E-mount for Nikon MILC, good idea or not?

Started Sep 23, 2017 | Polls thread
bobn2 Forum Pro • Posts: 57,696
Re: E-mount for Nikon MILC, good idea or not?

cosmicnode wrote:

bobn2 wrote:

cosmicnode wrote:

bobn2 wrote:

cosmicnode wrote:

bobn2 wrote:

cosmicnode wrote:

bobn2 wrote:

cosmicnode wrote:

bobn2 wrote:

Antonio G wrote:

One of the problems a lot of people points out against the use of the F-mount in future mirrorless models is the lens throat diameter that prevents designing/using the ultra-fast aperture lenses and the E-mount wouldn't bring any advantage there as it shows just 2mm more and still far from other competitors.

It's not just the diameter of the mount, it's the register. The E-mount has a register of 18mm whilst the F mount is 46.5mm. It means that the E-mount straddles the exit cone from the lens where it is much narrower, therefore the mount can be narrower. That being said, the F-mount can go down to f/1.2, and there haven't been many faster lenses at FF.

You have this back to front

I really don't.

the actual distance from the rear of a Sony lens to the sensor is approx 12mm taking into account the thickness the lens protrudes into the mount. this makes it harder to cover the sensor than the similar F mount with a 47mm registration distance where you can have a larger cone . Think of it like this, the largest opening in the lens barrel can be about 39mm, move this even closer to the sensor until it is touching and you eliminate the sensor corners completely from the exit cone.

That's a different effect. What we are talking about is what can be the subtended angle of that light cone. Yes, you're right, in that in that extreme, the mount would shade the edges of the frame. However, as a sanity check, think of the Classic Canon 7, which used the 39mm Leica screw mount (register 28.76mm) and came with an f/0.95 lens.

This lens didn't actually mount on the screw mount, but the clear aperture was 39mm.

Consider long telephoto lenses, they are in the position with the short registration distance of vignetting caused by the rear element being a good distance up the lens barrel, my 200-400 has the rear element some 80-100mm away from the sensor

Do your 200-400 or 80-100 have an f-number of faster than f/1.4? That's what we were talking about. In any case, pretty much every designed for digital lens has the exit pupil something like 100mm away from the sensor, even if they are wide angles. Just do the geometry, even in the far corners, hardly any of the exit pupil is shaded by the lens mount aperture, probably not enough for the vignetting to be noticed. Here's a diagram to show you what's happening

On the right is the exit pupil of an f/1 lens, 100mm from the image plane (actually the distance doesn't matter, because the light cone always subtends the same angle, in proportion to the f-number). On the right in cyan is the 43mm image diagonal. In green is the 40mm aperture of the lens mount at 18mm register. As you can see the dark green shaded area is the part of the exit pupil obscured, in the end maybe 1/3 of the exit pupil (can't be bothered to do the exact sum), 0.6 stop loss in the extreme corners. And of course, that is using a lens not designed for the mount, if the lens is designed with a positive rear element filling the mount circle, there need be no shading at all.

You need to re-draw that Bob the "rear" of the lens exit tube is only 12mm off the sensor, everyone forgets that the lens mount has a bayonette that has a thickness which protrudes into the mount throat, this is about 6mm which must be subtracted from the registration distance, this opening in the lens body is at "best" a max of 39mm, not the 43mm of the lens throat and this for mainly telephoto lenses where the elements are well up inside the lens body, wide angles and normal lenses often have focusing element tubes in this area reducing this diameter further. Only wide angle lenses will have the posative rear element filling the lens mount circle, normal and tele lenses are well inside the lens tube on such a short registration distance, it's these small details that make the difference to how perfect the lens mount can be . Make the lens mount Sony or Nikon 5-6mm bigger in diameter in order that no part of a 43mm diagonal sensor is obscured by any part of the lens material and it does not matter what the registration distance is. Make it slightly bigger still for the movement of the most aggressive IBIS, of course you will never get a new mount that will take a adaptor for Sony or Fuji lenses there simply is not enough room.

Your discussion is based on a whole load of incorrect assumptions

They are not assumptions but actual dimensions.

The assumptions aren't in the dimensions, they are in the understanding of optics.

The position of the rear element does not matter, for this discussion (I accept the point that when it's far enough back, it will physically shade the frame), it is the size and position of the exit pupil that does. Also, there is nothing making the lens protrude through the mount, if the designers choose not to do it that way. Then, check through what I wrote again. The 43mm refers to the image diagonal, I used 40mm for the lens throat opening. As for 'only wide angle lenses will have a positive rear element"

"filling the rear mount" don't cut up sentences to change the meaning.

I'm not sure what that means. What I said was 'if the lens is designed with a positive rear element filling the mount circle, there need be no shading at all.'. The reason is simple, a positive element at the rear of the lens provides exit pupil magnification at that point, the lens mount is no longer a restriction. Certainly, there will need to be other elements in front which compress the ray bundle sufficiently, but even with the positive rear element, there is no reason why that rear group cannot overall have a negative power, as would be necessary for a telephoto.

any lens can have a positive rear element if designed that way (a retrofocal lens needs a positive rear group, but that is different from a positive rear element).

As for the rest, sure a bigger mount has advantages, wat we were discussing is whether the E-mount restricted use of lenses faster than f/1.4 - it clearly doesn't.

I'm simply correcting assumptions based on incorrect dimensions simply taken off the net.

Which 'incorrect dimensions' were you referring to. The only one you mentioned was a reading error on your part, not an incorrect dimension.

Those taken of Wikipedia . they are correct but do not refer to the same parts of the lens mount so cannot be compared across mounts. the registration distance is from the front of the mounting face to the sensor not the rear of the lens barrel which in the case of Sony is around 12mm.

You're still barking up the wrong tree. The question is, how does the lens mount obstruct the light cone from the exit pupil. Sure, placing a stop at the register is a bit of a simplification, but not an extreme one.

The important dimensions will be the diameter of the exit of the lens and its distance to the sensor when working out light paths.

That's simply wrong, which is what I've been telling you. Since the rear element is refractive, the light does not take a straight path though it. If the rear principle plane is far enough back, a small element can illuminate all parts of the sensor (albeit, at oblique angles in the corners). The diameter of the exit of the lens is in proportion to its distance from the focal plane. The exit lens for an f/1.0 lens needs to be 12mm if it's 12mm away from the focal plane, 18mm if it's 18mm away and 46.5 mm if it's 46.5 mm away, and so on, if the rear principle plane of the lens is at the exit lens. Place it further forward, and it gets bigger if vignetting in the corners is to be avoided (which is one reason that 'design for digital' lenses have tended to restrict maximum apertures). For the F-mount, which gives a minimum distance for the rear element of 38mm, and has a clear diameter (with a slim contact block) of 37.5 mm (not off Wikipedia, just measured for myself) that gives a minimum achievable f-number of 1.01, which one could call 1 between friends. In practice f/1.2 shouldn't be a problem with an E type lens and a more sensible contact block than Nikon's standard design. For the E mount, with the same diameter and a smaller register, it clearly isn't a problem, a Nikon f/1.2 on an adapter can do it, so build on the adapter and there you are.

-- hide signature --

Tinkety tonk old fruit, & down with the Nazis!

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow