Originally posted by Wheatfield Something the measurebeators won't tell you is how absolutely meaningless dynamic range is once you get past 7 or so stops.
Very few scenes will take more than 6 or 7 stops of DR, so any camera that hits more than about 9 stops (I'm pretty sure that is all of them), has ample DR, and any additional headroom, while it looks good on the spec sheet, is wasted.
How true.
Just for every one's information, I did a test with my *istD where I ran 2 series of exposures with a lens (my smc-m 100F4 macro) using a uniformly lit concrete wall. The first set started at F32 and what the green button metering set as normal exposure (the result was a grey scale at about 120) I then left the shutter speed constant and opened the lens up 1 aperture click at a time, and took a shot.
I repeated the tests with the lens wide open, set exposure with the green button, (again about 120 greyscale value) and then stopped down one click at a time.
What I found was that if you potted greyscale vs stops, you could detect about 5 stops between 25 and 225 grey scale linearly spaced at about 40 greyscale/stop, and about 2 1/2 stops either side, of this range but no longer linearly varying, the first was about 17 greyscale the second was about 7 and the remaining 1/2 stop was 1-2.
The only difference between RAW and JPEG was raw had more data in the high end of the exposure, but what it demonstrates is that there is at least 10 stops of dynamic range within the sensor, with 7-8 stops that have reasonably good resolution within the stop, and 2-3 stops where there is some resolution of detail at the extremes.
That was 6 years ago.
I have not tried since, because as you say, every camera has the ability to record most scenes fairly well because the dynamic range of most scenes is well within the really good resolution 7-8 EV dynamic range in the middle of the curve
The