12

To keep the conversation simple, please focus only on native lens design for the mount.

So with the recent Nikon mirrorless announcement, Nikon has been touting the bigger mount. I am no expert in this area so I was wondering if someone would help me understand this topic.

Nikon indicated that the bigger mount is there so that lens can be made faster. Sony disagree. Based on products in the past, such as the current L series canon 50mm 1.2, seem like Nikon statement is false.

So that's both side of the argument. I don't have enough understanding in this matter to figure out who is right, and who is bullshitting.

Would someone with enough understanding explain who is wrong and who is right, and if there is/area actual benefit to having a larger mount when it comes to getting more lights to the sensor via faster lens?

Control lens. 85mm f/1.0.

Does having a larger mount mean that the design of this lens could be less complicated compare to the design of the same lens on a smaller mount? And if the lens design is less complicated, that mean cost can also be lowered.

  • I don't understand. Modern sensors do not like angled rays. Both short flange distance and large diameter back lenses increase the angle of the rays that hits the sensor. And we know how difficult to Leica (44mm & 27.8mm) to deal with this without excessive magenta drifting and vignetting in the corners. This large mount communication looks like pure marketting B.S. – user2512189 Oct 20 '18 at 5:48
  • I don't understand your not understanding :) — the point about short flange distance seems obvious, but doesn't it seem equally obvious that a larger rear element allows for rays from that element to the sensor to be closer to a perpendicular angle? – mattdm Oct 20 '18 at 12:47
  • @mattdm Some of them, but not all of them. Remember that light from a point source of light strikes every point on the front of a lens and is refracted through the lens to be focused back to a single point on the imaging plane. It would be interesting to see where a short flange distance reaches the point of diminishing returns in this respect for various focal lengths and sensor sizes.. – Michael C Nov 29 '18 at 4:39
  • ... but, the larger the flange distance the most perpendicular rays hits the sensor's extrem corner. I guess !! – Juan Feb 2 at 14:10
16

First: Sony doesn't necessarily disagree with Nikon's claim. It's just that Sony designed their 'E' mount with a throat diameter of 46.1 millimeters at a time when it appeared it would be an APS-C only mount for the NEX series of compact mirrorless ILCs.

Sony later made the decision to move into full frame territory using the all-electronic 'E' mount, rather than either using the existing mechanical 'A' mount¹ that dated all the way back to Minolta film cameras or creating yet another new mount for their FF mirrorless cameras. The 46.1 mm throat diameter of the 'E' mount is just large enough to accommodate the 43.27 mm diagonal of a 36x24 mm FF sensor.

Second: The all-electronic Canon EOS 'EF' mount, introduced in 1987, has a throat diameter of 54 mm. The existence of a 50mm f/1.2 lens in the Canon EF mount with its 54mm throat diameter does nothing to disprove Nikon's claim that their new 55 mm wide 'Z' mount allows better lens design than their previous 'F' mount, which had a narrow throat diameter of only 44 mm.

Basically, since 1987 when Canon introduced the 'EF' mount with a 44 mm registration distance and a 54mm throat diameter, Nikon has been technically limited from matching some of Canon's lens designs due to their own 46.5 mm registration distance and narrower 44 mm throat diameter.² Similarly, with the introduction of the Sony 'E' mount in 2010 and the application of it to FF cameras with the introduction of the Sony α7 series of cameras in late 2013, Nikon 'F' cameras (and Canon EOS cameras as well) were at a disadvantage with regard to designing very wide angle, very wide aperture lenses that can take advantages of the shorter registration distance to both simplify design, reduce size/weight, and match or improve lens performance in a smaller package.

Nikon is touting the newer throat diameter and much shorter registration distance of the 'Z' mount because it is one millimeter wider than the Canon EF mount has been since introduced in 1987 and two millimeters shorter than the Sony 'E' mount has been since introduced in 2010. It's also 11 mm wider and 30.5 mm shorter than their own 'F' mount.

For shorter focal length lenses with wider apertures, a larger throat diameter allows larger exit pupils. A shorter registration distance allows shorter focal lengths without needing to resort to a complex retrofocus design to make lenses with focal lengths shorter than the registration distance. Both of these factors combined mean that larger rear lens elements can be positioned closer to the imaging sensor. This allows for lens designs not possible using narrower throat diameters placed at greater distances from the camera's image plane.

Control lens. 85mm f/1.0

With an 85mm focal length lens, the difference between 16mm and 46.5mm registration distance is not really a factor at all because 85mm is considerably longer than even the 46.5mm registration distance of the Nikon 'F' mount. When one looks, for example, at 85mm lenses for the Sony E-mount and compares them to 85mm lenses with the same maximum aperture for the Canon EF mount or Nikon F mount, it's pretty easy to see that the lenses are about 30 millimeters longer for the 'E' mount to make up for the roughly 30 millimeters shorter mounting flange. The rear elements of the 85mm lenses for the Sony 'E' mount are about 30mm more recessed into the lens.

The wider throat diameter is a factor, because it allows light striking the edge of the imaging sensor to strike it at a more perpendicular angle than a narrower throat diameter would. This increases the amount of light falling on each photosite in the same way that each square meter of ground on the surface of the Earth receives more light/energy from the sun when it is high overhead than when it is low on the horizon. In fact, it would be even more analogous to how much sunlight, based on the angle of the sun, would strike a one meter square on the Earth's surface with a solid 1-2 meter high fence around it, since the pixel wells on typical ILC sensors have depth that usually exceeds their own width.

¹ The Minolta/Sony 'A' mount had a 44.5 mm registration distance and a 49.7 mm throat diameter.

² The previous Canon 'FL' and 'FD' mounts had 48 mm throat diameters that were 4 mm wider and sat 4 mm closer to the film than the Nikon 'F' mount, giving them a slight advantage when designing very large aperture lenses in medium focal length ranges. That's one reason Nikon's 58mm prime could be designed differently than their 50mm primes. The 3.5mm difference between 46.5mm and 50mm is not quite enough room to fit all of the lens elements needed for a higher quality yet simply designed 50mm lens.

10

Using your thumb and forefinger on each hand, make a finger frame, and hold it out at arms length.

enter image description here

Now imagine that your entire field of view is limited by the bounds of this finger frame. Anything outside of the frame is not visible to you.

If you wanted to see a larger (wider) field of view, you have two options:

  1. Move the finger frame closer to your eye. In fact, rest the finger frame on your forehead and nose. You'll still see the bounds of the frame, but you'll be able to see a large field inside the frame.

    This is the analog of moving from the > 40 mm flange depth of DSLRs to the much shorter 16–20 mm flange depth of mirrorless cameras.

  2. Increase the size of the frame by moving your hands apart.
    enter image description here

    This is the analog of increasing the flange mount size.

Nikon did both with its Z-mount system. But really, it had no choice — Nikon had to do both:

  • Manufacturers must have a viable mirrorless camera system to be competitive in today's camera market. Nikon did not have one prior to the Z-mount.

  • Nikon's existing F-mount system dated back to the 1950s. Its 45 mm throat imposed constraints on lens design that are much more relaxed with the 55 mm throat of the Z-mount.

Nikon indicated that the bigger mount is there so that lens can be made faster. Sony disagree.

No, Sony doesn't disagree. Yes, Sony's E-mount throat diameter of 46.1 mm is among the smallest of MILC lens mount throats. But that's because it was initially designed for APS-C -sized sensors, with a possible eye towards full-frame sensors. They simply didn't design for benefits of large throat diameters that might ease lens design in the future, because they were focused on the benefits of small camera bodies in 2010.

Based on products in the past, such as the current L-series Canon 50mm 1.2, seem like Nikon statement is false.

Nikon's Z-mount introduction accompanied by an unimpressive lens lineup didn't disprove the benefits of a large throat diameter; it merely shows that, upon introduction of the mount and initial lens lineup, they didn't take full advantage of the possibilities their new mount opened up to them.

The primary benefit of a large throat diameter on a short-flange distance camera body is not in designing an 85 mm ƒ/1.0 lens. The main benefit will be when (if) they want to introduce a wide angle fast lens, perhaps a 14 mm ƒ/1.4. Wide angle lenses in particular have been limited on DSLR bodies due to the need to use retrofocal optics to achieve the short focal length and wide angle of view. See also:

Nikon's announced Z-mount Noct 58 mm ƒ/0.95, is an example of what they can do with the wide mount that they couldn't do with the F-mount. Nikon's previous fast champion, the 58 mm ƒ/1.2 Noct-Nikkor, had a rear element that was pushed to the very edge of its bayonet boundary. The rear glass even had a groove cut into a portion of its circumference, to accommodate the aperture linkage lever:

enter image description here

Had the F-mount throat been 55 mm rather than 45 mm, it's not likely they would have produced a 58 mm lens substantially faster than the ƒ/1.2 at the time. But they certainly would have had the room to fit both the rear element and the aperture linkage lever without having to cram them together.

  • 1
    It's worth clarifying that a bit. There is a point at which the light from the lens crosses in both directions (hence the need for a pentaprism), at which the throat width could be essentially zero, and you would still get an image (ignoring that the mount would break easily). A larger rear lens element isn't precluded by a small flange size, but it requires either moving it farther away from the flange so the sensor can "see" the whole thing (retrofocal), or moving the flange closer to the sensor, which breaks lens compatibility anyway, so you might as well make the throat bigger. – dgatwood Sep 27 '18 at 5:22
  • You can demonstrate that by making circles with both hands, looking through one at the other. As you move the close one (the flange) closer to your eye, you can see more of the farther one (the rear element), and as you move the rear element farther away, you also see more of it. – dgatwood Sep 27 '18 at 5:24
  • 1
    @dgatwood I don't understand your point re: pentaprism. The pentaprism has nothing to do with capturing an image; it only is useful for seeing through the viewfinder. – scottbb Sep 27 '18 at 5:38
  • The rays cross between the lens and the sensor, which is why DSLRs require a pentaprism (or pentamirror) to prevent the viewfinder image from being upside down and backwards. – dgatwood Sep 27 '18 at 6:04
  • 1
    @dgatwood Rays do not usually cross between the rear of the last lens element and the sensor in a compound lens such as we are discussing. The 'crossover point' is generally inside the lens in front of the rear group. With a simple single lens, it would be exactly halfway between the lens and the sensor. With a (non-retrofocus design) compound lens it is usually still somewhere in the vicinity of half the lens' focal length in front of the sensor. With a lens that uses a retrofocus design it is usually even further forward than half the focal length from the sensor. – Michael C Sep 27 '18 at 6:47
2

Lens Speed

Given that (for one example) Canon built a 50 mm f/0.95 lens for a variant of the M39 mount, which had 39 mm throat, and that remains one of the fastest conventional1 lenses every built for a 35mm camera, I doubt the throat diameter is particularly relevant to building lenses that are simply extremely fast.

Digital Sensor

That leaves us looking for other reasons to do this. The obvious one would be to avoid a problem some have seen when mounting older lenses (especially wide-angle lenses) to mirrorless bodies. With a (non-retrofocal) wide-angle lens, the light can be traveling at a fairly steep angle of incidence when it reaches edges or (especially) corners of the sensor.

A wider throat makes room for a lens that doesn't need to project at such a steep angle to cover the corners of the sensor.

When the light is traveling at a steep angle, you typically expect to see more vignetting, and in extreme cases you can get a rather strange rainbow effect toward the corners.

The microlens in front of a sensor well isn't corrected for things like chromatic aberration. Normally there's no need or point, since all the light sensed by a single sensor well is treated as having a single color anyway. With the light at a steep angle of incidence, however, you can run into a problem: only a fairly narrow range of colors of light are refracted correctly to reach the sensor well at all.

enter image description here

With a wider throat, the lens can be designed so the light is traveling (closer to) perpendicular to the sensor as it leaves the rear of the lens, preventing this problem from occurring (or at least reducing it the point of insignificance).


  1. Leica did at one time build a lens with a higher effective f/stop rating, but it used electronic light amplification, so the actual optics weren't nearly as fast as the effective rating (and I don't believe they were as fast as the Canon f/0.95 either).
  • "... since a single sensor well senses only one color of light anyway..." Not true. Not true at all. No Bayer mask uses hard filters that completely eliminate wavelengths of light not within the range of that color channel. They just reduce sensitivity to other wavelengths in the same way color filters made objects that reflect other colors darker in B&W photos. But some of that other light still makes it through each of the three color filters used for Bayer Masks. – Michael C Oct 21 '18 at 2:12
  • @MichaelClark: Sure. Even an astronomer's H-beta filter doesn't admit only a single color of light. In a typical case, it admits the entire vast range of wavelengths between 475 and 480 nanometers (or so). But hey, why have writing that's understandable, when instead you could include all sorts of detail that's completely irrelevant to the question at hand? – Jerry Coffin Nov 28 '18 at 22:55
  • The point is, some of all wavelengths of visible light make it through all three of the colors used in a Bayer mask. Every sensor well on a Bayer masked sensor is somewhat sensitive to all visible colors, not "... only one color of light anyway." – Michael C Nov 29 '18 at 4:36
0

According to an Amateur Photographer interview with Kenji Tanaka, Senior General Manager of Sony's Digital Imaging Business group:

AP: What are your initial thoughts about the Canon, Nikon and Panasonic systems? For example Nikon says that its larger lens mount means it can produce faster, wider-aperture lenses. What’s your response to that?

Tanaka: The diameter of lens doesn’t matter. Honestly speaking it is very difficult to create a small size mount. But the quality of our new 400mm f/2.8 is better than the competitors. That means the diameter of the mount is not critical for our foremost lenses.

AP: Could you produce an f/1.0 lens?

Tanaka: Yes we could, but there is no market demand. Maybe some demand exists for an f/1.2, but an f/1.0? Technically we could produce an f/1.0, but it would not make business sense.

  • You can produce an f/1.0 lens with a throat diameter barely the size of a pinhole, but the consequence of that is that all of the light striking the edges and corners of the sensor is striking it at much less than ideal angles. Much more than film, digital sensors don't like light striking them at angles significantly less than perpendicular. What do you think Sony's G.M. is going to say? "Yeah, our decision to make our FF mirrorless cameras with the E-mount having only a 46mm throat diameter so we could rush it to market has put us at a competitive disadvantage over the long term?" – Michael C Nov 29 '18 at 4:45

protected by Community Feb 2 at 20:02

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.