Exposure is a problem everyone who uses a camera faces. The matrix or multi-segment metering of todays pro cameras are suppose to give closer to perfect readings for a given subject/background then the standard center weighted metering. This document describes the matrix metering capabilities of two of Nikon's top of the line cameras the D1X and the F100. The D1X is Nikon's top of the line Digital SLR that sports a 1005 segment color matrix meter. The F100 is Nikon's top of the line little brother to the flagship F5. The F100 sports a 10 segment matrix meter. An F5 would have been tested however, one was not available during testing. A complete descriptions of these cameras and their meters can be found at the NikonUSA website. It is my hope that other Nikon users will make good use of this data as I am. If one knows how a given meter will act given specific background/subject combos, then the correct compensation may be applied to achieve a perfectly exposed image.

Three backdrops were chosen as backgrounds for this test which were dark, middle tone, and light. Three subjects where also chosen and were dark, middle tone, and light. The dark background and subject where black in color chosen for their dark tones. The middle tone background was a brick red chosen because it metered very close to a Kodak grey card. The Kodak grey card was chosen as a middle tone subject. The light background is a sky blue color that metered lighter in tone then middle tone. The light colored subject was a white stuffed animal chosen for the white color and texture in the fur.

The morning was bright and clear at 10am when this experiment was done. The backdrops were set up outside with the sun at my back so that the subject and backdrop would have frontal lighting. A Sekonic L-508 set at zero compensation was used in incident mode and in spot mode with a Kodak grey card to get the correct base exposure for the experiment. The spot meter was used to measure the exposure of each of the three backgrounds (Table 1).

Table 1. Incident metered exposure and spot metered exposures

In turn, each of the backgrounds was metered with the D1X and F100 with each of the chosen foregrounds. An image was created using the D1X with each background/subject combo for a visual reference of the resulting exposure. All meter readings were taken with the background filling the viewfinder and the subject filling the reference circle within the viewfinder. I did not change my position during the test and this brings a point about the 1.5x crop of the D1X. To keep the subject within the reference circle with both cameras, a Nikon 28-105 was used. This allow me to zoom to whatever extent was needed to fill the reference circle with either camera. I worked in aperture priority and kept the aperture a f5.6 in all cases. In one instance with the F100, it could not make up its mind on the actual shutter speed it wanted to use. The slightest movement of the camera would cause the shutter speed to change. Basically it teetered back and forth between two shutter speeds. This was not a problem on the D1X as it picked a shutter speed and stuck with it right or wrong. The exposure data follows in Table 2 as well as the images from the D1X.

Table 2. Exposure data

The first thing we realize is that matrix metering is not perfect! The only time either camera metered the same exposure as the incident or spot metered grey card is when the middle toned background and middle toned subject were used. With a few exceptions we see that the most +/- compensation needed is 1/3 stop. The exceptions are indeed critical though especially with the Dk/Dk and Dk/Mt combos. The Dk/Dk combo over exposed the image by a full two stops trying to bring the tonal value to something lighter then it is. The Dk/Mt combo overexposed by one full stop again bringing the tonal values up. The Bk/Lt combo overexposed by a 1/3 a stop blowing out the whites in the stuffed animal. Indeed it looks like both cameras meters have the most trouble with an extremely dark background and overexposes in all cases. It should also be noted that the Mt/Dk combo overexposed by one full stop.

Up to this point the cameras metered in an identical manner. However, the Mt/Lt and the Lt/Mt combos yielded a 1/3 stop difference in exposure between the two cameras. With Mt/Lt the D1X underexposed by 1/3 stops while the F100 nailed the incident reading. This could account for a number of people claiming that the D1X underexposes compared to its film counterparts. It is my opinion that Nikon chose to do this with the D1X in an attempt to keep from blowing out the highlights in an average scene. The Lt/Mt combo turned up interesting results. The D1X underexposed by 1/3 a stop while the F100 underexposed by 2/3 a stop. Maybe it is just a fluke but I did try the F100 at different angles seeing if the exposure would change and it was no dice. The F100 was unyielding in it's 1/1600 shutter speed. Last both cameras underexposed by 2/3 a stop with the Lt/Lt combo.

It is my conclusion that a +-1/3 stop compensation is all that is needed for most scenes with the exception of Dk/Dk and Dk/Mt which needs a -2 and -1 stop compensation respectively. You may find other differences between the two cameras and other scenes may indeed need different compensation values. One set of combos I did not try was a white background with the various subjects. The reason this was not included is I did not have a white background. I even looked for a white sheet in the linen closet to no avail. It appears Robin does not purchase white sheets for the beds.