I apologize in advance for a long post; this is a complicated subject

When the idea that the D70 spot metering spot size is larger than that specified first came to my attention my initial conclusion was that the spot size is as advertised.
Shortly afterwards, after a little more testing, I changed my mind and thought the spot to be about 50-100% larger than expected.
Now that I have performed a rigorous (to my standards) test; I have come full circle (pun intended ).

I shot a series of 48 images of a white circle on a black background ranging from 0mm to about 25mm in diameter (3.2mm as measured on the DX sensor).
All exposures were with the 18-70mm lens @70mm f/4.5 ISO 200, manual focused, on a tripod using Aperture priority and Spot Metering.

The area of each image was determined in Photoshop. This also provides the radius and diameter.

With all other factors being held constant, the Exposure Time is inversely proportional to the brightness as measured by the meter.

The smallest circle where exposure time started to stay constant was 2.68mm in diameter.

At first blush this would seem to indicate the spot is larger than the specified 2.3mm
However, this is a trap and is where I (and many) went astray.
The faulty assumption is that our test circles are perfectly centered on the metering spot.

The white circles are almost 9 stops brighter than the black background so we can treat the background as truly back.
Therefore as we increase the diameter of the test circles exposure should be the reciprocal of the area of the circle.

Here is the plot from my test data:

Notice there is a discontinuity in the data.
This is where the largest test circle that fit inside the spot touched the spot.
I marked this point with a cyan diamond.
I also marked the smallest circle that encloses the spot with a green diamond.

Discovering these two points, and some math, yields a 2.3mm spot; as advertised.
The following diagram may help to understand this:

In this diagram the black circle is the spot.
The red circle is for the cyan diamond.
And the blue circle is for the green diamond.
So apparently my test circles were centered .29mm from the true center of the spot.

In my initial test I was much further off because I used the center of the viewfinder for my test circles.
As it turns out the center of my center spot is lower.
The set of data used in this post is from test circles closer to the true center of the spot.

(My AF sensors are also low, so it's more than likely that my viewfinder is actually high!)

So the good news is that the spot metering spot is really 2.3mm in diameter...
The bad news is that it might not be where you think it is

Respectfully,
Bill

Visit me, info and galleries at: Nikon Photos and Information

The open squares on this diagram are a second, more accurate set of data; based on ISO rather than Exposure Time.
(Rather cleverly gathered using ISO Auto in shutter priority at 1/125s)
(ISO Auto used 1/6EV increments )

Note that the blue dotted line is now curved, the original diagram was incorrect; this is correct.
The dotted cyan line, associated with the cyan diamond; is an extension of the dotted red line.
The dotted green line is horizontal at the level of the green diamond.
The cyan and green diamonds have been moved further apart because the placement error of the measuring circles is larger than I originally thought.
The cyan and green dotted lines intersect at the true Area of the Spot.

This diagram now shows the "advertised" spot as well as the measured spot which is slightly larger at 2.48mm versus 2.3mm
Given that even the second set of data is still only measured in 1/6EV this discrepancy might not be statistcally significant.

Respectfully,
Bill

Visit me, info and galleries at: Nikon Photos and Information