The only time you have to worry about lengthening the lens changing light transmission is if you add additional kit, such as close up tubes or a teleconverter between the lens and the camera.
Originally Posted by bohica
Comparisons need a set of criteria and assumptions as a basis. Conclusions are then drawn. Sometimes the set of criteria and assumptions do not correspond to the practical situation we are faced with, so we should have some idea of how changing one of the criteria or assumptions changes the conclusions. Surely this is a factor that differentiates between an academic discussion and one with practical usefulness. DoF is all about the acceptability of images that are actually out of true focus, so there are plenty of assumed values.
For DoF comparisons we need to choose the limiting circle of confusion. This is not a fixed value for a given format: what is acceptable to someone using T-Max 3200P handheld at 1/15 will probably not be acceptable to someone using Tech Pan in a camera fixed to a solid tripod, for an extreme example. The choice of CoC affects the distance at which the change from DoF being effectively independent of focal length to being dependent on focal length occurs (for the same aperture and magnification). The larger the CoC, the nearer the change will occur (this could be one of the differences between my view of the usefulness of the change and Dan’s).
Perhaps of more practical use: The theoretical and practical examples of DoF being independent of focal length generally assume that the magnification required for comparison is the magnification of an object in the plane of focus. I often find that this is not the case: there is one particular item that fixes the magnification, and it isn’t conveniently in the right place in relation to the zone that I want in acceptable focus.
Having suggested that it is dangerous to make blanket statements, here’s a general observation: if the magnification is set by an object in the near part of the required zone of acceptable focus then the DoF will be greater for a shorter lens. This is the most common situation that I come across. If the determining object is in the far part of the zone, then the DoF will be greater for a longer lens.