Bart, your comments about DOF vs diffraction and shutter speed confuse me.
Hi, it was not my intention. Maybe it's due to the confusing responses that denied the generally shallower DOF of MF,
unless one starts to change apertures to (over-)compensate. Apertures change DOF, which should be no surprise.
Firstly, to get equal FOV and DOF in a smaller format, one adjusts both focal length and aperture ratio in proportion to (linear) format size, which is to say keeping equal effective aperture diameter. That gives circles of confusion smaller in the same proportion as the image formed by the lens is smaller, so equally large on prints (or any other display medium) that have the image appearing equally large. From the examples given, this seems possible in 35mm format in comparison to most if not all MF lenses.
Yes, it it possible to compensate for the differences in DOF that are caused by the different focal lengths needed to adjust the FOV. No disagreement on that, as far as I'm concerned.
But one
will have to adjust the Aperture to achieve that, and that also requires a change in exposure time to compensate. There may be constraints (e.g. too long an exposure time to avoid a certain level of motion blur on the MF gear).
Secondly, the lower aperture ratio used in the smaller format to get equal DOF also gives equal diffraction effects on equal sized prints: like he circles of confusion, the diffraction spot sizes are also smaller in proportion to the size of the image formed at the focal plane, so come out the same size on same-sized prints.
The aperture number (which determines the angular aperture or Numerical Aperture) determines the diameter of the Airy Diffraction pattern. So regardless of the focal length, the same f-number will produce the same amount of diffraction on the sensor plane. With the required (usually) smaller output magnification of MF, the diffaction in output will be less. Obviously, if we had compensated the DOF with changed apertures, things like diffraction would also be equal. One exception, if the required narrower aperture number for the MF exceeds the diffraction cut-off frequency of the sensor, there will be a total loss of resolution, while there might still be some marginal resolution on the wider aperture for the smaller format.
Thirdly, as to shutter speed: this could be kept the same if the ISO speed can be reduced enough (in proportion to image area) which would also lead to roughly equal photon counts and do equal S/N ratio as far as photon shot noise goes.
A lot can happen when ISO is changed (depends on the sensor technology and supporting electronics), but it doesn't change the number of photons that a given aperture number allows to pass (per unit time). The physically larger aperture of the longer MF focal length lets in more photons, but they are spread over a larger sensor area, so total exposure per pixel remains the same, until we change the aperture numbers between formats to compensate for DOF differences. Changing aperture numbers will also need to be compensated for by adjusting exposure time or flash intensity, because different numbers of photons will be let in. ISO does not change that.
In cases where the ISO speed cannot be reduced enough, the consequence is only that the smaller format might have to use a higher shutter speed, which is usually not a problem.
Or the larger format has to use a slower shutter speed, which might get to be an issue at large magnifications (although MF usually requires less output magnification).
In the rare cases that this higher shutter speed is undesirable, an ND filter can be used – however that would leave the larger format with the advantage of higher photon counts and so potentially better IQ through higher S/N ratios and such.
Higher photon counts, but also spread over a larger surface! Local exposure of larger magnified detail is where the S/N ratio counts, and where MF has opportunities (in addition to higher MTF due to lower spatial frequencies of the more magnified projected scene), especially in postprocessing it will give more quality to work on. Unfortunately aliasing artifacts may prevent exploiting some of the advantages.
Cheers,
Bart