Roger Cicala - Variance in 50mm-58mm primes

[ahsanford]

As always, a fascinating read from Uncle Rog and his pal Olaf:
http://www.lensrentals.com/blog/2015/07/variation-measurement-for-50mm-slr-lenses

A few fascinating bits:

• The new Canon 50mm f/1.8 STM (with its old optical formula) is off the charts for high consistency / low copy-to-copy variation:
  
  "The Canon 50mm f/1.8 STM was amazingly consistent, and I'm not sure why. It's a simple design, but so are several of the other 50mm lenses. I suspect there might be something different in the manufacturing process of this very new lens, but until I take one apart and look inside (we haven't yet) I'm only speculating."
  
  
• He has already run this data on 24mm primes, and in comparison, the 50mm primes are more consistent in general.
  
  
• We have a habit of believing that newer lenses are more consistent than older ones. Roger's data -- granted, taken in a limited cross-section across many manufacturers -- would imply that this is not such a clear trend. In fairness, there was a very wide spread of simple 50s (Canon and Nikon budget f/1.8 designs) and very complex 50s (Sigma Art, Zeiss Otus 55) in this mix.

- A

 

[ahsanford]

And this data clearly proves, beyond any doubt, that this is what we need next:

 

[sulla]

Wow, the Canon 50 1.8 STM is the most consistent lens of them all.

Consistency is only a virtue if you're not a screw-up. IQ-wise.

So yes, ahsanford, this is really what we want. Canon could have my money as well, if only they wanted it...

 

[Mt Spokane Photography]

That's good to hear, I do expect that the AF design using the STM motor benefits from improved design and manufacturing processes.

However, I'd not want a 50mm f/'1.4 STM, I'd want pro level AF.

 

[ahsanford]

That's good to hear, I do expect that the AF design using the STM motor benefits from improved design and manufacturing processes.

However, I'd not want a 50mm f/'1.4 STM, I'd want pro level AF.

Yep. I don't the red ring or magical draw of the 50L. I don't even need f/1.4 -- f/2 would serve me just fine.

I just need a proper/modern/reliable/fast/consistent USM 50mm lens that's 90% as sharp as the Art/Otus glass in half the size and weight. That's not an unreasonable ask.

- A

 

[YuengLinger]

Ashanford continues his odd crusade against lenses faster than f/2.

Now that's consistent! :

 

[ahsanford]

Ashanford continues his odd crusade against lenses faster than f/2.

Now that's consistent! :

Well played, sir. [Tips his cap.]

I'd love the 50 f/nooneknows IS USM to be f/1.4, but that would likely mean:

• Canon can't put out something the same speed as the Sigma Art and not try to compete against it optically. (The current EF 50 f/1.4 USM does not compete against the Sigma Art from a host of optical perspectives.) That likely means abandoning the tiny basic optical design they have now. That likely means the lens must get bigger and heavier. Boo.
  
  
• A 50 f/1.4 IS USM would be very very attractive to all Canon users, including the 50L owners -- a new Canon 50 will likely mop the floor with the 50L from a sharpness perspective (heck, the current 50 f/1.4 USM already does that after f/2.8 ). Though sharpness isn't everything to the 50 prime crowd, not all 50L users are slamming that thing wide open, and as such I think such a sharp new lens would bite into 50L sales. Canon could combat this with either a higher asking price for the new lens (say $999) or add a red ring, a weathersealing gasket and put a much higher asking price to it.

I guess, to try to sum that up: Canon isn't going to offer another 50 f/1.4 with IS on it as a pleasant little upgrade to the current 50 f/1.4 USM because such a lens creates price point problems for Canon and will not optically compete against the Sigma Art.

It might (and I'm not certain) make more sense to make a killer alternative to the Sigma Art with a razor sharp 50 f/2 IS USM like the 35mm f/2 IS USM refresh of a couple years ago. Not making it f/1.4 makes it a different animal to the market and the natural comparisons to the Sigma Art will drop. It also would be smaller and lighter, which would be terrific. The pitch would be "90% as sharp as the Art, 50% as big, 50% as heavy, and this comes with IS." That would sell like hotcakes.

Truth be told, I just don't want a pickle jar for a staple walkaround lens. For me, there is immense value in smaller size + all the latest tech (internal focusing, modern USM, IS, solid build quality, etc.).

- A

 

[YuengLinger]

My hands are not the steadiest, for sure, but I'd snap up a version II of the 50mm 1.2 L that had no more trouble with focus shift than the 85mm 1.2 L II, and slightly less tendency than the current 50 L to produce nervous bokeh in outdoor situations with trees/shrubs.

Could a 1.2 II have faster AF than the current? I really don't know. The elements are very heavy, the DoF so shallow. But if it had sharpness, accurate AF without focus shift complications, and slightly better f/2.8 - f/5.6 bokeh, even at double the Sigma Art's price, I think a lot of Canon shooters would buy it. A lot. Awareness of the beauty and importance of a dynamite 50mm in the quiver is higher than ever--thanks in part to Sigma!

With a fantastic 50mm 1.2, rarely would I be shooting above f/2.2 in low to medium light, so my shutter speeds are going to be fast enough where IS isn't crucial. I like IS, certainly, and it helps make the ef 16-35mm f/4 a joy (but I still plan to sell that if Canon releases a non-IS version in f/2.8, just as sharp across the frame).

But I don't want to be stuck at f/2! Too much potential, magic going wider at 50mm.

I'm so eager for an update of the 50mm 1.2 L. I only tried one copy of the Sigma Art. I had trouble with the AF on my 5DIII, but, beyond that, I wasn't so impressed with any aspect of the IQ other than sharpness. A sharp lens is very important to me, but I felt the Sigma Art had a dark character, so to speak, probably the vignetting and me not having learned to expose more to the right with it.

BTW, the Sigma 35mm Art is one of my favorite, lenses. Can't praise it enough. I just didn't catch the great feeling so many have for the 50mm Art, and I rarely see anything that impressive being done with it in online images.

 

[siegsAR]

Old optic formula and consistency, on a cheap lens its a very good thing.

 

[ScottyP]

I will never ever understand people who object to a great sharp lens based on a few ounces of weight. If a few inches length and a few ounces of weight are really the be all end all, then get a 40mm pancake and delight in it. All of my favorite lenses have some heft to them. Big deal unless you are free climbing with your cameras.

If they could come out with a 50mm f/1.4 with IS and that was as sharp as Sigma, I promise Canon they would never, ever, ever, ever hear me bleating about the awesome lens being an inch too long and 3 ounces too heavy. Rather the contrary.

Same goes 3x if they would do that with an 85mm or a 135mm.

 

[ahsanford]

Could a 1.2 II have faster AF than the current? I really don't know. The elements are very heavy, the DoF so shallow. But if it had sharpness, accurate AF without focus shift complications, and slightly better f/2.8 - f/5.6 bokeh, even at double the Sigma Art's price, I think a lot of Canon shooters would buy it. A lot.

I've shot the 50L and found that at f/1.2, my AF hit rate for frame filling head and shoulders portrait was relatively poor. At first, I thought it was the 50L commonly discussed 'finnicky AF' that needs some TLC through AFMA to dial in. But that wasn't it. The damn AF boxes (I have a 5D3) were sufficiently large that I'd occasionally grab a bridge of the nose or eyebrow and it threw things off a bit. (I was using off-center boxes and not recomposing, so it wasn't that.)

My point is, there comes a point that your DOF is so small that perhaps MF is the way to go...

With a fantastic 50mm 1.2, rarely would I be shooting above f/2.2 in low to medium light, so my shutter speeds are going to be fast enough where IS isn't crucial.

Must disagree, but that's a personal thing. For me, IS is speed -- it makes everything better when you live on the dark end of lighting.

I occasionally shoot concerts in wretched, wretched light, and you live in the ISO 6400 - 8000 neighborhood. If your subject isn't moving, IS makes any lens better as it's a stop for stop trade of IS stops for ISO stops, i.e. if I have a 3 stop IS lens, I can walk my ISO back 3 stops and still net the shot. That's massive when you are at the edge of the earth for acceptable high ISO quality.

But yes, quick glass can trump low light as well. If you are routinely working in good light, not shooting video, it's not a terribly long FL, etc. then IS is far less pressing of a need.

- A

 

[ahsanford]

I will never ever understand people who object to a great sharp lens based on a few ounces of weight. If a few inches length and a few ounces of weight are really the be all end all, then get a 40mm pancake and delight in it. All of my favorite lenses have some heft to them. Big deal unless you are free climbing with your cameras.

If they could come out with a 50mm f/1.4 with IS and that was as sharp as Sigma, I promise Canon they would never, ever, ever, ever hear me bleating about the awesome lens being an inch too long and 3 ounces too heavy. Rather the contrary.

Same goes 3x if they would do that with an 85mm or a 135mm.

You are not alone at all. From all my discussions on this lens, I'd say a good 80% of people would prefer the best instrument possible despite the weight, and 20% would willingly give up a stop of speed or possibly a little bit less on the IQ front for a smaller/lighter lens.

Ask folks who own the Sigma 35 Art vs. the Canon 35 f/2 IS USM. It's a perfect case-study to peg who wants the best IQ vs. who values 'other things'. Neither are right, neither are wrong -- both are fine lenses -- but I think the percentage mix of who owns each lens would be telling (market-wise).

- A

 

[Mt Spokane Photography]

I'd love a 50mm f/1.2 that had near perfect optics, but that's not likely to happen, the cost and size might be out of sight. Its one thing to pay $12K for a big white, but for a 50mm to get a extra stop, it would not sell, not even at $3K. About the best we can hope for is a f/1.2 in the under 2K range with a substantial improvement, and hopefully truly excellent when stopped down. I've been surprised before, so there is hope.

 

[TeT]

Amazon Prime Day; hoping to get a cheap 50 1.8 STM...

 

[Hjalmarg1]

Ashanford continues his odd crusade against lenses faster than f/2.

Now that's consistent! :

I found my 50mm 1.8 STM to be very consistent. Way better than all my previous Canon's 50mm (1.2, 1.4 & 1.8) lenses.
I don't use it much but when needed it offers very acceptable IQ. I'd also like to see a 50mm 1.4 IS lens (with similar quality of 35/2 IS) but, it will kill 90% of 50L sales, except from those who need weather resistant lenses. If Canon releases such a good lens they will take my money right away.

 

[sulla]

The damn AF boxes (I have a 5D3) were sufficiently large that I'd occasionally grab a bridge of the nose or eyebrow and it threw things off a bit. [...]
My point is, there comes a point that your DOF is so small that perhaps MF is the way to go...

Indeed, the current 50 1.4 has already a very shallow DOF at portrait distances, not to speak of the 85 1.2.
And working with such shallow DOFs is very difficult. That's why I used to have my 5D3 set to single-point AF and even take the center only at this AF point (in the viewfindner it is the small point-like box within an AF-box), this very often is able to nail focus to where I want it. I don't know if this setting makes optimal use of the 5D3's capable AF-system, though.

But yes, using an aperture so large and a distance so short that the iris of an eye is in focus and the tips of the eyelashes are already out-of-focus is... not easy to work with. Not a keeper rate of 100% here as well...

But at larger distances wide apertures are extremely practical, think e.g. street photography at 1.4 in the night, not a big problem.

In my view the primes compete against the zooms: For me I decided a prime needs to be at least 1.4 to bring me a sufficiently big advantage over the 2.8 zooms to spend my money on it. I won't spend it on a 50/2 IS USM, but would do so on a 50 1.0 - 1.4 USM, as long as IQ competes with sigma. Size and weight do not matter for me, as the competition of the primes is against the zooms. If weight is the limiting factor, then there is always the EF 40.

 

[sulla]

I'd love a 50mm f/1.2 that had near perfect optics, but that's not likely to happen, the cost and size might be out of sight.

Size not necessarily. You can do fantastic optics if you make widespread use of aspherical lens elements. And in this area technical progress is quick: Formerly it was necessary to grind aspherical lenses very very carefully and this was very expensive. Thus lenses like the EF 50 1.4 do not have aspherical lenses, and comparatively poor optics.
But in 2015 aspherical lenses can be simply moulded (at least smaller lenses can), and this makes it possilbe to put aspherical elements in small lenses like the EF 40.

So, I think it would be possible to improve upon the IQ of the current EF 50 1.0 / 1.2 / 1.4 by making ample use of aspherical lenses (which are not so expensive any more today) and special glass types (which are still expensive today).

Doesn't need to be retrofocus designs with twice the number of lens elements, size and weight.

 

[Haydn1971]

Personally, I'm seeing Canon gradually swamp the 50mm market with better products.

We currently have the;

• 50mm f1.2 - No one comes close on full frame for aparure size - Nikon can't do f1.2 in its current mount, I doubt Sigma would make a Canon only prime as fast.
• 50mm f1.8 STM - Cheap, well built, sharp - it's a bargain

So what are the issues moving forward;

• IS hasn't yet been fitted to f1.4 glass
• Buyers want better performance and IS
• Buyers want faster than f1.8
• The f1.2 is flawed
• Macro ?

So where does that leave us (in my opinion) ?

• Theres a market for a "better" optical lens with IS and macro abilities - This could be either f1.8 or f2.0 IS USM and I'd see customers being happy to pay £400 for such a lens
• Theres a market for a Sigma Art killer 50mm f1.4 lens - I don't see this having IS, but I do see it being optically amazing - Customers would be happy to pay more than the current f1.2 for such a lens

Range moving forward...

• 50mm f1.8 STM
• 50mm f1.8/2.0 IS USM
• Macro adapter for the above
• 50mm f1.4 L
• 50mm f1.2 L

I don't see the f1.2 being replaced anytime soon unless Sigma have a go at a Canon only version.

 

[YuengLinger]

The damn AF boxes (I have a 5D3) were sufficiently large that I'd occasionally grab a bridge of the nose or eyebrow and it threw things off a bit. [...]
My point is, there comes a point that your DOF is so small that perhaps MF is the way to go...

Indeed, the current 50 1.4 has already a very shallow DOF at portrait distances, not to speak of the 85 1.2.
And working with such shallow DOFs is very difficult. That's why I used to have my 5D3 set to single-point AF and even take the center only at this AF point (in the viewfindner it is the small point-like box within an AF-box), this very often is able to nail focus to where I want it. I don't know if this setting makes optimal use of the 5D3's capable AF-system, though.

But yes, using an aperture so large and a distance so short that the iris of an eye is in focus and the tips of the eyelashes are already out-of-focus is... not easy to work with. Not a keeper rate of 100% here as well...
A
But at larger distances wide apertures are extremely practical, think e.g. street photography at 1.4 in the night, not a big problem.

In my view the primes compete against the zooms: For me I decided a prime needs to be at least 1.4 to bring me a sufficiently big advantage over the 2.8 zooms to spend my money on it. I won't spend it on a 50/2 IS USM, but would do so on a 50 1.0 - 1.4 USM, as long as IQ competes with sigma. Size and weight do not matter for me, as the competition of the primes is against the zooms. If weight is the limiting factor, then there is always the EF 40.

My very sharp pupils/catch-lights rate with the 85mm 1.2 II is high. I never use center AF & recompose--only the AF point which suits the composition.

This took some time to learn, and I occasionally have to tweak AFMA. But the lens does its job very well.

Why would the same in a refreshed 50mm 1.2 be more difficult for Canon? Isn't DoF thinner at 85mm?

 

[rfdesigner]

Personally, I'm seeing Canon gradually swamp the 50mm market with better products.

We currently have the;

• 50mm f1.2 - No one comes close on full frame for aparure size - Nikon can't do f1.2 in its current mount, I doubt Sigma would make a Canon only prime as fast.
• 50mm f1.8 STM - Cheap, well built, sharp - it's a bargain

So what are the issues moving forward;

• IS hasn't yet been fitted to f1.4 glass
• Buyers want better performance and IS
• Buyers want faster than f1.8
• The f1.2 is flawed
• Macro ?

So where does that leave us (in my opinion) ?

• Theres a market for a "better" optical lens with IS and macro abilities - This could be either f1.8 or f2.0 IS USM and I'd see customers being happy to pay £400 for such a lens
• Theres a market for a Sigma Art killer 50mm f1.4 lens - I don't see this having IS, but I do see it being optically amazing - Customers would be happy to pay more than the current f1.2 for such a lens

Range moving forward...

• 50mm f1.8 STM
• 50mm f1.8/2.0 IS USM
• Macro adapter for the above
• 50mm f1.4 L
• 50mm f1.2 L

I don't see the f1.2 being replaced anytime soon unless Sigma have a go at a Canon only version.

Some very good points. Many people do seem to be forgetting the macro in dicussions

Thinking function first and looking at replacing the current lenses:

Macro and landscape.
50 f2.5 or 2.8 IS Macro USM built around a very flat image plane, very sharp even in ther corners and supremely so stopped down, at the expense of Boke, AF aimed at accuracy rather than speed

Portrait & Sport:
50 f1.4 USM with improved optics re flair and edge but not corner sharpness at 1.4, merely so-so stopped down, much improved contruction, not so close focussing and no IS, but lighting fast AF.

The 1.2L is a creative lens, so different again and not particulalry threatened by the above, but could still be updated.