A New 50mm Lens is Being Tested in the Wild [CR2]

[hendrik-sg]

'upgrade' the exact same optical formula

I think, if it would be that easy, all manufacturers would produce such lenses. Without knowing lens design in Detail, we can not judge on what is possible and what not. There are so many different abberations, which must be optimized against each other.

fact is, over all brands, at 50mm there are cheap and averagely performing gauss lenses, better (and bigger) designs are very rare.

There was one interesting interview with a Canon technician, about the developpment of the 85mm IS lens, which was some kind of creative process to find a solution with IS. Compromise is, it's not the best abailable lens in this range.

other example for me is the 16-35 iii lens, which sems to be fantastically sharp, but with terrible vignetting. Only in the very center of the frame it's brighter than the F4 IS version. It's plausible, that a lens can be sharer, if the corner light is vignetted away, and can no longer disturb center sharpness. (this as well is a unscientific and maybe wrong Imagination)

 

[Bennymiata]

Might be a 50mm 1.8 for M's.
Good length for a crop sensor and it wouldn't be hard to convert the current 50 1.8 for use on an M.

 

[ahsanford]

other example for me is the 16-35 iii lens, which sems to be fantastically sharp, but with terrible vignetting. Only in the very center of the frame it's brighter than the F4 IS version. It's plausible, that a lens can be sharer, if the corner light is vignetted away, and can no longer disturb center sharpness. (this as well is a unscientific and maybe wrong Imagination)

That may be more of a design tradeoff than anything else. 16-35 2.8 zooms are expected to have front filtering in a non-absurd / somewhat standard diameter. Other UWA zooms that don't feel the need to satisfy that can opt for larger, more bulbous front elements. The Tamron 15-30 2.8 foregoes the front filterability, opts for a large + bulbous front element and has half the vignetting levels as the 16-35 f/2.8L III.

- A

 

[jolyonralph]

Could you imagine if Canon did something batsh-- crazy and make a slow L prime, say a relatively tiny 50 f/1.8L or f/2L? Folks' heads would explode. 'Cuz aperture is better, yo. :

I'd much rather have an f/1.8 lens that was sharp wide open than an f/1.4 lens that you need to stop down to f/2 to be of any use

The Sony FE 55mm f/1.8 is a beautiful lens.

 

[RayValdez360]

I love how Canon upgrades and replaces every lens but their MOST f'ed up lens ever.

 

[melgross]

Companies look at the potential market size for a product before coming out with it. Price is in the equation. Don’t sell too many 600mm, but at the price it’s profitable, and keeps these customers within Canon.

So, realistically, what is the biggest customer for a new 50 1.4? Is it the pro? The semi-pro, or the average customer who just wants a 50 1.4?

If it’s the pro, then very high IQ is required, and competitively, that means a big, heavy lens. It may mean IS, but maybe not.

For the semi pro, it needs excellent IQ, but maybe not quite as high, and a somewhat lower price. IS is likely more of a necessity.

For the average user, price and size is more important, as long as performance is good. IS adds too much cost, and adds to the size and weight, which the average user is more sensitive to.

So... if Canon is coming out with a new 50, what market are they aiming it at?

I’m not even thinking about a 1.2.

 

[ahsanford]

So... if Canon is coming out with a new 50, what market are they aiming it at?

I’m not even thinking about a 1.2.

Sure, we get why the 50 f/1.4 doesn't get much priority, but under a project by project review sort of rationale, a lens can actually sit without update for 25 years. Eventually, you have to do one of the following things:

• Update it --> EF 50 f/1.4 IS USM or EF 50 f/1.4 USM II. Do it as part of a 'middle-level' refresh series alongside the rest of that USM line: 20 f/2.8, 28 f/1.8, 85 f/1.8, 100 f/2, etc.
  
  
• Put out a different product that captures its userbase and possibly addresses some new unmet needs --> EF 50 f/2 STM 1:2 illuminated macro perhaps.
  
  
• Update the lower price point with nicer features that draw some of the old product's folks in --> an affordable 50 f/1.8 IS STM (or possibly Nano USM).
  
  
• Update the higher price point option to be more affordable (unlikely), or create a second premium price point under the top dog --> EF 50 f/1.4L IS USM, which is surely possible.
  
  
• Obsolete the product and funnel people to the other existing 50mm options. Doubt this will happen. That takes a small profit product off of Canon's rolls and surely just hands some money to the Sigmas and Tamrons of the world. That's not Canon's MO unless the product is a huge PITA to build in spec, creates portfolio support problems over time, etc. Consider: if they haven't obsoleted the EF 20 f/2.8, 28 f/1.8, 100 f/2, etc., they won't do it here.

- A

 

[ahsanford]

But, to answer your question, Mel, no one knows.

I think a slightly underwhelming 50 f/1.4 USM II (with Nano) sliding in around $499 would make sense, tent up the price in this segment, and put a non-Yongnuo-able product on the board. It's the logical, simple, rational thing for Canon to do -- both at 50 and with the 85 f/1.8 USM refresh. Hopefully that would be a new optical design, but who knows these days?

I think an EF 50 f/1.4L IS USM around makes a ton of sense based on the recent 85L IS, but I'm not sure it would be cheaper or lower 'prestige' than the 50 f/1.2L @ $1299. Canon may want to put out a skull-splittingly sharp retrofocal design like the Sigma 50 Art and sell it like the 35L II, say $1699 or so. So the future of the 50L space in general potentially overlaps / clouds the call on the non-L slot -- if a 50 f/1.4L IS arrives, I'm not expecting a new mid-level 50 to be offered for years afterwards.

But of course, I'd love the fancy non-L: New optical design, Ring USM, internal focusing, IS, double gauss / not huge, etc. Canon would ask $799-ish for that based on the 24/28/35 IS lenses, though I recognize that's a mint for a non-L / non-DO product currently sitting at a $329 price point. So I'm not confident that the lens I want is going to happen. We'll see.

- A

 

[slclick]

Time and time again, from pundits to optical experts, folks are saying double gauss only can get so good and you have to go retrofocus to obtain the IQ corner to corner without abberations and therefore that is why Sigma,Tamron and Zeiss have gone that route. Why would people think Canon won't do a retrofocus over 35mm? Just because they personally want a lighter smaller lens? I can't help but think it will be a pickle jar for sharpness, distortion and room for modern IS and AF internals.

Please tell me how I am wrong if you think so and how DG could be improved.

 

[ahsanford]

Time and time again, from pundits to optical experts, folks are saying double gauss only can get so good and you have to go retrofocus to obtain the IQ corner to corner without abberations and therefore that is why Sigma,Tamron and Zeiss have gone that route. Why would people think Canon won't do a retrofocus over 35mm? Just because they personally want a lighter smaller lens? I can't help but think it will be a pickle jar for sharpness, distortion and room for modern IS and AF internals.

Please tell me how I am wrong if you think so and how DG could be improved.

I always wondered if you could just supersize the elements diametrically with a DG design to get better results from the center of that glass (i.e. the FF sensor needed portions of it). In other words: oversize the elements like it was a medium format image circle so the FF output would be more central on the glass. In theory, such a lens would stay short front to back but step up in diameter (and presumably weight) considerably. (I wonder if the current 50L has done just that -- it's a 72mm front element while a 50 1.2 doesn't need anything so large for basic light gathering.)

But I'm no optical whiz and can't comment on if/how that might work. This idea could be nonsense.

- A

 

[stevelee]

I wish you guys who are pining for a better 50mm lens all the best, but as I read these threads, I keep remembering that I have never put my 50mm f/1.4 on my 6D2 that I bought in September. It was a useful lens on my T3i, but I've not thought of a need for it since. I use the 24-105mm kit lens all the time, and I guess if I looked at the metadata of the shots I've made with it, some of them would have been made in the general area of that focal length, but never felt the need for a faster lens or blurrier background. The IS and the low noise at higher ISOs keep me from wanting to use a faster lens. And I find that I use the 100mm macro much more for other things since I shoot FF, taking on much the role that the 50mm took on the Rebel.

The basketball videos I shot this month came close. But there was plenty of light in the arena (even if odd color and causing the flicker warning to go off). And I found with the zoom that even around 45mm was a little too tight to make sure I caught all the action. I guess with the 50mm, I would have just moved to a higher seat. (I also wanted to experiment with the autofocus on the zoom for that situation. It turned out that videos were consistently sharper just to focus manually on the basket rims—I was about equidistant from the two—and let DOF cover the whole court.)

If I have a point, it is that I just can't relate to your obsession with this focal length. I think back when I shot film, my "normal" lens was a 55mm f/1.2. I liked the look it gave, as I recall. But for general purposes, in that era when zoom lenses were ginormous and pretty bad, I would take a 28mm, an 85 (f/1.8?), and a 200mm when I went out to shoot pictures unless I had some specific application in mind that favored my bringing along a different lens.

 

[ahsanford]

If I have a point, it is that I just can't relate to your obsession with this focal length. I think back when I shot film, my "normal" lens was a 55mm f/1.2. I liked the look it gave, as I recall. But for general purposes, in that era when zoom lenses were ginormous and pretty bad, I would take a 28mm, an 85 (f/1.8?), and a 200mm when I went out to shoot pictures unless I had some specific application in mind that favored my bringing along a different lens.

Believe it or not, my fav FLs are in the 24-35 neighborhood... but Canon puts new things out in those FLs so I'm all set on that front.

50 is -- to me -- is an inflection point of the shortest FL where you get real bang for buck compositionally with large aperture. Longer than 50 you surely can exploit this effect more, but the lenses get huge quickly. I just kind of love a small tool for large aperture work, and 50s magically (i.e. via simple DG designs) can be that small and fast tool.

So 50 is not a favorite of mine, not at all -- it's just a neglected slice of the wonderful EF portfolio, and that neglect limits my use of that FL. Again, I may just pick up a 50L and call it good (quirks, whiffing AF and all).

- A

 

[melgross]

Time and time again, from pundits to optical experts, folks are saying double gauss only can get so good and you have to go retrofocus to obtain the IQ corner to corner without abberations and therefore that is why Sigma,Tamron and Zeiss have gone that route. Why would people think Canon won't do a retrofocus over 35mm? Just because they personally want a lighter smaller lens? I can't help but think it will be a pickle jar for sharpness, distortion and room for modern IS and AF internals.

Please tell me how I am wrong if you think so and how DG could be improved.

I always wondered if you could just supersize the elements diametrically with a DG design to get better results from the center of that glass (i.e. the FF sensor needed portions of it). In other words: oversize the elements like it was a medium format image circle so the FF output would be more central on the glass. In theory, such a lens would stay short front to back but step up in diameter (and presumably weight) considerably. (I wonder if the current 50L has done just that -- it's a 72mm front element while a 50 1.2 doesn't need anything so large for basic light gathering.)

But I'm no optical whiz and can't comment on if/how that might work. This idea could be nonsense.

- A

That’s basically what Canon does with their tilt/shift lenses, ergo the rumors that Canon will go to medium format.

 

[zim]

I think it will be a 50mm L 1.4 IS like the 85mm L IS 1.4... but what I really would like to see is a better 50mm 1.0L. Why? because it's possible to make

Or because it's just not necessary with modern sensors

The bokeh-fanatical masses of the world all simultaneously retorted: "WHAT?! You shut your mouth, sir! Good day! I SAID GOOD DAY." ;D

;D

I honestly think back in the day 1.0 was all about overcoming the ISO/grain limitations of film to just get the shot.

FWIW my 50 replacement/update senario would be

1.8 -> 2.0IS
1.4 -> 1.4L (like 35L1.4)
1.2L -> 1.2L (a refresh, same design, updated USM and coatings)

 

[Tom W]

A replacement for the 1.4 would be logical.
But I still would like to see another 50/1.0. just because.

 

[mb66energy]

Time and time again, from pundits to optical experts, folks are saying double gauss only can get so good and you have to go retrofocus to obtain the IQ corner to corner without abberations and therefore that is why Sigma,Tamron and Zeiss have gone that route. Why would people think Canon won't do a retrofocus over 35mm? Just because they personally want a lighter smaller lens? I can't help but think it will be a pickle jar for sharpness, distortion and room for modern IS and AF internals.

Please tell me how I am wrong if you think so and how DG could be improved.

I always wondered if you could just supersize the elements diametrically with a DG design to get better results from the center of that glass (i.e. the FF sensor needed portions of it). In other words: oversize the elements like it was a medium format image circle so the FF output would be more central on the glass. In theory, such a lens would stay short front to back but step up in diameter (and presumably weight) considerably. (I wonder if the current 50L has done just that -- it's a 72mm front element while a 50 1.2 doesn't need anything so large for basic light gathering.)

But I'm no optical whiz and can't comment on if/how that might work. This idea could be nonsense.

- A

Modifying a basic double gauss standard design with

- aspherical surfaces,
- alternative glass / plastic / blue refractive stuff showing other dispersion functions

should give a lot of chances to change the IQ - maybe to change it in a positive
way. I am not in optics too but if you have spherical aberration (due to the spherical
lens surfaces) you change the surface to correct that type of aberration which is
best practice in modern lenses. If chromatic aberrations occur you try to find a material
for one or more lenses where the dispersion function (refractive index depending on light color)
cancels the function of other elements. The EF 40mm 2.8 seems to be such a
derivative of a double gauss with at least one aspherical surface/lens
( http://cweb.canon.jp/ef/lineup/standard/ef40-f28stm/spec.html scroll down to corr. image)

Maybe the non-existence of new compact high aperture lenses has another reason:
Photographic equipment is - besides of taking photographs - a status symbol. Pay 1000
EUR/$ for a small lens or pay the same for a lens where others say: What a cool big lens?

 

[snoke]

But of course, I'd love the fancy non-L: New optical design, Ring USM, internal focusing, IS, double gauss / not huge, etc. Canon would ask $799-ish for that based on the 24/28/35 IS lenses, though I recognize that's a mint for a non-L / non-DO product currently sitting at a $329 price point. So I'm not confident that the lens I want is going to happen. We'll see.

- A

What you want: L-lens but not L-price. Good wish.

 

[Ozarker]

Decisions, decisions. TS-E? 85L? Future new 50L (Probably released next year.)? Hmmmm.... Ramen or beans?

 

[Canoneer]

I'd like to see a new incarnation of the 50mm F/1L. Probably an absurd wish - the level of optical corrections, exotic flint/crown glass, and multi coatings needed for a usable F/1 aperture at 50mm would make the thing cost a fortune.

Canon needs to concentrate on getting a competitive 50mm F/1.4 on the market - Image stabilization, weather sealing, IQ on par with the Sigma ART 50mm F/1.4, and an L-series designation would make it a suitable replacement for the 50mm F/1.2L; plus they could make a dumbed-down variant without weather sealing and OIS to replace the current 50mm F/1.4.

 

[melgross]

I think it will be a 50mm L 1.4 IS like the 85mm L IS 1.4... but what I really would like to see is a better 50mm 1.0L. Why? because it's possible to make

Or because it's just not necessary with modern sensors

The bokeh-fanatical masses of the world all simultaneously retorted: "WHAT?! You shut your mouth, sir! Good day! I SAID GOOD DAY." ;D

;D

I honestly think back in the day 1.0 was all about overcoming the ISO/grain limitations of film to just get the shot.

FWIW my 50 replacement/update senario would be

1.8 -> 2.0IS
1.4 -> 1.4L (like 35L1.4IS)
1.2L -> 1.2L (a refresh, same design, updated USM and coatings)

High Speed Ektachrome, yeah! Because of the stupid filtering in the film, daylight was 320, but where you really needed the speed, tungsten was just 160. Of course, we could push it for the wonderful green, and golf ball grain.