Patent: Canon EF 58mm f/1.4

[sdsr]

It is very easy to beat the mediocre (and overpriced) Nikon 58mm, but is difficult to overcome the Sigma 50mm Art. :

Nikon 58mm F1.4

Sigma 50mm F1.4 Art

It all depends on what you want; and if a lens tests badly that may say more about the test than the lens. Consider this, for instance:

https://photographylife.com/reviews/nikon-58mm-f1-4g

It's far more expensive than I would want to spend on such a lens, but that's partly because I have several elderly lenses that I like a lot, for which I paid vastly less, and which, I suspect, would test far worse!

 

[Coldhands]

Re: Canon patent for a 58mm f/1.4 seen in photorumors

50L is a front focusing Design. Are you sure that is "floating elements"? FL means as i understand it, that both groups of lenses - the front group and a rear group - chance independendly and at once their position to the focal plane. This will need two AF-mechanisms in the lens. This means the position of the rear group is not fixed, as in a front focusing design.

Yes, I stand corrected, you are right: 50/1.2 L is no floating element design. EF 85/1.2L is. Seems to be the reason why the 50/1.2 L has that pronounced focus shift issue that should be corrected in any successor lens, no matter whether it is a Mk. II or something different (f/1.2 and/or 1.4, 50mm and/or 58 mm).

Autofocus is definitely possible also with floating element design. It may be more difficult to implement and it may well be the reason why the 85/1.2 L has focus-by-wire AF.

If you're curious, this article does a good job of showing the focus system for a lens with a floating element. Quite a simple mechanism, in fact. Floating elements are quite common in wide-angle lenses and almost universal in true macro lenses.

By the way, the focus shift in the 50mm f/1.2 is caused by under-corrected spherical aberration - floating elements wouldn't have any effect on this.

 

[ajfotofilmagem]

It is very easy to beat the mediocre (and overpriced) Nikon 58mm, but is difficult to overcome the Sigma 50mm Art. :

Nikon 58mm F1.4

Sigma 50mm F1.4 Art

It all depends on what you want; and if a lens tests badly that may say more about the test than the lens. Consider this, for instance:

https://photographylife.com/reviews/nikon-58mm-f1-4g

It's far more expensive than I would want to spend on such a lens, but that's partly because I have several elderly lenses that I like a lot, for which I paid vastly less, and which, I suspect, would test far worse!

Yes, it depends on the price.
If the Nikon 58mm F1.4 would cost up to $ 600 might be a good choice. But in US1700 is just mediocre when wide open. Chromatic aberration screams aloud: I am a 1970 project that has been repackaged and overpriced.

Hopefully Canon NOT do the same thing with his hypothetical 58mm F1.4.

 

[ajfotofilmagem]

For those who did not understand my criticism:

Canon 50mm F1.4 from 1993, costing US$350

Nikon 58mm F1.4 from 2013, costing US$1700

 

[Rudeofus]

At the moment there are two commonly employed ways to make a 50mm lens for slr cameras:

[list type=decimal]
[*]reverse telephoto design
[*]double gauss design
[/list]

The first design (used by Zeiss Otus and Sigma 50 Art) gives unwieldy lenses with exceptional sharpness. The second designis the more common one, it gives compact lenses with moderate sharpness (see Canon's current 50mm lineup). Main limitation of double gauss design is that you can't properly correct such a lens under its space constraints, there has to be sufficient space between rear element and film/sensor plane. A 58mm lens would give quite a bit more space, and therefore allow construction of a better lens which still uses the simple double gauss design.

As the patent implies, Canon still isn't there. 38mm back focus length is still too short for full frame slr cameras, so unless they make this a crop or mirrorless lens (which I think is extremely unlikely), this patent is not going to lead to a product anytime soon.

 

[privatebydesign]

At the moment there are two commonly employed ways to make a 50mm lens for slr cameras:

[list type=decimal]
[*]reverse telephoto design
[*]double gauss design
[/list]

The first design (used by Zeiss Otus and Sigma 50 Art) gives unwieldy lenses with exceptional sharpness. The second designis the more common one, it gives compact lenses with moderate sharpness (see Canon's current 50mm lineup). Main limitation of double gauss design is that you can't properly correct such a lens under its space constraints, there has to be sufficient space between rear element and film/sensor plane. A 58mm lens would give quite a bit more space, and therefore allow construction of a better lens which still uses the simple double gauss design.

As the patent implies, Canon still isn't there. 38mm back focus length is still too short for full frame slr cameras, so unless they make this a crop or mirrorless lens (which I think is extremely unlikely), this patent is not going to lead to a product anytime soon.

Are you sure about that? With a flange distance of 44mm and a mount thickness of 5mm the current design gives 39mm of clearance, the electrical connections add another 2mm and the mirror doesn't swing past the connectors in the body. I don't see why back focus of 38mm isn't possible with the EF mount, the lens rear element would be extreme, but all these new lenses do seem to be pushing boundaries.

 

[MayaTlab]

At the moment there are two commonly employed ways to make a 50mm lens for slr cameras:

[list type=decimal]
[*]reverse telephoto design
[*]double gauss design
[/list]

The first design (used by Zeiss Otus and Sigma 50 Art) gives unwieldy lenses with exceptional sharpness. The second designis the more common one, it gives compact lenses with moderate sharpness (see Canon's current 50mm lineup). Main limitation of double gauss design is that you can't properly correct such a lens under its space constraints, there has to be sufficient space between rear element and film/sensor plane. A 58mm lens would give quite a bit more space, and therefore allow construction of a better lens which still uses the simple double gauss design.

As the patent implies, Canon still isn't there. 38mm back focus length is still too short for full frame slr cameras, so unless they make this a crop or mirrorless lens (which I think is extremely unlikely), this patent is not going to lead to a product anytime soon.

Are you sure about that? With a flange distance of 44mm and a mount thickness of 5mm the current design gives 39mm of clearance, the electrical connections add another 2mm and the mirror doesn't swing past the connectors in the body. I don't see why back focus of 38mm isn't possible with the EF mount, the lens rear element would be extreme, but all these new lenses do seem to be pushing boundaries.

Agreed, otherwise nearly all of Canon's patents on Egami would be inappropriate for the EF mount, such as this one filed in 2011 for a rather double gaussish design : http://egami.blog.so-net.ne.jp/2013-06-11

 

[Rudeofus]

Are you sure about that? With a flange distance of 44mm and a mount thickness of 5mm the current design gives 39mm of clearance, the electrical connections add another 2mm and the mirror doesn't swing past the connectors in the body. I don't see why back focus of 38mm isn't possible with the EF mount, the lens rear element would be extreme, but all these new lenses do seem to be pushing boundaries.

Given what you write 38mm should be doable, and the 85L seems to be one of these extreme designs where the rear element goes into the mount. I stand corrected.

What this still shows, is that double gauss design with better performance is only possible for fast glass with focal length > 50mm, although the Nikon 58mm still suggests that even 58mm isn't enough to create Otus like performance.

 

[sdsr]

What this still shows, is that double gauss design with better performance is only possible for fast glass with focal length > 50mm, although the Nikon 58mm still suggests that even 58mm isn't enough to create Otus like performance.

You seem to assume that Nikon was trying to come up with Otus-like performance. If that assumption is true, they spectacularly failed, but the review I linked to earlier and the blurb on Nikon's website - which goes out of its way to mention the beautiful bokeh etc. (which isn't even mentioned in the blurb for most of their other 50-85mm lenses) of this lens - suggest that they weren't trying to. Either that, or they're attempting to make the most out of a failure....

Of course, even if the 58 1.4 performs as Nikon planned, I wonder how much better it is (if at all) than, say, the Minolta 58mm 1.4; you can buy one of those in excellent condition from KEH for $56.00 - i.e. c. 1/30 the price of the Nikon!

 

[Rudeofus]

I am quite confident, that both Nikon and Canon could have created a reverse telephoto type 50mm lens, which would at least match the Sigma 50 Art in terms of sharpness. Both companies are not exactly newcomers, and both have created some incredibly sharp lenses before.

At the same time, Canon's 50L and 85L, while not terribly sharp wide open, have a passionate following due to their beautiful bokeh. What Nikon (and now Canon) may have aimed for was a double gauss 58mm lens, which is a tad sharper than what they could have done with a 50mm lens, but still has decent bokeh.

 

[Mika]

There are several things mentioned in this thread that I'd like to know where they originate from. You don't need to spare the technical details, I do optical design full time and can probably follow.

First is the notion that combination of a floating element and autofocus does not work. Where does this come from?

Second is: what do people mean with floating element? To me it means part of the objective moves at a different rate than the rest, and the rest may not move at all. This can be accomplished with a rather standard rail and cam system, and does not require several AF motors.

50/1.4 does not use a floating element, the entire objective barrel moves - I own this one and could check this is the case. 50/1.0L does have a floating element and focuses at least by extending the front, as does its sister objective 85/1.2. I don't have the 50/1.2L and can't say whether the entire assembly moves or whether just the front part moves.

EDIT:
The way I read the Canon 50 mm patent is that they are trying to think of a way of getting some retrofocus benefit without increasing the objective size or mass considerably while providing an improved image quality. They don't particularly need to do better than 50 mm Art or 55 mm Otus, just improve the EF50/1.4 and not increase the mass substantially.

And yes, 58 mm instead of 50 mm allows slightly more room for balancing the back focal distance with aberration correction. A back focal length of 38.5 mm should work fine with EF mount. One cannot go much closer, though. From the patents that I've seen, I think it is either 50/1.0 or 85/1.2 that has the shortest BFL of the EF objectives.

 

[ahsanford]

EDIT:
The way I read the Canon 50 mm patent is that they are trying to think of a way of getting some retrofocus benefit without increasing the objective size or mass considerably while providing an improved image quality. They don't particularly need to do better than 50 mm Art or 55 mm Otus, just improve the EF50/1.4 and not increase the mass substantially.

Yep. The 50 f/nooneknows IS USM crowd -- specifically, the folks here who think this lens is not the future L but instead it's the future 50 f/1.4 USM replacement -- want to keep the size of this lens on or about where it is now. Save the big pickle jars for the next L.

Honestly, as much as we all want that lens to be a razor for sharpness, the 50 f/1.4 USM not wretched today. By far, the biggest improvements that lens needs has nothing to do with the optical formula. That lens needs an AF upgrade and a build quality upgrade far, far more than an optical upgrade. I know I'll wind people up and get this thread covered with everyone's soft f/1.4 shots taken on that 50 f/1.4, but people don't scream at that lens for its wide open softness -- they scream at the AF and the wimpy mechanical external focusing nonsense of that design.

For me -- and you may disagree -- just slapping proper modern fast reliable USM AF + internal focusing on the exact same 50 f/1.4 optical design is worth a good $600-700 to me. That said, here's hoping we get a top to bottom new lens like the 24/28/35 IS refreshes. I'd buy that on day one.

- A

 

[Rudeofus]

And yes, 58 mm instead of 50 mm allows slightly more room for balancing the back focal distance with aberration correction. A back focal length of 38.5 mm should work fine with EF mount. One cannot go much closer, though. From the patents that I've seen, I think it is either 50/1.0 or 85/1.2 that has the shortest BFL of the EF objectives.

What's your prediction then? Nikon's 58mm faces harsh criticism for it's reported softness wide open, what are the odds that Canon will fare better in this regard? How would such a Canon 58mm F/1.4 fit into their lineup, given that their prosumer 50 is said to be acceptably sharp, whereas their 50L claims to be bokeh king?

 

[Mika]

And yes, 58 mm instead of 50 mm allows slightly more room for balancing the back focal distance with aberration correction. A back focal length of 38.5 mm should work fine with EF mount. One cannot go much closer, though. From the patents that I've seen, I think it is either 50/1.0 or 85/1.2 that has the shortest BFL of the EF objectives.

What's your prediction then? Nikon's 58mm faces harsh criticism for it's reported softness wide open, what are the odds that Canon will fare better in this regard? How would such a Canon 58mm F/1.4 fit into their lineup, given that their prosumer 50 is said to be acceptably sharp, whereas their 50L claims to be bokeh king?

Somehow I don't think Canon is going to launch 58 mm line-up. I think this patent is to hinder anybody else's possibility of providing 58 mm retrofocus objective.

What it comes to modernizing Canon's 50/1.4, the motor really has to improve. But I wouldn't pay more than 400 € for it. To be able to squeeze 600 € for that from me, it has to: [list type=decimal]
[*] Remain light and small enough (so I can use it abroad)
[*] Have more reliable and accurate autofocus mechanism (everybody says Sigma drifts, but my EF50/1.4 is the only objective where I've noted it)
[*] Keep it F/1.4
[*] Provide a modest improvement in resolution.
[/list]

This is because I already have 50A and that serves me well enough for local photography, but it is too heavy for travel. Then again, Canon launched 50/1.8 STM, but I'd like to have the always on manual focus possibility. If they don't release 50/1.4 update soon, it's going to be the 50/1.8 STM then.