[c a sugg]

14 stops with the K-5? Really?
I'm not sure if that's possible with B&W film and still have a usable tonal separation.

[telfish]

Sounds more like a statement than a question. The DR was tested at over 14 stops. End of story.

[falconeye]

14 stops with the K-5? Really?
I'm not sure if that's possible with B&W film and still have a usable tonal separation.

I don't like the above answer as I think the original post asks a valid question.

As a matter of fact, there is no ISO or DIN standard how to measure dynamic range.

Consider for a moment that the sensor in the K-5 can register either 1 photon or the maximum of photons before it runs over (for equal illumination). Which are 16 million times 40,000 photons. Or 640 billion photons. The ratio of 640 billion : 1 is 2^39 : 1 or 39 stops. So, every dynamic range statement below this theoretic maximum is based on some definition.

DxO defines dynamic range as the ratio of luminosity for bright white and a dark tone where noise approaches the same magnitude as the signal itself; when looking at the image resized to 8 Megapixel. This is the best dynamic range definition up to now.

And yes, according to this definition, the K-5 has 14.1 stops DR. In the same way as a 7D has 11.7 stops DR or about 2 1/2 stops less. This difference in DR is real and largely independent from an actual definition. The absolute value however doesn't mean much w/o digging into the definition how DR was defined.

For practical purposes, I would subtract 3 EV from all DxO figures when it comes to decide how much exposure lattitude you can count on. Which means you can still underexpose by 3 stops to preserve full 8 stops of output dynamic range w/o sacrfycing in tonality.

When it comes to camparing with film, you first would have to come up with a good definition of real lab measurement for film DR. My own impression is that DR in a K-5 is on par with medium format film. For B&W, it would depend a lot on the gradation too. But I'd like to see a serious treatment about the question myself. I sometimes see densitometer RMS measurements for film but that can only be a starting point.

[nanok]

what falk said. the other problem when comparing to film is that film does not have a linear response as digital sensors do, so the comparison becomes very difficult at the "edges" of the dynamic range. so as Falk said, you would have to properly define dr on film (separately from the definition on digital, but so that it enables you to get similar results in practice for similar figures obtained. not easy, i should think)

[falconeye]

The other problem when comparing to film is that film does not have a linear response as digital sensors do, so the comparison becomes very difficult at the "edges" of the dynamic range.

That's true but it isn't fundamentally different.

Digital (JPG) has an output DR of 8 Bit (which actually is 11.6 stops rather than 8 stops because unlike common belief, it isn't linear either). The output DR of film would be the Dmax or density value. Dmax can be close to 4, like 3.9 and means 10^3.9 or 13 stops. That's the trivial part on both sides which most reviews confine themselves too.

The interesting part is input DR, the DR in a scene which can be captured without too much loss in tonality or too much noise. DxO measures this for digital. The same could be done for film but I know no source who actually did.