Quote Originally Posted by mablo View Post
I noticed the matrix metering problem with F100 and b&w films several years ago. I tried to raise the issue but people wouldn't notice or told me to fiddle with ISO values. This is nonsense if you know how matrix works. My theory is the matrix exposure scenes were adjusted for shooting chromes. Remember, everyone was shooting chromes when F100 came out. It all makes sort of sense if you think of how to expose for the chromes versus exposing for b&w. You will get underexposed shots for b&w in situations where contrast is wide (bright sky, dark woods for example). The only good cure is to use weighted average for exposure. I've never had exposure problems since I switched to weighted average.
I don't have an F100 but I do have an F5. I do a lot of test with the F5 matrix metering and sending it back to Nikon for calibration twice. My take on the Nikon matrix is this.
With flat lighting and a low contrast scene, it's fine and read the same as center weighted.
With high contrast lighting it bias toward the highlight. It seemed that it try to determine the highlight that is important (fairly large, at the focus point etc..) and try to keep that high light from burning out and that's about 2 stops above middle gray or so.

With that it works great with slide film and also with digital. But I shoot mostly color negative film and B&W film is about the same so the matrix meter would be underexposing in such a situation. Biasing the exposure by putting a + compensation or changing the ISO wouldn't work because in a flat scene it would over expose.