Mount size
With Sony’s E mount 10mm smaller than that of new mirrorless cameras from Canon and Nikon, is there a justification for a large lens mount?
One of the most commented on features of the new Nikon Z-series cameras is the large lens mount diameter. In fact, it’s not as large as the rumour mill would have had it before the full details were known. At 55mm, it possesses a bare one millimetre larger in throat diameter than the Canon EF lens mount, which is also the diameter chosen by Canon for its new RF mirrorless mount. Nonetheless, it raises the question, what is the advantage of the larger lens mount, given that the competitive Sony E mount is a full 10mm smaller and appears to function perfectly satisfactorily.
If you do a web search on the benefits of a large diameter mount, most of the pages found will talk about the ability to support large aperture lenses, yet the venerable Leica M mount has a clear diameter of only 44mm, and has had available lenses as fast as f/0.95. There is a simple principle that dictates how fast a lens a mount can serve. Conventionally, the f-number of a lens is defined to be ratio of the focal length to the aperture. Thus a 50mm lens with an f-number of 0.95 would have an aperture of 50mm/0.95, or 52.6mm. The same ratio applies to the position of the exit pupil and its diameter. If the exit pupil is projecting a cone of light to a point of focus at the centre of the frame, then if the lens mount is not to obstruct this cone, some simple geometry shows that the ratio of the distance of the lens mount from the sensor (its register) and its diameter must also be the same. This tells us that the Leica M’s 44mm diameter and register of 27.8mm could support a lens with an f-number of 0.63, which should be fast enough. The new Nikon Z mount, with its 16mm register and 56mm diameter could accommodate an f/0.28 lens, which is physically impossibly fast for a wellcorrected lens.
However, projecting light to the centre of the frame is not the whole story. Light also needs to be projected into the corners of the frame. Whilst an exit pupil positioned right in the lens mount can project to the whole of the sensor, in the corners of the frame, the light cone strikes the sensor at a very oblique angle. This worked quite well with film, but is problematic for silicon sensors fitted with microlenses, which are used to enhance sensor efficiency.
A partial solution to this is to offset the microlenses in the corners and edges of the sensor towards the centre, so that they can more easily accept light projected from that direction. An optically better solution is to design lenses which have an exit pupil further from the sensor, allowing the light cone to strike the sensor with more perpendicularity. However, this requires a larger sensor throat. The diagram, left, shows how this works.
The black line on the right represents the 43mm diagonal of the sensor, while the diverging dotted lines show how the width of an f/0.95 exit pupil varies with its distance for the focal plane. The green exit pupil is placed at the Leica M mount’s 28mm flange distance, the red at 50mm and the blue at 100mm. The blue ‘L’-shaped lines represent the flange of the Z mount, and from this we can see how it can accommodate an f/0.95 exit pupil at 100mm. Meanwhile the green ‘L’-shaped lines represent the Leica mount, and clearly only the 28mm exit pupil can properly illuminate the corners. Thus, the new Nikon mount allows for more digital-friendly fast lenses.