.mark. wrote in post #9100502
Sorry if this has been asked before but I've searched and not found anything...
Ok, so we talk all day every day about lenses. 24-70mm this, 85mm that... but who comes up with the number.
I'll talk about primes to keep it simple. Take a 50mm lens. What physical property are we talking about which measures 50mm? I thought it might be something like the distance between two glass elements in the lens.
To keep it short: the FL of a lens is determined by the distance from the lens plane to an image sharp in focus when the lens is focused at infinity.
As an example: when a lens is focused at the sun, which for all intents and purposes is at infinity in this discussion, the focal length is the distance from the middle of the lens to the focused image of the sun you use to burn a hole in a piece of paper. And this is true for any lens.
For a composite lens, or objective, as we use in photography, this generally is the distance from the rear lens plane to the sensor. This is also called the back focus point, or rear nodal point, etc.
There seems to be a correlation between the physical length of a lens and the focal length, e.g. an 800mm lens is much longer than a 50mm lens.
Yes and no. A true telelens like an 800 mm may well be built a lot shorter than its FL indicates, as additional converging lens elements are used to shorten the barrel. A lens of, let's say, > 100 mm FL with a length, including the mount to sensor distance which is equal to the FL, IOW, not shortened by additional lens elements, is very aptly called a long lens, rather than a telelens or telephoto lens which includes the converging lens elements construction.
But then how can a 35mm lens also be longer than a 50mm lens, and the 24mm lens physically longer still? (comparing the Ls)?
Because of the fact that FL requires a certain distance from the lens to the sensor (or film for that matter), with dslrs there is a problem with lenses which have a shorter FL than approximately 45 mm, generally speaking, as the mirror box and mechanism get in the way of using short lenses.
In order to get around this, optical engineers came up with the opposite of what they applied with long lenses to shorten those, namely a set of diverging lens elements to distance a short lens further from the sensor, to make way for the mirror box and all the moving elements therein. This design is called a retrofocus design, and is typical for (d)slrs. Range finder cameras don't need such a design, there is no mirror for viewing the actual framing etc. You will see that a WA lens for an RF camera generally does extend into the camera body, however, obviously depending on the thickness of the body, or the distance from lens mount to sensor, if you will.
The FL doesn't change with the retrofocus design. It essentially introduces a virtual rear nodal point, which doesn't interfere with the mirror.
Using such a design has a few consequences, however. First of all, it requires much more optical corrections, and in addition to the glass already introduced for the retrofocus design, it also requires more glass for these corrections. Furthermore, the aperture plane, essentially the spot where the true focal length is determined for a lens, the true lens plane, is now positioned rather far away from all the other glass required to form the image. Because of this and because of the rather wide AoV (angle of view) of a WA lens, it requires the image capturing elements to be rather large, other wise there would be a lot of mechanical vignetting. This gets worse when the aperture gets larger, like is the case with L WAs.
This all adds up to a lot of glass, and large glass especially at the front, in order to make these lenses. Hence the size of these lenses.
HTH, kind regards, Wim