Exposure calibration - How to test if metering is correct

Photos seem slightly dark on new 5D mk3. Probably just me not being used to it, but I would like to do some kind of simple test to see if metering is performing "correctly". I imagine some standard items (sheet of paper?) would give a specific histogram that I could compare with?

Thanks!

In order to calibrate a meter, you need another meter whose calibration is traceable to a known standard.

You should completely forget the histogram for "calibration", as it does not show anything about the light levels coming from the elements in the scene that you are photographing. It merely shows the distribution of different brightness levels in the recorded scene.

You could shoot some photos of ordinary outdoor scenes and post them here and we who have calibrated monitors can tell you if they look proper or whether or not you might want to use exposure compensation in your camera.

Thanks SkipD:

Here are 2 random images. exif should be preserved. No exposure compensation used. These are from RAW using DPP without me doing any adjustments:

In my experience, shooting when overcast will almost always result in under exposure. Many times I end up dialing in +1 and sometimes as much as +2 EC.

Martin Dixon wrote in post #14874872
TThese are from RAW using DPP without me doing any adjustments:

Those exposures are off by several stops from the old standby "Sunny 16" guideline.

I think you may be being fooled by DPP's actions. Try shooting in .JPG mode and see how things look then. After that, we can tackle the RAW conversions.

They are a few stops off, however as stated above the overcast day is going to provide a bit softer light and more difused.

The meter on the camera is never going to be 100%.

If you really want to know where the cameras meter stands, you can get a gray card (18% gray), light it evenly and set up the camera so it is only metering off of the gray card (spot metering). Make sure your WB is right and then take a picture with the meter reading centered for the gray card. Open it in PS and check the values. The values for the gray card should be close to 128/128/128.
Granted you can't calibrate the cameras meter this way, but it will tell you if the meter is going to slightly under or slightly over expose the scene. Of course, outside it is a moot point if you can't recognize the tonal values of the scene and understand how the camera would meter them..

I brought it into LR4 and it looks better pushed 1 - 1 1/3 stops but I have no idea how dark the overcast was that day or exactly what tone the sand should be. +2 looks way overexposed to me so I'd say calling it underexposed by "several stops" is a bit of an exaggeration. Your exif didn't show the metering mode. I'm guessing evaluative. The light colored sand is going to make the camera underexpose just like shooting someone with a white wall behind them will.

Slow your roll. Add 1/2 stop exposure when metering with an 18% gray card. The instructions on the back of the card should say this but most people don't bother to read instructions. The meter is calibrated for about 12.5% gray not 18%. The story (I don't know if it is true or not) is the cards are 18% because Ansel Adams wanted it that way for the zone system. Kodak wanted 12% but Ansel Adams won the argument. Seriously, this is not a joke and it's a pain in the @$$.

Scatterbrained wrote in post #14875411
If you really want to know where the cameras meter stands, you can get a gray card (18% gray), light it evenly and set up the camera so it is only metering off of the gray card (spot metering). Make sure your WB is right and then take a picture with the meter reading centered for the gray card. Open it in PS and check the values. The values for the gray card should be close to 128/128/128.
Granted you can't calibrate the cameras meter this way, but it will tell you if the meter is going to slightly under or slightly over expose the scene. Of course, outside it is a moot point if you can't recognize the tonal values of the scene and understand how the camera would meter them..

An 18% reflectance card is actually about 12.5% (it's a long story). However, you can shoot a blank sheet of paper, a gray card, a white balance card, or just about any other solid featureless neutral object and all should result in the image having the same brightness of about middle gray. Canon typically has their nominal exposure set about a third stop below this point for better contrast and saturation of in-camera JPG images.

Martin Dixon wrote in post #14874872
... Both taken on overcast day at the beach. How "accurately" are they exposed?

The accuracy of the exposure for the second image is exactly as expected. The lighting is very flat and the luminance values of the sky, water and beach all are clustered around middle gray as shown in the luminance histogram below. Almost everything falls between the two blue lines marked on the histogram. If the image were underexposed, the cluster of data would have been shifted to the left. In order to have a histogrfam with a broader distribution of data (i.e., more dynamic range) would require lighting that creates shadows. Otherwise, this is to be expected.

Boss302 wrote in post #14875719
Slow your roll. Add 1/2 stop exposure when metering with an 18% gray card. The instructions on the back of the card should say this but most people don't bother to read instructions. The meter is calibrated for about 12.5% gray not 18%. The story (I don't know if it is true or not) is the cards are 18% because Ansel Adams wanted it that way for the zone system. Kodak wanted 12% but Ansel Adams won the argument. Seriously, this is not a joke and it's a pain in the @$$.

I'm aware of that yet it has no bearing on the technique. It's not my technique, but one that I've used. Although it's designed to calibrated a light meter to a camera. You can see the technique explained here:http://www.frankdoorho​f.com …t-meter-some-quick-notes/

and a quote: