Understanding your in-camera meter 101

[sk66]

Originally Posted by Doug Pardee

Huh. The article states that "light meters are color-blind." Nikon's metering systems have been full-color for some time, and Canon started switching to 2-color metering systems a year or so ago.

The description of Evaluative metering (I wonder why they chose the Canon term) is a bit on the generic side. In the case of Canon, Evaluative metering consists of a spot or partial metering (depending on the camera) at the locked AF point (or one of the locked AF points, or the center of the frame in manual focus), followed by an automatic exposure compensation that's calculated from a weighted analysis of the data from the entire metering matrix. What really matters though, is this: Evaluative metering is unpredictable. If you're trying to control your exposure, don't use Evaluative metering because it'll override your efforts. If you don't have time to calculate your exposure, let Evaluative metering make its best guess.

Yup.
I really wish sites would put "publishing dates" on their articles more....

[autofocus]

I really can't comment on this because quite honestly I'm not sure. I know my Nikon will track and focus on a moving subject using color. I'd have to believe, though, that all the meter cares about is the various tones in an image as it's trying to get to 18% grey, from lightest to darkest, and that it doesn't need to consider color while assessing the data. I also believe that sensors see in black and white, and have RGB filters in front of them to create the colors, usually with twice as many greens as the other two colors. But, like everything else, always the chance of interpreting bad data on my part??

[sk66]

essentially a sensor sees in B/W....it only sees light strength and has no ability to determine the wavelength. The RGB filter tells the camera what wavelength it is (with bias).

Color metering starts as an 18% grey metering but then takes the color information, and balance of color, to fine tune. (most reds/blues will bias towards black if recorded in B/W)

With strictly 18% grey metering cameras used to tend to overexpose (bringing reds/blues "up" towards 18%), with color metering cameras seem to tend towards underexposure more, but are overall more accurate.

At least thats my understanding/experience.

[Doug Pardee]

Originally Posted by autofocus

I also believe that sensors see in black and white, and have RGB filters in front of them to create the colors, usually with twice as many greens as the other two colors.

That's the imaging sensor. DSLRs have a separate metering sensor, and that's the one we're talking about.

By the way, here's what Canon has to say about its 2-color iFCL metering:

Electronic sensors in general are more sensitive to red light. This means when photographing subjects with lots of red in them – skin tones for example – the sensor receives a stronger signal as it only detects brightness levels. This can lead to the wrong assumption that there is more light than there really is.

The Dual-layer system overcomes this by having one layer sensitive to red/green light and one layer sensitive to blue/green light. Both these layers measure the light in their respective spectra and the metering algorithm then combines the two to provide an accurate light reading. In this way, accurate exposures can be attained in a wide range of shooting situations and irrespective of the colour of the subject being metered.

I can't say that I'm personally satisfied with the details in that description. Canon's older monochrome metering sensors use Silicon Blue Cells. SBCs are called that because there's a blue filter in front of them to compensate for the bias toward red. If the description above from Canon were correct, the extra red being picked up by the metering sensor would cause it to underexpose the image. However, many of us with older Canon DSLRs find that they tend to overexpose on reds.

[Doug Pardee]

Originally Posted by autofocus

it's trying to get to 18% grey

This is an Old Wives' Tale, or more correctly an Old Photographers' Tale. The 18% figure isn't correct. The full details on this get to be a big technical mess, and if you really care you can start with this article which made the first big "splash" about the 18% figure being bogus, and then this paper (PDF) which discusses what's really going on.

Then there are some big complicating factors on top of all that.

First, most metering these days is done with a multi-zone metering system, whether it be called Evaluative, Matrix, Pattern, or whatever. These metering systems do not aim for any particular brightness. They aim for a figure that gives exposure guidance which, if followed, will produce whatever the manufacturer thinks the customer will think is a well-exposed picture.

Second, the "18%" is meaningless as applied to metering alone. The "18%" is only meaningful when we're talking about exposure settings. Basically when we're talking about auto-exposure. Furthermore, they're only meaningful for camera-produced images, not Raw — there is no brightness percentage in Raw, which is just a bunch of sensor data.

Third, the metering results depend on the ISO value, or more specifically the Exposure Index. Virtually all camera manufacturers today use the "Recommended Exposure Index" technique for assigning ISO values, and that technique can be summarized as "whatever the manufacturer thinks the customer will think is a well-exposed picture." And customers think that a picture that averages out at 18% gray is dark and dull, so it's almost unheard-of for ISO values to be set to give an 18% gray target. (The Canon Rebel XTi/400D apparently did, and it was roundly criticized for the "dark pictures" it produced.)

18% is the brightness of the standard "gray card." It isn't any sort of in-camera target for metering or auto-exposure.

[autofocus]

Originally Posted by Doug Pardee

That's the imaging sensor. DSLRs have a separate metering sensor, and that's the one we're talking about.

By the way, here's what Canon has to say about its 2-color iFCL metering:
I can't say that I'm personally satisfied with the details in that description. Canon's older monochrome metering sensors use Silicon Blue Cells. SBCs are called that because there's a blue filter in front of them to compensate for the bias toward red. If the description above from Canon were correct, the extra red being picked up by the metering sensor would cause it to underexpose the image. However, many of us with older Canon DSLRs find that they tend to overexpose on reds.

thanks for clearing that up Doug

[autofocus]

Originally Posted by Doug Pardee

This is an Old Wives' Tale, or more correctly an Old Photographers' Tale. The 18% figure isn't correct. The full details on this get to be a big technical mess, and if you really care you can start with this article which made the first big "splash" about the 18% figure being bogus, and then this paper (PDF) which discusses what's really going on.

Then there are some big complicating factors on top of all that.

First, most metering these days is done with a multi-zone metering system, whether it be called Evaluative, Matrix, Pattern, or whatever. These metering systems do not aim for any particular brightness. They aim for a figure that gives exposure guidance which, if followed, will produce whatever the manufacturer thinks the customer will think is a well-exposed picture.

Second, the "18%" is meaningless as applied to metering alone. The "18%" is only meaningful when we're talking about exposure settings. Basically when we're talking about auto-exposure. Furthermore, they're only meaningful for camera-produced images, not Raw — there is no brightness percentage in Raw, which is just a bunch of sensor data.

Third, the metering results depend on the ISO value, or more specifically the Exposure Index. Virtually all camera manufacturers today use the "Recommended Exposure Index" technique for assigning ISO values, and that technique can be summarized as "whatever the manufacturer thinks the customer will think is a well-exposed picture." And customers think that a picture that averages out at 18% gray is dark and dull, so it's almost unheard-of for ISO values to be set to give an 18% gray target. (The Canon Rebel XTi/400D apparently did, and it was roundly criticized for the "dark pictures" it produced.)

18% is the brightness of the standard "gray card." It isn't any sort of in-camera target for metering or auto-exposure.

Thanks again Doug..2 good reads, the 18% myth is broken. The author of the second article poses a good question, if 18% gray cards will yield 1/2 stop underexposure, why haven't the card manufacturers addressed this problem by simply making 12.7% gray cards? I (we) use a gray card occasionally, so it will be good to know that you want to add 1/2 stop up exposure to that reading.

[JFSanders]

Vince, you definitely married up! This has been a great thread and has given me some real good information.

Jim

[autofocus]

Originally Posted by JFSanders

Vince, you definitely married up! This has been a great thread and has given me some real good information.

Jim

You know Jim, I think I can safely say that we're both "old timers" chronologically, and when it comes to how long we've been into photography. And the old adage of "you're never to old to learn" is as relevant today as the day it was written. The wealth and depth of knowledge here on DPS is often nothing short of astounding.