Hi Jaap & Charles,
jaapv said:
Philipp, the surface of a sensor is a plane that behaves like the perfect surface and is flat and two- dimensional,
a film has thickness, and is not perfectly flat, so three-dimensional. Thus the tolerances on film a wider than on a sensor. This has an effect on DOF as well. It is more defined on a sensor. A film produces a cone of light inside the fim, elliptic towards the edges with reflections and refraction inside the emulsion. A sensor has disks controlled mainly by the resolution and pixel size. Therefore DOF is more defined and steep on a sensor than on film.
usccharles said:
thats not entirely true. its because film isn't completely flat like a digital sensor when a picture is taken. the film plane will have slight variences in curvature which will give soom 'room' for the critical focus point.
yes the rules of DOF ar strictly geometric, but the rules become a little different when you change from film (with alittle flex) to a digital (completely flat and secure in position to the lens). you would thus have much less tolerance on a digital back compared to a film back. thats atleast the reasoning i've read around here and it seems to make sense to me
Sorry, I think you are both wrong. I think we need to distinguish between two different geometrical aspects in which films and sensors may or may not differ, and that's (a) flatness and (b) thickness.
Thickness is what Jaap alludes to when he talks about cones of light inside the film that we don't have in a digital sensor. However, there's two remarks to be made about this. Firstly, in a film camera the film shows the emulsion side towards the lens, so that the thickness of the
film itself is actually irrelevant as long as it's constant and standardised across films, what's relevant is the thickness of the
emulsion. Secondly, film manufacturers take great pains to avoid reflections etc. inside the film itself and inside the emulsion, which is why most films don't have a perfectly clear base, and why you have all sorts of flare protection layers inside the emulsion which get washed out when fixing (with a film like Lucky you get all sorts of weird flare effects). This is also why unexposed film is not transparent. Thirdly, the thickness of the emulsion is actually as close to zero as it gets; an ideal B&W film has exactly one layer of silver halide crystals, and these are very, very small in relation to the dimensions we are talking about. In particular, the crystals are a lot smaller than the acceptable circles of confusion that form the basis of any discussion about DOF (otherwise the whole concept of circles of confusion and DOF falls down and makes no sense). Fourthly, a sensor has a thickness, too, otherwise concepts like the Foveon sensor wouldn't work where a photon's colour is determined by its penetration depth. So film and sensors aren't fundamentally different in this respect.
Flatness is also alluded to in Jaap's post and forms the basis of Charles' post. The argument is, if I understand correctly, that if the film is less flat and bulges a little bit, then there is more room for the film to fall in the imagined plane of sharpness. However, this is just plain wrong. Focusing with any film camera is based on the assumption that the film is in a known location; focusing works by altering the lens geometry in such a way that the image is projected exactly onto the location of the recording medium. If the film isn't flat, this does not add any depth of field; on the contrary, because the film isn't where it's expected to be, it moves those areas of the negative that should be in focus out of focus. So if the film isn't flat, not only does DOF
not get any bigger (in fact in practice DOF should be minimally affected, if at all), but also the focus is wrong. I don't see a gain in that. The best thing that could theoretically happen at this point is that areas that should be out of focus happen to be in-focus, because the cone of light from their projection is focused at a point in front of where the film plane should be, and now the film accidentally bulges into exactly this place. However, this is not only undesirable from the photographer's point of view, but it also does not add any
depth of field either, it just results in an uneven distribution of sharpness across the frame. Camera manufacturers take great pains to stop the film from bulging and keep it flat. If the film isn't flat, it's a problem, not a source of added DOF. It's exactly the same as if saying that having a wobbly lens mount adds DOF, because the lens wobbles back and forth.
DOF has very little to do with what is happening behind the lens. It's just a way of thinking about how the human eye can be fooled into perceiving unsharp projections as sharp by keeping the unsharpness below a certain threshold. If the film moves back and forth behind the lens, this does not add any DOF. What is confusing in this respect is I think that when we talk about "depth" of field and about film bulging three-dimensionally in space, our mind assumes that both are connected with depth, and is thus fooled into thinking that they are somehow related.
Philipp
