I still like to shoot with film (I use a Nikon F3) but I like the immediate feedback of digital. I'm thinking of buying a used DSLR to help me get the framing and lighting right when I shoot static subjects. Then I'd take the final shot with my F3. But would the metering be transferable? If the shot looks perfect at 1/125 with f/2 on a DSLR, will the same settings work on a film camera?
Apparently no one has mentioned the reciprocity error (Schwarzschild effect) when measuring for film, and it's probably the only things OP should be concerned about. I'm actually surprised Philip Kendall suggesting there is little difference in 't-stops' when this is actually the only HUGE difference between the two.
You see, the response of the digital sensors was designed from the film response in the 1/10000 - 1" range, which was applied lineally to the rest of the range.
In digital, to get an extra stop you just multiply your exposure time by two, and that is true for any given range (the maximum most cameras work with is 30", but this rule applies for longer ones). The response is linear.
But in film this 'linear' approach only works as an approximation in a very small range (The previously mentioned 1/10000 - 1"). The response is NOT linear AT ALL, and it depends a lot on the film you are using (Refer to the datasheets for more information). Let me explain this with a handy chart you will be using a lot (It was taken down from flickr but I treasure a copy in all of my thumb drives just in case):
As you can see, once you pass the 1 second mark, things get wild. If you measure something with your digital camera and decide that 10" is OK, it will be OK in digital. In fact it will probably look OK in any digital camera you get your hands on.
If you try to take that photo on film with the same parameters, it will look... Well, very bad. If, for example, you are shooting Tri-X (Why would you shoot with any other thing?), you need to add 2+1/3 stops to that time to get the same exposure. That is, 10"*2*2 (40 seconds) +1/3, which total around 50 seconds.
You can see how different the times can be for not-so-long exposures. I'm not going to go through why this happens, you have plenty of info on that pretty much everywhere (Any serious book on film will cover it, I always recommend the Ansel Adams trilogy for anything film-related).
The same thing happens when you shoot faster than 1/10000. The film does not behave in a linear manner and you should correct for this effect.
Other than that, measuring for film is the same as measuring for digital, and you get the same effects modifying the same parameters (Depth of field, motion blur, etc), with the only exception that you get more grain instead of digital noise when pushing film or shooting with sensible emulsions.
To a very good approximation: yes, assuming you've done the obvious and set your DSLR ISO to whatever film speed you're using. The definition of "ISO" is the same for film and for digital.
There are a number of reasons why you might not get exactly the same exposure between the digital and film setups:
Note that the above applies only for "regular" shutter speeds - if you go above around about 1s, then the Schwarzchild effect will start applying and you'll get a big divergence between the digital and film responses - see Achifaifa's excellent answer for a detailed explanation.
Finally, a note on sensor sizes: if all you want from your DSLR is to act as a "super light meter" for your setup, then it doesn't matter what size sensor you use. However, if you want to check other aspects of the composition (depth of field, etc) then you'll need to be using a DSLR with the same size sensor as your film camera (i.e. full frame in your case - which will push your cost up significantly). If you are only looking for a "super light meter", then it's possible you don't need a DSLR at all and could get away with just a compact camera. Again, see jrista's answer (and the comments) for more details on this.
So, absolutely you can do that. In some ways you're suggesting a similar approach to the old polaroid and medium format professional shooting. Basically, the photographer would take a polaroid shot to confirm light, shadows, and general scene before shooting with the big camera. It's always more expensive to get back to the darkroom and discover that you don't have what you need.
Translate that forward and film has become more rare and more expensive to develop. The digital camera would provide the same instant feedback of information as polaroid used to do and even better, will give histograms and other scene information that polaroid couldn't do. The settings should be the same or very close, bearing in mind that the ISO setting on the digital camera is analogous to ISO on film, but not exactly the same. In other words, it could be a fraction of stop off.
The gotcha? I suspect you may find yourself moving more and more to digital. You can do things at ISO levels that film just can't touch anymore.
You can indeed use a DSLR to help improve your film photography skills. At the very least, you can use the DSLR to meter, and determine what your exposure settings and ISO need to be, assuming you use an identical lens.
It should be noted that if you use a film SLR, it is probably 35mm format. Most DSLR cameras these days are APS-C, or cropped frame. Instead of being 36x24mm, the sensor of a cropped sensor is around 23x15mm. Cropped frame DSLRs are cheaper, and often below the $1000 mark, but they probably won't help you the way you would need.
To ensure you have the easiest time achieving your goal of using a DSLR to help you expose properly for a 35mm film SLR, you should get a full frame DSLR. This is due to a set of key fundamental differences between FF and "cropped sensor" (APS-C) cameras. The final results, in terms of exposure, noise, dynamic range (for the most part...some key differences between film and digital, for overall they shouldn't matter), and composition, require different settings on FF/35mm film than on APS-C. To demonstrate, the following two cameras and exposures will produce identical results:
The following will NOT produce identical results, despite the identical exposure:
Nor will this:
The key difference is framing, motion blur and depth of field. To produce an identical image on APS-C as on FF, you must reduce the focal length to counteract the crop factor, and use a wider aperture to achieve the same depth of field. This necessitates a reduction in ISO, because you cannot change shutter seed as that could result in different subject blur.
This also assumes that you have the option of using the appropriate focal length on either the FF or APS-C in order to achieve the desired results. If you are using an APS-C DSLR to teach you how to expose properly, with a 135mm lens, you would need a 200mm lens on your FF SLR to get the same framing. This either necessitaes having a very broad range of lenses to support your goals, ensuring you have the necessary focal lengths on both APS-C and FF, and be willing to swap lenses when necessary...or requires a FF DSLR and simply swapping the same lens between it and the film SLR.
It is certainly possible to convert exposure settings between FF and APS-C, however since your goal is to use a DSLR to help you learn how to use a 35mm film SLR, the best tool for the job would be a FF DSLR rather than an APS-C (cropped) DSLR.
Nikon currently offers the D610, which is a fairly reasonably priced full frame DSLR that would probably fit your needs. If price is not an object, you could look into the D800, which is similar to the D610 with a higher resolution sensor and more professional grade features. Both cameras offer EXCEPTIONAL dynamic range at low ISO, and more resolving power than any other cameras in their class. Being Nikon, they should work with all of your lenses as well (although there are certainly some caveats there, the F mount has a long and varied history.)
Yes, you can use a digital camera to take test shots in much the same way pros used Polaroid backs for many years. Just keep in mind that both film ad digital have some leeway with regard to ISO. Many digital cameras can be 1/2 stop or more less sensitive than they claim for a particular ISO. One of the things DxO Mark tests is actual sensitivity for each full ISO stop. Here's a link for the Nikon D610. Different films also vary slightly from the exact ISO they are rated at. You might run into a case where both the film and digital vary by the same amount in the same direction and so they are both almost equally sensitive or you might find a combination where the film is more sensitive and the digital camera is less sensitive and wind up with differences that add to one another.