[soalle]

Whenever I read comments like this, I have to respond ...

A digital revolution is nice. But it won't beat the laws of physics. At least, there are no signs whatsoever this will happen any time soon

Read about quantum efficiency and photon shot noise. And you'll see that current dSLRs are deceivingly close to the physically possible. Another stop in quantum efficiency, another stop now lost in color filters and maybe another stop in black noise. And that it is. Only exception is dynamic range which can be made arbitrarily large by adding a digital frame buffer into the sensor. Needing more? Use bigger lenses. Better images need more photons. As simple as this. Really. Look at owls

Of course, you can play with stacking images in camera, aligning and removing movable parts etc. All possible in the digital domain. Already possible now in post processing and I use it. I know the limitations. It helps in some cases (mostly where a tripod would do too) but won't help in others (like in action photography). So, in general, by 2020, things won't be much different from now.

And "post-shot selective DOF" (read Lytro) is another take at raping off venture capital firms. Lytro's Ng can't beat Heisenberg's uncertainty relation (not even a Stanford degree can help here ). So, read optics textbooks and understand why a plenoptics light field camera only produces low resolution (sub MP) images. German Raytrix does a bit better (actually now owning all the patents) but use a kind of prefocus, a suitable compromise between fully refocussable and fuly prefocussed. This achieves resolution suitable for movies but not still images.

god has spoken... so please silence!

[Unsinkable II]

Whenever I read comments like this, I have to respond ...

A digital revolution is nice. But it won't beat the laws of physics. At least, there are no signs whatsoever this will happen any time soon

Read about quantum efficiency and photon shot noise. And you'll see that current dSLRs are deceivingly close to the physically possible. Another stop in quantum efficiency, another stop now lost in color filters and maybe another stop in black noise. And that it is. Only exception is dynamic range which can be made arbitrarily large by adding a digital frame buffer into the sensor. Needing more? Use bigger lenses. Better images need more photons. As simple as this. Really. Look at owls

Of course, you can play with stacking images in camera, aligning and removing movable parts etc. All possible in the digital domain. Already possible now in post processing and I use it. I know the limitations. It helps in some cases (mostly where a tripod would do too) but won't help in others (like in action photography). So, in general, by 2020, things won't be much different from now.

And "post-shot selective DOF" (read Lytro) is another take at raping off venture capital firms. Lytro's Ng can't beat Heisenberg's uncertainty relation (not even a Stanford degree can help here ). So, read optics textbooks and understand why a plenoptics light field camera only produces low resolution (sub MP) images. German Raytrix does a bit better (actually now owning all the patents) but use a kind of prefocus, a suitable compromise between fully refocussable and fuly prefocussed. This achieves resolution suitable for movies but not still images.

Interesting points regarding sensor efficiencies. 3 stops isn't as much as probably many, including myself, hoped, but it's not bad.

I certainly take your point regarding owls! However, the breakthroughs will come in electronics, not optics. I can't begin to imagine what those breakthroughs will be, but they will come.

[Class A]

god has spoken... so please silence!

If god has spoken then ~13.7 billion years ago.

No need for silence, but some talk makes more sense than other.

[falconeye]

However, the breakthroughs will come in electronics, not optics.

That was my point exactly.

[devorama]

There's a new rumor now that the big sister of the Q is coming soon.

Mirrorless Rumors | Blog | A biggers sisther of the Pentax Q coming soon?

I would love to see a mirroless Pentax that would still accept k-mount lenses without an adapter. The newer mirrorless specific lenses could put the bayonet mount near the front of the lens instead of the back. So it would have most of the lens elements behind the bayonet. Although I guess this would limit the diameter of the lens elements that are close to the sensor. They could alter the mount just enough that the lens would not be able to fit inside the standard k-mount bayonet, so one couldn't accidentally try to insert one in a dslr and damage their mirror. I know someone mentioned an idea like this before. Sounds like an amazing idea to me.

[Caat]

There's a new rumor now that the big sister of the Q is coming soon.

Mirrorless Rumors | Blog | A biggers sisther of the Pentax Q coming soon?

I would love to see a mirroless Pentax that would still accept k-mount lenses without an adapter. The newer mirrorless specific lenses could put the bayonet mount near the front of the lens instead of the back. So it would have most of the lens elements behind the bayonet. Although I guess this would limit the diameter of the lens elements that are close to the sensor. They could alter the mount just enough that the lens would not be able to fit inside the standard k-mount bayonet, so one couldn't accidentally try to insert one in a dslr and damage their mirror. I know someone mentioned an idea like this before. Sounds like an amazing idea to me.

There are several issues with this idea:

1). You've already mentioned this: placing restrictions on the maximum diameter of the rear, and possibly middle, elements.

2). The mount would only be compatible with K mount lenses - you couldn't adapt other brands because the registration distance would be the same as a K-mount.

Whilst you are mainting backwards compatibility with existing K-mount lenses by keeping the same registration distance you are effectively binding the new lenses to a 40 year-old mount. This is OK if you are going to continue to produce DSLRs and therefore K-AF lenses but if Pentax stops making them then the USP of their new K mount is no longer relevant (except from a legacy users perspective). The system with a new lens attached is only going to be slightly shorter front to back than with an original K mount lens attached.

You're comprimising the mount purely for backwards compatibility. I'm not convinced this is a good idea.

I say comprimising the mount because new lenses for the mount are bound to be more complicated to design. This added complexity brings no inherent benefit to the lens itself.

To my mind Pentax either have to stick with the K-AF mount as is, remove the mirror and work on shrinking all other parts of the system (bar the registration distance) OR develop a brand new mount. In my opinion the latter is the most sensible and viable option.

[ghelary]

There are several issues with this idea:

1). You've already mentioned this: placing restrictions on the maximum diameter of the rear, and possibly middle, elements.

Yeah, so that means that we wouldn't have non-retrofocus 35mm/f1.2 lenses...

The advantage of having elements close to the sensor is for making wide angle lenses with non-retrofocus compatc design. If we are ready to through out compactness for large aperture, then retrofocus would just do it right.

To me this issue is non signifcant for the concept, which is about adding compactness to the system.

2). The mount would only be compatible with K mount lenses - you couldn't adapt other brands because the registration distance would be the same as a K-mount.

Whilst you are mainting backwards compatibility with existing K-mount lenses by keeping the same registration distance you are effectively binding the new lenses to a 40 year-old mount. This is OK if you are going to continue to produce DSLRs and therefore K-AF lenses but if Pentax stops making them then the USP of their new K mount is no longer relevant (except from a legacy users perspective). The system with a new lens attached is only going to be slightly shorter front to back than with an original K mount lens attached.

You're comprimising the mount purely for backwards compatibility. I'm not convinced this is a good idea.

I say comprimising the mount because new lenses for the mount are bound to be more complicated to design. This added complexity brings no inherent benefit to the lens itself.

To my mind Pentax either have to stick with the K-AF mount as is, remove the mirror and work on shrinking all other parts of the system (bar the registration distance) OR develop a brand new mount. In my opinion the latter is the most sensible and viable option.

There are already Next and m4/3 camera for the purpose of adpating every mount. And the new Next 7 seems like the perfect camera to do the job.

With an evolved K-mount, you can still use M42 and K lenses, which makes already a very broad choice in weirdo lenses.

Oh and by the way, the F mount is 54 year old, have seen the introduction of electronics, 2 versions of AF and different sensor formats, and Nikon users seem quite happy with it.

The K-mount have shown as well as much versatility, the MX have shown that making compact is not a problem with the mount, but with all the gitzmos around, AF motor, sensors, LCD screens, CPUs, batteries... All of those will eventually become much more compact shrinking the camera around the image sensor.

[falconeye]

Sounds like an amazing idea to me.

The idea is amazing.
But I agree that in the end the disadvantages will dominate (I won't repeat them all now).

An aspect I'd like to mention though is that the K mount not only is an SLR mount (with too large a registration distance) but also a 35mm FF mount (with too large a diameter).

I know and agree that other mirrorless mounts (NEX, NX, even µFT) have the same big diameter. But that doesn't mean that Pentax has to do the same mistake. Esp. as they have to differentiate (I know, a mount too large allows for a mirrorless FF down the road -- but personally, I have no probs to have Pentax then use the K mount or a derivative of it).

Pentax has the opportunity to restart APSC mirrorless from scratch, with a mount allowing to deliver on the promise of mirrorless: small. The Auto110 showed how small Pentax once was able to make FourThird lenses (because this is what they were). So, imagine the same principles applied to a Pentax APSC mirrorless. No space for the K mount left

[Caat]

Yeah, so that means that we wouldn't have non-retrofocus 35mm/f1.2 lenses...

The advantage of having elements close to the sensor is for making wide angle lenses with non-retrofocus compatc design. If we are ready to through out compactness for large aperture, then retrofocus would just do it right.

To me this issue is non signifcant for the concept, which is about adding compactness to the system.



There are already Next and m4/3 camera for the purpose of adpating every mount. And the new Next 7 seems like the perfect camera to do the job.

With an evolved K-mount, you can still use M42 and K lenses, which makes already a very broad choice in weirdo lenses.

Oh and by the way, the F mount is 54 year old, have seen the introduction of electronics, 2 versions of AF and different sensor formats, and Nikon users seem quite happy with it.

The K-mount have shown as well as much versatility, the MX have shown that making compact is not a problem with the mount, but with all the gitzmos around, AF motor, sensors, LCD screens, CPUs, batteries... All of those will eventually become much more compact shrinking the camera around the image sensor.

I didn't mean the K-mount is anachronistic. For a DSLR it's a modern mount and like the F mount has evolved from a fully manual mount to a largely electronic one.

I doubt very much that a mirroless Nikon mount camera will be an evolution of the F mount.

[gazonk]

Pentax has the opportunity to restart APSC mirrorless from scratch, with a mount allowing to deliver on the promise of mirrorless: small.

My thoughts, too. Shrinking the K-mount by the APS-C crop factor should give us a mount 2/3 the diameter of the K-mount. But that's just about the size of the Q mount, isn't it? I wonder why it's that big.