How much light does a polarizer really eat?

My recent post about the amazing abilities of modern cameras was based on my bewilderment that I ended up with ISO2500 (using auto ISO) for an otherwise pretty "normal" photo, except that I was using a polarizer. Following that post was a nice discussion on Buzz too, it was interesting to read all the different thoughts.

Jao was rightly wondering what I was really doing there :) as the "sunny 16" rule should apply and that the polarizer would maybe cost some amount of light. Today, I tried it. Another mountain hike, about the same conditions. Bright sun, elevation about 1800m, polarizer... sunny 16 rule with polarizer yields this result:


Sunny 16 rule with polarizer (NIKON D700, 1/200s @ ISO 200; f/16, 24 mm (in 35mm)

Now, add a whopping 3 1/3 stops and you have the correct exposure, as indicated by the camera's light meter. Yes, 3 1/3 stops. Not one or two, more than THREE! Good heavens. Now I know why Singh Ray can get away with that price for their "light and bright" polarizers. ;)


Add 3 1/3 stops... (NIKON D700, 1/20s @ ISO 200; f/16, 24 mm (in 35mm)

The hike went to the peak of the Karlkogel at 2195m; I didn't make it to the peak of the Kammerlinghorn (2483m) (which was my original intention) because I'm about to catch a cold as it seems, I felt too exhausted to continue on to the Kammerlinghorn once I reached the peak of the Karlskogel.

Both are peaks of the Hochkalter or more precisely Hocheis group. Even with an itchy nose and a scratchy throat, the hike was extremely delightful - the area is unbelievably beautiful at this time of the year - turning leaves, larches changing their color from green to orange, the pastures turning from a thick green to a mixture of green and beige (photos will of course follow in a Picasa Web Album.) The Nationalpark Berchtesgaden is a real gem.