A hypothesis concerning the RF mount

[blackcoffee17]

I don't think anything will be announced on 5th, except lenses. Hard to believe there are no credible leaks only days before the announcement.

 

[applecider]

A moveable sensor doesn’t sound so bad to me. I’d think that the sensor would only move in a macro way with lens changes, I change lenses on a given body less than ten times a day even with busy shooting. If another lens is needed it usually goes on a separate body. So a moveable sensor perhaps on a rail, could be called upon with lens type changes. EOS R for railed sensor.

Canon already has patents for conformational sensor changes, so this is not a bridge too far.

Unlike others I see a ML as initially a cross grade rather than an upgrade, unless a new sensor technology comes with it.

Now a fifty MP, ten plus FPS ML with full screen cross type focus points handling f8 focus would be tempting.

 

[bks54]

My hypothesis, FWIW, and based on the limited information we've seen here, is that there is some type of automatic flange extension that pushes out the required 1/4 to 1/2 inch when an EF lens is mounted, but locks in the recessed position when it is an "R" lens.

Seems complicated, but maybe.

Such a system would probably require more than than 1/4 to 1/2 inch extension; more like 26 mm (more than an inch) since EF is at 44 mm and thin mirrorless much shorter (18 mm for EF-M). If they want to do something like this it would be better done on the lens rather than in the camera. I hope Canon makes a thin mirrorless, but they could make some new WA lenses for it that could extend into the live-view mirror-box of EF DSLRs so everyone can use them.

 

[amorse]

Is allllll that trouble worth a measly 1/4-1/2"?

I need something sexier than that to sign up for all his trouble. Faster than ring USM on SLR focusing. A modern 50 prime to go on that new mount. Something.

- A

An electronically curved sensor would be sexy, especially if it can change its curve to suit the lens

The description on this literally says the curve would adapt to the needs of the lens, so it may not be limited to one lens design as previously thought.

 

[Adelino]

Why do so many people assume this is a do or die situation for Canon. It's entirely possible to have an EF mount now and wait to introduce a new mount later, one that will work with curved sensors or Medium Format sensor. Also these Rube Goldberg shifting sensor mechanisms don't sound so attractive especially since Canon are known for reliability, quality and performance. Keep it simple Canon, EF mount now sit back and plan something truly revolutionary for the next mount.

 

[Adelino]

Considering that the 5D series is a little too thin for my hands, I certainly wouldn't be interested in a thinner body than that.

Right! I certainly don't have large hands but I prefer the size of my 7D to my 6D.

 

[Bob Howland]

So what's the likelihood that Canon has already implemented the protocol switching in the M-series series cameras and never mentioned it? One thing that has always puzzled me is that M-series cameras have an extra pin with which to communicate with the lens.

 

[bks54]

I'm sceptical of this EF-depth body with R-mount lenses that sit deep in the body. This suggests the mount is offset part way up the lens barrel.
In addition, those R-mount lenses intended for this first-generation mirrorless will have an outer barrel thin enough to go through the EF mount which means that they will be narrower lens elements than native lenses.

As and when Canon go full-bore mirrorless in 5-10 years time and start to make bodies that are thinner and lighter, I presume they will have a full-on 54mm R mount and Canon will have to design a whole new set of R-mount lenses - and the lenses they buy now will not be compatible with the future mirrorless body.

Or am I missing something?

Many wide angle lenses have rear elements that are already much smaller in diameter than the 54 mm EF mount. And, non-retrofocus designs of the past almost always had small rear elements. Nice article here: https://www.bhphotovideo.com/explor...review-non-retrofocus-ultra-wide-angle-lenses

 

[Adelino]

Keeping the native registration distance of EF lenses sounds like a terrible idea to me. I sincerely hope Canon doesn't do that. It would mean needlessly fatter bodies, etc.

Not so needlessly, full compatibility with all those EF lenses.

 

[Mikehit]

Many wide angle lenses have rear elements that are already much smaller in diameter than the 54 mm EF mount. And, non-retrofocus designs of the past almost always had small rear elements. Nice article here: https://www.bhphotovideo.com/explor...review-non-retrofocus-ultra-wide-angle-lenses

That is wide-angle sorted, but what about standard to telephoto, or zooms?

 

[Adelino]

In an interview with the Canon team, one of the designers stated that the EOS was not designed for FF.

EOS not designed for FF? It was designed for 35mm film, the original FF.

 

[Mt Spokane Photography]

Forgive me, I always thought EF and F were sufficiently close on flange distance that adapting would be impossible. (Aren't they only a couple mm apart?)

- A

I have several F to EF adaptors, I keep them on each of my Nikon lenses. They have been available for many years. There are a very few that adapt Nikon lenses with no aperture rings, but they exist.

https://photographylife.com/reviews/novoflex-nikon-canon-lens-adapter

 

[BillB]

That is wide-angle sorted, but what about standard to telephoto, or zooms?

Longer than about 40mm, the EF mount works fine--no need to recess the lens.

 

[NWPhil]

I really hope that EF lenses can be mounted without too much fuss. For what I see with other FF mirror less systems (except Leica rf) the lenses are almost as big and as expensive....however legacy lenses will be left in the dark.sort of.
After all how much more is needed to make it work? Isn't live view sort of a temporary mirrorless already?

 

[bks54]

That is wide-angle sorted, but what about standard to telephoto, or zooms?

Standard and tele lenses don’t need elements close to the sensor. Wide zooms mostly have narrow rear elements.

 

[amorse]

I have several F to EF adaptors, I keep them on each of my Nikon lenses. They have been available for many years. There are a very few that adapt Nikon lenses with no aperture rings, but they exist.

https://photographylife.com/reviews/novoflex-nikon-canon-lens-adapter

Yes, but adapting an EF lens to F would mean the lens needs a reduced flange distance than the f-mount can provide (if I understand correctly). I would suspect that any adapter which allowed an EF lens to be mounted to F mount would need optics?

 

[3kramd5]

A moveable sensor doesn’t sound so bad to me. I’d think that the sensor would only move in a macro way with lens changes, I change lenses on a given body less than ten times a day even with busy shooting. If another lens is needed it usually goes on a separate body. So a moveable sensor perhaps on a rail, could be called upon with lens type changes. EOS R for railed sensor.

Canon already has patents for conformational sensor changes, so this is not a bridge too far.

It’s not a bridge too far to another patent. Production-worthy might be a long haul though.

 

[Mt Spokane Photography]

Such a system would probably require more than than 1/4 to 1/2 inch extension; more like 26 mm (more than an inch) since EF is at 44 mm and thin mirrorless much shorter (18 mm for EF-M). If they want to do something like this it would be better done on the lens rather than in the camera. I hope Canon makes a thin mirrorless, but they could make some new WA lenses for it that could extend into the live-view mirror-box of EF DSLRs so everyone can use them.

Really difficult and expensive to have a moving sensor and keep it parallel to 1/10000 in tolerances. Canon puts shims under each sensor to keep it parallel to the lens mount, its difficult. The shim thickness is marked on the sensor, but a check with a special tool is also a requirement.

 

[Kit.]

I'm assuming that because that's the most sensible design once the mirror is removed. A mirror box space with no mirror is very inelegant.

It is not only mirror that can go into that space. For example, EF mount videocameras keep their ND filter wheels there.

 

[3kramd5]

1) MLU + Liveview on an SLR does not allow you to hold the camera up to your eye unless you design a hybrid VF or use some comical eye loupe on the back LCD.

... or unless it’s given a second viewfinder (an electronic one, specifically). You made blue the pertinent caveat


Personally I don’t consider the hybrid VF to be the same capability as it doesn’t look through the lens in optical mode.

2) An SLR design -- even with MLU + liveview -- cannot adapt other people's lenses.

Sure it can. But to do so without optics, the lenses have to come from a larger format. The same restriction applies to mirrorless short flange cameras. The Nikon Z for example likely won’t be able to adapt Sony mirrorless lenses without adding optics. But a 35mm SLR can adapt for example medium format lenses.

But fair enough: they can adapt a smaller set of lenses. So there’s one thing. In terms of core functionality, however, I’m still coming up short:

You can give an SLR an EVF, IBIS, electronic shutter, sensor-based metering and AF, etc.
You can’t give a mirrorless TTL OVF (without removing the sensor from the path while composing); you can’t use dedicated and optimized sensors for AF and metering without a mirror.

While a dual VF may be a bit of a kludge, I think I’d take one rather than trade away the advantages the mirror brings.