curt

The peak should be at 110 on the luminous histogram, left of center.
Digital Imaging: Essential Skills - Google Book Search
See p 66. I get tired posting all the same links re: meter calibration.
Re: Using a 113 gray standard in digital photo
Or these:...............
A middle visual gray (L* = 50) has these equivalents:
Reflectance = 18.419%
Density = 0.73474
Pixel value for gamma 2.2 (e.g. Adobe RGB) = 118
Pixel value for gamma 1.8 (e.g. ProPhoto) = 100

A true 18% gray card (18% reflectance) has these equivalents:
L* = 49.496
Density = 0.74473
Pixel value for gamma 2.2 (e.g. Adobe RGB) = 117
....................................
118..110 oh well close enough... it's not 127 (which is the center.... 40 "points" to a stop in the center of the histogram (roughly), edges get more logrithmic so it doesn't work the same way.)
Pixel value for gamma 1.8 (e.g. ProPhoto) = 98

[MNCurt-K10D]

The peak should be at 110 on the luminous histogram, left of center.
Digital Imaging: Essential Skills - Google Book Search
See p 66. I get tired posting all the same links re: meter calibration.
Re: Using a 113 gray standard in digital photo
Or these:...............
A middle visual gray (L* = 50) has these equivalents:
Reflectance = 18.419%
Density = 0.73474
Pixel value for gamma 2.2 (e.g. Adobe RGB) = 118
Pixel value for gamma 1.8 (e.g. ProPhoto) = 100

A true 18% gray card (18% reflectance) has these equivalents:
L* = 49.496
Density = 0.74473
Pixel value for gamma 2.2 (e.g. Adobe RGB) = 117
....................................
118..100 oh well close enough... it's not 127 (which is the center.... 40 "points" to a stop in the center of the histogram (roughly), edges get more logrithmic so it doesn't work the same way.)
Pixel value for gamma 1.8 (e.g. ProPhoto) = 98

Thanks! I'm going to have to think about this a little before I can fully absorb all this good info... but for a rough assessment, does this say the 'right' spot is about 1/4 stop below the center?

Curt

[Tom M]

If you're shooting in RAW, you only need to get close.

[stewart_photo]

From a couple of posts here and there, I've gotten the impression that if I shoot an image of a standard 18% Gray card the resulting histogram should peak at the center if everything is working properly. ...(snip)... seemed to wander around a bit more. ...(snip)

Not necessarily. The internal software algorithms behind meters in modern cameras are smarter (digital logic) than older meters which simply measured a single point of 18% grey alone. Instead, these algorithms combine multiple metering points, WB calculations, lens info, and so on, and then interpret the overall scene using calculations with 12-18% grey as a base. Since these interpretations may not always be same, even with indentical scenes, from image to image (each image is interpreted uniquely), there may be some slight differences (the wandering you describe) in exposure and in the resulting histograms. However, other than Raw, those differences are often attenuated by later in-camera image processing, resulting most often in very acceptable exposure in the final image.

stewart

[jeffkrol]

Thanks! I'm going to have to think about this a little before I can fully absorb all this good info... but for a rough assessment, does this say the 'right' spot is about 1/4 stop below the center?

Curt

Another interesting quote (and more visuals w/ the link)
I later found out that my original assumption regarding camera meters was incorrect. It turns out that camera meters are not actually calibrated to make whatever they see medium gray. Instead, they are calibrated, according to ANSI standards, to make whatever they see slightly darker than medium gray by about half a stop. Ansel Adams supposedly argued to have this changed but was unsuccessful. To read more about the confusion regarding the calibration of meters, click here.

So, it turns out that if you expect your camera to make whatever it sees medium gray, your images will be darker than you expect. A possible reason that other D-SLRs don't seem to exhibit the same characteristic may be because manufacturers compensate for the half stop discrepancy by calibrating their sensors to be half a stop more sensitive than indicated. Fortunately, Nikon's engineers designed all their D-SLR cameras to be flexible enough to be tailored for individual preferences. For those of us who have a problem with the ANSI standard, we can either use a permanent exposure compensation of +0.33 EV or +0.5 EV or we can use a custom tone curve. As we shall see later, there are several advantages to using custom tone curves over exposure compensation.
.......unquote
Fotogenetic - D100 and D70 Custom Tone Curves
nikon and Pentax are pretty much the same here.... The "fudge the iso to make the images look bright" refers to Canon.
You will find all that here in a few articles.....
Articles by Doug Kerr

[jeffkrol]

Not necessarily. The internal software algorithms behind meters in modern cameras are smarter (digital logic) than older meters which simply measured a single point of 18% grey alone. Instead, these algorithms combine multiple metering points, WB calculations, lens info, and so on, and then interpret the overall scene using calculations with 12-18% grey as a base. Since these interpretations may not always be same, even with indentical scenes, from image to image (each image is interpreted uniquely), there may be some slight differences (the wandering you describe) in exposure and in the resulting histograms. However, other than Raw, those differences are often attenuated by later in-camera image processing, resulting most often in very acceptable exposure in the final image.

stewart

Center-weighted, spot or even matrix would make little difference w/ a monotone wall/ greycard but yes, metering method will affect "real images". But white balance/lens speed/mechanical tolerances/lens contrast ect also will.

[MNCurt-K10D]

Again, thanks to all for the responses... now I have much to contemplate as tonight's temperature plunges well below zero!

[jeffkrol]

........as tonight's temperature plunges well below zero!

You and me both..........

[Canada_Rockies]

You and me both..........

And blowing snow with the cold here. I almost froze a thumb with the snowblower this am. 40 below (either scale) with wind chill.