Could the A or E mount support a medium-format sensor?

[Phyltre]

There is a rumor that a 50MP Fuji MF camera may be released in 2017 with a Sony sensor. If a Sony medium-format camera gets released this year on (presumably A) mount, that would make sense since Sony usually delays third-party cameras using their own sensors by a year. There are also several rumor sites dropping hints that new sensor tech is on the way from Sony, while Sony continues to tease that 2016 will be more revolutionary than 2015 was in terms of cameras released, and Hirai himself was conspicuously fondling Nikon DSLRs at CP+. No way that wasn't a PR hint.

That seems to support speculation that we will get new "professional" Sony models this year both in A and E mount--E to match the G Master lenses/update the A7__ line with A6300 advancements, and A mount because Sony has repeated themselves several times that A Mount is not dead but the A7__ family has come so far they've eaten up a lot of what would have been most of the A Mount business (of course I'm not implying they're equally useful in all the same situations), leaving the A Mount in search of new justifications beyond just "A7RII with faster AF" (although that would probably be enough for many!) I could certainly see Sony justifying A Mount by placing it above the A7__ family by virtue of medium-format going forward.

So my question is, could the A Mount theoretically support a medium-format sensor? If so, I suspect that will be its territory going forward. If not, maybe they have other plans.

 

[uhligfd]

Yep, the a-mount might support a medium format sensor, but all old or current a-mount lenses would give you just a round center part image on that sensor.

So what is the point of you asking this question? To use a medium format sensor, you, we, I would need to buy all new lenses if we want to cover the sensor with light and so why not go for a new mount anyway. Who would care about the actual mount?

 

[Phyltre]

I'm just curious what Sony's next moves are since they're claiming to stay full-steam in E Mount but still release new A Mount bodies. They wouldn't have released G Master lenses in E Mount if A Mount was going to stay their only professional lineup, in my humble opinion.

 

[JimKasson]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is. I suppose it would be possible to design a line of lenses that had exit pupils close enough to the sensor that they could illuminate a big sensor through such a small hole, but why would you want to saddle lens designers with that kind of handicap?

And forget about trying to use Hassy, Pentax, Rollei, or Contax MF lenses on such a camera. The mount would occlude the periphery of the sensor.

Jim

--
http://blog.kasson.com

 

[Graham Best]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

Yea, Sony really pushed the envelope.

Sony E mount throat diameter - 46.1mm

Leica M - 44mm

Nikon F - 44mm

Minolta SR/MC/MD - 44.97

Olympus OM - 46mm

Rolleiflex SL 35 - 46mm

 

[JimKasson]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

Yea, Sony really pushed the envelope.

Sony E mount throat diameter - 46.1mm

Leica M - 44mm

Nikon F - 44mm

Minolta SR/MC/MD - 44.97

Olympus OM - 46mm

Rolleiflex SL 35 - 46mm

But consider the FFD.

 

[Graham Best]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

Yea, Sony really pushed the envelope.

Sony E mount throat diameter - 46.1mm

Leica M - 44mm

Nikon F - 44mm

Minolta SR/MC/MD - 44.97

Olympus OM - 46mm

Rolleiflex SL 35 - 46mm

But consider the FFD.

The Flange Focal Distance may prevent certain 3rd part lenses from being adapted with optimal results (as many of the wide rangefinder lenses tested poorly), but that shouldn't suggest Sony made a design mistake.

The optical performance of Sony's FE lenses prove that lenses designed for the camera can be as good or better than any camera/lens, including camera designs with greater FFD's. That the great percentage of adapted lenses (including almost all retro-focus lenses) also perform well is a testament to Sony's design.

 

[JimKasson]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

Yea, Sony really pushed the envelope.

Sony E mount throat diameter - 46.1mm

Leica M - 44mm

Nikon F - 44mm

Minolta SR/MC/MD - 44.97

Olympus OM - 46mm

Rolleiflex SL 35 - 46mm

But consider the FFD.

The Flange Focal Distance may prevent certain 3rd part lenses from being adapted with optimal results (as many of the wide rangefinder lenses tested poorly), but that shouldn't suggest Sony made a design mistake.

I don't consider that the RF lenses not working so well has anything to do with the 18mm FFD. They don't work so well because they were designed for a stack thickness of zero or very small, and they have the exit pupil close to the focal plane.

Putting the exit pupil close to the focal plane actually makes it easier, not harder to illuminate the corners without occlusion.

The optical performance of Sony's FE lenses prove that lenses designed for the camera can be as good or better than any camera/lens, including camera designs with greater FFD's.

I am not arguing the point that you are making. What did I say that makes you think that I am questioning the performance of lenses designed for the E Mount used on FF sensors?

That the great percentage of adapted lenses (including almost all retro-focus lenses) also perform well is a testament to Sony's design.

I'm not saying that the lens mount is defective, only that there's no way it will handle even slightly larger sensors without lenses designed for the mount.

Jim

 

[120 to 35]

So my question is, could the A Mount theoretically support a medium-format sensor? If so, I suspect that will be its territory going forward. If not, maybe they have other plans.

The outside diameter of the Contax 645 medium format mount is around 80mm compared to E-mount at 60mm. The internal clearance is just less than 60mm. I recall the A-mount is somewhere in-between.

So while the A-mount might support a sensor slightly larger than full frame, it won't support the larger medium format sensor sizes used in digital backs. The Pentax sensor is not too large though.

 

[Graham Best]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

I am not arguing the point that you are making. What did I say that makes you think that I am questioning the performance of lenses designed for the E Mount used on FF sensors?

Forgive me, I obviously misinterpreted.

 

[JimKasson]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

I am not arguing the point that you are making. What did I say that makes you think that I am questioning the performance of lenses designed for the E Mount used on FF sensors?

Forgive me, I obviously misinterpreted.

Perhaps I should have made my point with numbers. The diagonal of a FF sensor is slightly over 43mm. The throat of the E-mount is slightly over 46mm. No occlusion for parallel ray bundles. No occlusion for diverging ray bundles with source apertures less than 46mm. Larger apertures may suffer occlusion, but that will be ameliorated by exit pupil distance being finite; the smaller the FFD, the less amelioration.

The diagonals of the Hassy and P1 big chip MF sensors are slightly over 67mm. We can calculate the distance from the sensor surface to a point source that just starts to occlude by solving the equation:

46/(x-18) = 67/x

x = 57.4mm

which puts quite a restriction on the exit pupil location of a lens.

That's a best case number. As the diameter of the source that the sensor sees grows from zero as assumed in the above calculation, the lens exit pupil needs to be even closer to the sensor.

Jim

--
http://blog.kasson.com

 

[JimKasson]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

I am not arguing the point that you are making. What did I say that makes you think that I am questioning the performance of lenses designed for the E Mount used on FF sensors?

Forgive me, I obviously misinterpreted.

Perhaps I should have made my point with numbers. The diagonal of a FF sensor is slightly over 43mm. The throat of the E-mount is slightly over 46mm. No occlusion for parallel ray bundles. No occlusion for diverging ray bundles with source apertures less than 46mm. Larger apertures may suffer occlusion, but that will be ameliorated by exit pupil distance being finite; the smaller the FFD, the less amelioration.

The diagonals of the Hassy and P1 big chip MF sensors are slightly over 67mm. We can calculate the distance from the sensor surface to a point source that just starts to occlude by solving the equation:

46/(x-18) = 67/x

x = 57.4mm

which puts quite a restriction on the exit pupil location of a lens.

That's a best case number. As the diameter of the source that the sensor sees grows from zero as assumed in the above calculation, the lens exit pupil needs to be even closer to the sensor.

To see how the FFD plays into this, run the above calculation with Nikon numbers:

44/(x-46.5) = 67/x

x = 135.5mm

Jim

 

[120 to 35]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

I am not arguing the point that you are making. What did I say that makes you think that I am questioning the performance of lenses designed for the E Mount used on FF sensors?

Forgive me, I obviously misinterpreted.

Perhaps I should have made my point with numbers. The diagonal of a FF sensor is slightly over 43mm. The throat of the E-mount is slightly over 46mm. No occlusion for parallel ray bundles. No occlusion for diverging ray bundles with source apertures less than 46mm. Larger apertures may suffer occlusion, but that will be ameliorated by exit pupil distance being finite; the smaller the FFD, the less amelioration.

The diagonals of the Hassy and P1 big chip MF sensors are slightly over 67mm. We can calculate the distance from the sensor surface to a point source that just starts to occlude by solving the equation:

46/(x-18) = 67/x

x = 57.4mm

which puts quite a restriction on the exit pupil location of a lens.

That's a best case number. As the diameter of the source that the sensor sees grows from zero as assumed in the above calculation, the lens exit pupil needs to be even closer to the sensor.

To see how the FFD plays into this, run the above calculation with Nikon numbers:

44/(x-46.5) = 67/x

x = 135.5mm

Jim

--
http://blog.kasson.com

It's not clear what these x distances actually allow.

The examples look reasonable. If you shine a beam of light from a point 57mm from the sensor, it should clear the opening and reach every corner of the large MF sensor in an E-mount camera.

Likewise if a Nikon mount camera with a large sensor was made, the same beam would clear the opening from 135mm (and closer). This is because the opening is closer to the source of the beam.

Now what does it mean for lens design?

 

[JimKasson]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

I am not arguing the point that you are making. What did I say that makes you think that I am questioning the performance of lenses designed for the E Mount used on FF sensors?

Forgive me, I obviously misinterpreted.

Perhaps I should have made my point with numbers. The diagonal of a FF sensor is slightly over 43mm. The throat of the E-mount is slightly over 46mm. No occlusion for parallel ray bundles. No occlusion for diverging ray bundles with source apertures less than 46mm. Larger apertures may suffer occlusion, but that will be ameliorated by exit pupil distance being finite; the smaller the FFD, the less amelioration.

The diagonals of the Hassy and P1 big chip MF sensors are slightly over 67mm. We can calculate the distance from the sensor surface to a point source that just starts to occlude by solving the equation:

46/(x-18) = 67/x

x = 57.4mm

which puts quite a restriction on the exit pupil location of a lens.

That's a best case number. As the diameter of the source that the sensor sees grows from zero as assumed in the above calculation, the lens exit pupil needs to be even closer to the sensor.

To see how the FFD plays into this, run the above calculation with Nikon numbers:

44/(x-46.5) = 67/x

x = 135.5mm

Jim

--
http://blog.kasson.com

It's not clear what these x distances actually allow.

The examples look reasonable. If you shine a beam of light from a point 57mm from the sensor, it should clear the opening and reach every corner of the large MF sensor in an E-mount camera.

Likewise if a Nikon mount camera with a large sensor was made, the same beam would clear the opening from 135mm (and closer). This is because the opening is closer to the source of the beam.

Now what does it mean for lens design?

It's a limit on the location of the exit pupil so that occlusion will not occur. Bigger numbers mean the exit pupil can be farther from the sensor. But note that these are best-case numbers, assuming a point source at the lens, which means it's stopped way way down. Things get worse (the exit puuil has to be closer) as the aperture gets wider.

My point was that it's not just the throat diameter (as was implied by a poster earlier in this thread), but that the FFD plays into the occlusion calculation, and the E-mount has a FFD that makes it tougher to make a larger sensor work with no occlusion than, say, an F-mount.

By the way, the E-mount best case calculation puts the farthest exit pupil distance closer to the sensor than the FFD of many MF cameras, including Mamiya 645, Pentax 645, Hassy H, Hassy V, which means that lenses for those cameras would have to have exit pupils behind the flange, or they'd occlude even at f/64.

Jim

--
http://blog.kasson.com

 

[120 to 35]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

I am not arguing the point that you are making. What did I say that makes you think that I am questioning the performance of lenses designed for the E Mount used on FF sensors?

Forgive me, I obviously misinterpreted.

Perhaps I should have made my point with numbers. The diagonal of a FF sensor is slightly over 43mm. The throat of the E-mount is slightly over 46mm. No occlusion for parallel ray bundles. No occlusion for diverging ray bundles with source apertures less than 46mm. Larger apertures may suffer occlusion, but that will be ameliorated by exit pupil distance being finite; the smaller the FFD, the less amelioration.

The diagonals of the Hassy and P1 big chip MF sensors are slightly over 67mm. We can calculate the distance from the sensor surface to a point source that just starts to occlude by solving the equation:

46/(x-18) = 67/x

x = 57.4mm

which puts quite a restriction on the exit pupil location of a lens.

That's a best case number. As the diameter of the source that the sensor sees grows from zero as assumed in the above calculation, the lens exit pupil needs to be even closer to the sensor.

To see how the FFD plays into this, run the above calculation with Nikon numbers:

44/(x-46.5) = 67/x

x = 135.5mm

Jim

--
http://blog.kasson.com

It's not clear what these x distances actually allow.

The examples look reasonable. If you shine a beam of light from a point 57mm from the sensor, it should clear the opening and reach every corner of the large MF sensor in an E-mount camera.

Likewise if a Nikon mount camera with a large sensor was made, the same beam would clear the opening from 135mm (and closer). This is because the opening is closer to the source of the beam.

Now what does it mean for lens design?

It's a limit on the location of the exit pupil so that occlusion will not occur. Bigger numbers mean the exit pupil can be farther from the sensor. But note that these are best-case numbers, assuming a point source at the lens, which means it's stopped way way down. Things get worse (the exit puuil has to be closer) as the aperture gets wider.

My point was that it's not just the throat diameter (as was implied by a poster earlier in this thread), but that the FFD plays into the occlusion calculation, and the E-mount has a FFD that makes it tougher to make a larger sensor work with no occlusion than, say, an F-mount.

By the way, the E-mount best case calculation puts the farthest exit pupil distance closer to the sensor than the FFD of many MF cameras, including Mamiya 645, Pentax 645, Hassy H, Hassy V, which means that lenses for those cameras would have to have exit pupils behind the flange, or they'd occlude even at f/64.

Jim

--
http://blog.kasson.com

Thanks for the explanation. So the real requirements is a mount opening that is larger than the maximum aperture opening of the lens, which can be as large as 60mm for an f:1.4/85mm lens, as well as the mount being wider than the cone of light extending from the lens to the sensor.

To put it simply, the 18mm flange distance of E-mount might allow a sensor about 2mm larger than full frame, while the A-mount, being both wider and further from the sensor, might allow a larger sensor.

 

[JimKasson]

Dunno about he A mount, but the E mount can barely handle a FF sensor as it is.

I am not arguing the point that you are making. What did I say that makes you think that I am questioning the performance of lenses designed for the E Mount used on FF sensors?

Forgive me, I obviously misinterpreted.

Perhaps I should have made my point with numbers. The diagonal of a FF sensor is slightly over 43mm. The throat of the E-mount is slightly over 46mm. No occlusion for parallel ray bundles. No occlusion for diverging ray bundles with source apertures less than 46mm. Larger apertures may suffer occlusion, but that will be ameliorated by exit pupil distance being finite; the smaller the FFD, the less amelioration.

The diagonals of the Hassy and P1 big chip MF sensors are slightly over 67mm. We can calculate the distance from the sensor surface to a point source that just starts to occlude by solving the equation:

46/(x-18) = 67/x

x = 57.4mm

which puts quite a restriction on the exit pupil location of a lens.

That's a best case number. As the diameter of the source that the sensor sees grows from zero as assumed in the above calculation, the lens exit pupil needs to be even closer to the sensor.

To see how the FFD plays into this, run the above calculation with Nikon numbers:

44/(x-46.5) = 67/x

x = 135.5mm

Jim

 

[meanwhile]

Given lenses are round (ignoring baffles for a minute), and mounts are round, could they go square*? Or even 3:4 instead of 2:3?

Makes a smaller camera more practical if you don't need to go vertical to shoot portrait.

* I'm presuming that the costs involved in creating a round sensor are prohibitive

 

[JimKasson]

Given lenses are round (ignoring baffles for a minute), and mounts are round, could they go square? Or even 3:4 instead of 2:3? Makes a smaller camera more practical if you don't need to go vertical to shoot portrait.

Sure. I'd like that, being a long time Hassy shooter. But I bet it would be a tough sell to many.

Jim

 

[meanwhile]

But I bet it would be a tough sell to many.

Hmm, why a tough sell? Because of the extra cost that would likely be involved, or the practical applications of it?

I've been wanting more options in the Sony aspect ratio options for a long time (3:2 & 16:9 being the only ones currently). The options that some of the m43 cameras have with an oversized sensor (Panasonic? Or was it the LX cameras, rather than m43) are great.

 

[ProfHankD]

There is a rumor that a 50MP Fuji MF camera may be released in 2017 with a Sony sensor. ...

So my question is, could the A Mount theoretically support a medium-format sensor?

The basic answer is that the 50MP "medium format" sensor Sony made was sized so that it actually could give a variety of useful crops within the normal image circle of a designed-for-FF lens. I suspect that's why it's rather seriously undersized as medium-format goes... but this was all discussed here back when Sony first announced their 50MP sensor.

As for true medium format sensors, well, it's always possible, but both E and A would impose quite a few constraints that would make lenses harder to make (especially fast lenses) and would prevent use of many older medium-format lenses. It would make more sense for Sony to just come out with a larger mount that can swallow an E adapter....