near future lens size and resolution

[darth mollusk]

My interest is primarily with wildlife photography and in this thread I want to explore the near future (the next decade - if you can consider that near) of telephoto lens inovation. Over the past 2 or 3 decades telephoto lens size appears to have changed little - is this going to change over the next 10 to 20 years?

1. What specifically restricts lens size with existing DSLR technology? Does the distance from the sensor to the lens make enough of a difference that future telephoto lenses on mirrorless systems will be substantially smaller? (forgive me for assuming that mirrorless cameras will replace DSLR)

2. Ten years from now, when you're shooting with a 100mp mirrorless camera body, will it still be an advantage to shoot with a massive 600mm f4 lens, or will cropping to a 20mp image with a significantly smaller lens eliminate the need for a super-telephoto? There are certainly differences that will be difficult to overcome, such as tracking wildlife from a distance and accurately focusing at that distance with a smaller mm lens. Do you think these challenges will be met?

In a nutshell, 10 years from now (when I finally have 10k to spend on a lens) will I still want to purchase a 600mm 5kg lens for optimal image quality?

 

[SwissBear]

my thoughts for this matter:

1) the sensor-flange distance is only a problem for UWA's, not teles. To get smaller lenses, only DO helps significantly, where advanced tech might reduce the flaws.

2) If you get 100MP, then on rather large sensors (fullframe and maybe crop) due to physical limitations. But as FF sensors, or even bigger sensor will always exist because of DOF, the size of the big whites will not change, but as we've seen before, the next round might be again a bit ligther. If a lens is built for a smaller sensor, then it is of course smaller or faster.
Although, higher megapixels also have some draw backs: diffraction, noise, filesize and others, where the first can only be countered with better (and therefore bigger&heavier) lenses with build tolerances so small that they cost again a fortune.

My prediction goes only as far as the next (real) generation, where i hope to see a clumsy big megapixel studio body with superb IQ at native ISO but nothing much beyond, and the 1DXII 5DV, 6DIII will have a slight increase in MP count, but some 2 stops more DR&ISO.

 

[Knut Skywalker]

Maybe in the future when we have ultra-high megapixel counts, they adopt the methods of the Nokia Pureview. Slap a super high-resolving sensor in the device and zoom in with cropping the sensor.

 

[ecka]

I think that in the future things will get much simpler. Super-tele lenses for wild life could be replaced by any small lens + a flying invisibility suit (or that + downloading images directly from your brain). ;D

 

[Hannes]

Maybe in the future when we have ultra-high megapixel counts, they adopt the methods of the Nokia Pureview. Slap a super high-resolving sensor in the device and zoom in with cropping the sensor.

I doubt it, there's a limit to what we can coax out a sensor before we run into pesky things like the limits of physics. Digital zoom has never been a good idea and I very much doubt it'll change in the future.

 

[darth mollusk]

my thoughts for this matter:

1) the sensor-flange distance is only a problem for UWA's, not teles. To get smaller lenses, only DO helps significantly, where advanced tech might reduce the flaws.

2) If you get 100MP, then on rather large sensors (fullframe and maybe crop) due to physical limitations. But as FF sensors, or even bigger sensor will always exist because of DOF, the size of the big whites will not change, but as we've seen before, the next round might be again a bit ligther. If a lens is built for a smaller sensor, then it is of course smaller or faster.
Although, higher megapixels also have some draw backs: diffraction, noise, filesize and others, where the first can only be countered with better (and therefore bigger&heavier) lenses with build tolerances so small that they cost again a fortune.

My prediction goes only as far as the next (real) generation, where i hope to see a clumsy big megapixel studio body with superb IQ at native ISO but nothing much beyond, and the 1DXII 5DV, 6DIII will have a slight increase in MP count, but some 2 stops more DR&ISO.

Thanks SB, your thoughts are pretty close to what I have suspected - the physical size of super-tele's isn't going to change anytime in the foreseeable future. Are you aware of any articles that explore the physics of these limitations in a little more detail? Just want to wrap my head around how and why megapixels and sensor size affect the amount of glass required to achieve optimal image quality.

 

[9VIII]

My interest is primarily with wildlife photography and in this thread I want to explore the near future (the next decade - if you can consider that near) of telephoto lens inovation. Over the past 2 or 3 decades telephoto lens size appears to have changed little - is this going to change over the next 10 to 20 years?

1. What specifically restricts lens size with existing DSLR technology? Does the distance from the sensor to the lens make enough of a difference that future telephoto lenses on mirrorless systems will be substantially smaller? (forgive me for assuming that mirrorless cameras will replace DSLR)

2. Ten years from now, when you're shooting with a 100mp mirrorless camera body, will it still be an advantage to shoot with a massive 600mm f4 lens, or will cropping to a 20mp image with a significantly smaller lens eliminate the need for a super-telephoto? There are certainly differences that will be difficult to overcome, such as tracking wildlife from a distance and accurately focusing at that distance with a smaller mm lens. Do you think these challenges will be met?

In a nutshell, 10 years from now (when I finally have 10k to spend on a lens) will I still want to purchase a 600mm 5kg lens for optimal image quality?

1. As mentioned, It makes a big difference in wide angle lenses, once the flange distance is greater than the focal length the lens needs to use a retrofocus design, which adds a lot of complexity.
(I have little doubt, not that it counts for much, that SLR will end up a niche product compared to these newfangled thing cameras (NTC?) http://www.2cameraguys.com/mirrorless-milc-evil-csc-cameras.htm).

2. I actually think there's potential here for improvements (relatively speaking). If you take current Micro Four Thirds or Nikon 1 series cameras and stick a really sharp 300mm f4 lens in front of them, you might be able to get similar results to a large lens at a fraction of the cost (and weight).
You only need a 26mp Micro Four Thirds sensor to equal 100MP Full Frame, and the latest Nikon 1 series already have the same pixel density as FF@100MP. If you want a 100 megapixel full frame sensor just to crop it for distant subjects, then in effect we already have the same thing. The only real trouble is lens selection.

 

[Mt Spokane Photography]

My interest is primarily with wildlife photography and in this thread I want to explore the near future (the next decade - if you can consider that near) of telephoto lens inovation. Over the past 2 or 3 decades telephoto lens size appears to have changed little - is this going to change over the next 10 to 20 years?

1. What specifically restricts lens size with existing DSLR technology? Does the distance from the sensor to the lens make enough of a difference that future telephoto lenses on mirrorless systems will be substantially smaller? (forgive me for assuming that mirrorless cameras will replace DSLR)

No

2. Ten years from now, when you're shooting with a 100mp mirrorless camera body, will it still be an advantage to shoot with a massive 600mm f4 lens, or will cropping to a 20mp image with a significantly smaller lens eliminate the need for a super-telephoto? There are certainly differences that will be difficult to overcome, such as tracking wildlife from a distance and accurately focusing at that distance with a smaller mm lens. Do you think these challenges will be met?

Canon has filed a number of patents over the past few years which have the effect of reducing the length of a lens, but not the diameter. So, lenses are going to be shorter.

In a nutshell, 10 years from now (when I finally have 10k to spend on a lens) will I still want to purchase a 600mm 5kg lens for optimal image quality?

In 10 years you may need 25K to buy a new lens. Prices are doubleing with each new generation of lens. Many of the used big whites sell for more than they cost new.

The size of a lens is based on the focal length, aperture, and size of the image circle needed to cover a sensor. For the long telephotos, the aperture and focal length determine the diameter, so the sensor size has no affect.

 

[TrumpetPower!]

Yes and no.

In terms of sharpness, resolution, enlargement size, contrast, that sort of thing, there's still a fair (but not huge) amount of room before we start hitting hard physical limits. There's probably not much point in much more than 80-100 megapickles in a 135 format ("full frame") camera. Much past that and you're just rendering diffraction that much more faithfully. Few, if any, lenses can match that resolution, but there's no reason other than money and R&D why they couldn't.

Translate that into a smaller format, and I could imagine an hypothetical future micro 4/3 camera with a 200 mm f/4 "supertelephoto" the same size as the 85 f/1.8 that would have comparable (or even better) image quality to a 5DIII / 1DX with a 400 f/2.8 today.

But.

With one huge caveat.

The depth of field / background blur (not the same thing, but close enough for this discussion) from such a combination shot wide open is going to be comparable to a 400 f/2.8...shot at f/16 or so. Which rather renders moot one of the two reasons for a lens that fast that long. (The other is its ability to gather lots of light, enabling fast shutter speeds at low ISOs in dim conditions, but we'll assume that high ISO performance in the future eliminates this as a practical differentiator.)

Over here:

http://www.canonrumors.com/forum/index.php?topic=13837.msg250058#msg250058

I posted a comparison of a 400 f/2.8 and a 70-200 f/2.8 @ 200 cropped to the same field of view as the 400. That would basically be the shot you'd get with a 200 f/2.8 on m4/3. Now, imagine stopping down to f/4, how much more depth of field you'd have, and there you go.

You'll also notice that the crop, though noticeably inferior, still holds up remarkably well. It's not hard to imagine future smaller and lighter cameras doing even better.

But none of them will ever completely blow out the background the way a Great White can; that's just physics (and mostly simple geometry).

A 400 f/2.8 will always need a 143mm physical / virtual aperture, which means it's always going to have at least a 5 5/8" front element. You can use exotic materials to lighten the weight. You can probably use some other exotic materials and fancy design to shorten it somewhat. Diffractive optics, for example, holds real (as-yet-unrealized) promise for both. But you're always going to have that 6" lens cap to deal with, no matter what.

Cheers,

b&

 

[darth mollusk]

Yes and no.

In terms of sharpness, resolution, enlargement size, contrast, that sort of thing, there's still a fair (but not huge) amount of room before we start hitting hard physical limits. There's probably not much point in much more than 80-100 megapickles in a 135 format ("full frame") camera. Much past that and you're just rendering diffraction that much more faithfully. Few, if any, lenses can match that resolution, but there's no reason other than money and R&D why they couldn't.

Translate that into a smaller format, and I could imagine an hypothetical future micro 4/3 camera with a 200 mm f/4 "supertelephoto" the same size as the 85 f/1.8 that would have comparable (or even better) image quality to a 5DIII / 1DX with a 400 f/2.8 today.

But.

With one huge caveat.

The depth of field / background blur (not the same thing, but close enough for this discussion) from such a combination shot wide open is going to be comparable to a 400 f/2.8...shot at f/16 or so. Which rather renders moot one of the two reasons for a lens that fast that long. (The other is its ability to gather lots of light, enabling fast shutter speeds at low ISOs in dim conditions, but we'll assume that high ISO performance in the future eliminates this as a practical differentiator.)

Over here:

http://www.canonrumors.com/forum/index.php?topic=13837.msg250058#msg250058

I posted a comparison of a 400 f/2.8 and a 70-200 f/2.8 @ 200 cropped to the same field of view as the 400. That would basically be the shot you'd get with a 200 f/2.8 on m4/3. Now, imagine stopping down to f/4, how much more depth of field you'd have, and there you go.

You'll also notice that the crop, though noticeably inferior, still holds up remarkably well. It's not hard to imagine future smaller and lighter cameras doing even better.

But none of them will ever completely blow out the background the way a Great White can; that's just physics (and mostly simple geometry).

A 400 f/2.8 will always need a 143mm physical / virtual aperture, which means it's always going to have at least a 5 5/8" front element. You can use exotic materials to lighten the weight. You can probably use some other exotic materials and fancy design to shorten it somewhat. Diffractive optics, for example, holds real (as-yet-unrealized) promise for both. But you're always going to have that 6" lens cap to deal with, no matter what.

Cheers,

b&

Thanks TP - that was clearly written and exactly the kind of explanation I was looking for, kudos for taking the time to spell it out. Understanding the limitations of the technology currently available (along with quenching curiosity) makes it a little easier to draft a probable roadmap for my future gear investments. The 400 2.8ii is on my radar and 10K for an entusiast on a biologist salary is a substantial investment - knowing that its dimensions and some of the key advantages are constrained by simple physics helps.

 

[CarlTN]

My interest is primarily with wildlife photography and in this thread I want to explore the near future (the next decade - if you can consider that near) of telephoto lens inovation. Over the past 2 or 3 decades telephoto lens size appears to have changed little - is this going to change over the next 10 to 20 years?

1. What specifically restricts lens size with existing DSLR technology? Does the distance from the sensor to the lens make enough of a difference that future telephoto lenses on mirrorless systems will be substantially smaller? (forgive me for assuming that mirrorless cameras will replace DSLR)

2. Ten years from now, when you're shooting with a 100mp mirrorless camera body, will it still be an advantage to shoot with a massive 600mm f4 lens, or will cropping to a 20mp image with a significantly smaller lens eliminate the need for a super-telephoto? There are certainly differences that will be difficult to overcome, such as tracking wildlife from a distance and accurately focusing at that distance with a smaller mm lens. Do you think these challenges will be met?

In a nutshell, 10 years from now (when I finally have 10k to spend on a lens) will I still want to purchase a 600mm 5kg lens for optimal image quality?

Don't wait 10 years, just buy a 6D or 5D3 now, and a 300 f/4L, and a 1.4x TC. You will find that you can't shoot anything as well with any smaller sensor'd camera. If you need a lot of reach in bright light, try an SX50...1200mm equivalent focal length. You can wait until prices on it fall to below $300...which will probably happen before Christmas.

So there you go, a 6D + 300 f/4 + a good used 1.4x ii = approximately $3550...a lot less than $10,000!

10 years from now, sensors in the 36mm width, will still be widely used. I can promise you that. I'm not sure "crop sensors" will be as widely used...

And I agree, 10 years from now, a 600mm f/4, will cost a lot more than $10,000. If inflation skyrockets, and depending on demand...such a lens could cost $80k by 2023. Of course in today's dollars, it would probably be equal to $16k or so.

 

[bycostello]

as long as they are glass based can't see lenses changing