Nicolai Grossman and I got into a debate lately about his weblog entry asking why we film-based photographers still shoot slide film.
In short, yes, slide film has a smaller dynamic range, but it also has a higher contrast range. The larger dynamic range of print film comes at a cost.
If we cover about 3 stops of exposure range, the resulting change in density on slide film is going to be about .6. For slide film, the density is often times at least 1.8, often more. With Provia 100F, the change in density is 2.
What does this mean?
Well, if you look at the characteristic curves of slide film, there's about 3 stops of linear film response and then you start getting into the toe and shoulder range, where you lose detail. Whereas print film will have more like 6 stops, maybe more, of linear film range. This means that, yes, print film has more stops of dynamic range, so if you have a huge contrast from dark to light you will be able to pull detail from the highlights and the shadows.
However, let's consider just those 3 stops of exposure range. Print film with a Drange of .6, Slide film with a Drange of 1.8 -- that's at least three times the Drange.
Let's consider scanning each of these pieces of film with a CCD scanner that has 8 bits of "real" bit-depth resolution over the DRange of 1.8. This means that for each stop, there are 85 possible density values over these 3 stops on slide film, but only 28 possible density values from print film.
This means that as you increase the contrast in an image, print film will start to give you banding effects before slide film will. You won't notice this on a stucco wall, but you might with eggshell paint.
With negative film, you capture a low-contrast image and have the option of increasing the contrast later. With slide film you capture a high-contrast image and are required to manage your highlight and shadow detail (lower the contrast) if the scene is too contrasty with diffusers or flash or graduated neutral density filters. Studio shooting, for example, allows you to control quite well your contrast range..
Furthermore, it's not always necessary to worry about shadow and highlight detail. As long as there isn't color crossover (digital cameras are really bad at this) it's OK to let the sparkley highlights off of the ocean lose detail.
Some of this also has to do with just the way that emulsions have developed over time. Currently, there is significantly better data available for long-exposure reciprocity with slide film and the film itself (notably Fuji Provia and the upcoming T64 films) are simply better for long-exposure.
Because color landscapes were first primarily shot with Kodakchrome and then Velvia, people have come to judge that sort of tonality as the way things ought to be. Some of this is just because of the different contrast ratio, but some of this is because of the different way that the C41 and E6 process manage dye purity and other issues. This is the same reason why black and white folks shoot Tri-X even though it has been technologically bested for at least 20 years or why tube amplifiers are still popular even though vacuum tubes have been technologically bested for at least 40 years.
Finally, we still send stuff out to service bureaus. I can and have told a scanning service, "Gee, that doesn't match the slide," and they agreed and rescanned it. You still cannot argue that point with negatives.
I kept telling myself for a good amount of time that slide film wasn't worth the loss in dynamic range. But right now, I mostly shoot slide film.