The 2020 RF lens roadmap, up to 8 new lenses coming in 2020

justaCanonuser

Grab your camera, go out and shoot!
Feb 12, 2014
456
266
Frankfurt, Germany
Like a 50mm 1.2 that is bigger and heavier than the EF one
It is bigger because it is optically better controlled, so its design needed more glass. I have and like the EF 50mm 1.2, in particular because it is quite compact and can deliver a nice bokeh. But optically it is definitely on the vintage side, including sometimes more softness wide open than I'd love to have. It was clear that Canon had to come up with an up-to-date design with the RF version. On the other side, the RF is clean and perfect but does not have this old school character of the old EF version...
 

ColinJR

EOS-R
Nov 27, 2018
32
26
robertsonrep.com
I say this as someone who has stuck with Canon primarily because of their tilt-shifts—I don't think they need to focus on tilt-shifts for RF, but I'd be damn curious to know what kind of improvements they might be able to bring to TS-E lenses designed for the RF mount.

As is, tilt-shifts work great on my R—the adapter makes accessing the shift knobs easier! lol
 

Michael Clark

Now we see through a glass, darkly...
Apr 5, 2016
1,056
512
It is bigger because it is optically better controlled, so its design needed more glass. I have and like the EF 50mm 1.2, in particular because it is quite compact and can deliver a nice bokeh. But optically it is definitely on the vintage side, including sometimes more softness wide open than I'd love to have. It was clear that Canon had to come up with an up-to-date design with the RF version. On the other side, the RF is clean and perfect but does not have this old school character of the old EF version...

Designing a lens to have a high degree of flat field correction in order to do well from center to edge at the same focus position when imaging flat test charts seems to suck the soul out of the rendering character of a lens.

Just look at the entire Sigma ART range: every one of them is sharp as a tack when imaging flat test subjects, yet none of them have the kind of rendering character when used in the three dimensional world that many portraitists desire. It's the same thing as using a 90-100mm Macro lens for portraiture. Those macros are optimized to be best shooting flat subjects at very close distances, not rendering pleasing out of focus highlights at typical portrait distances.

What most lens tests don't show is that many of the lenses from the past that were prized for their rendering character when shooting portraits is that the edges are sharper when the lens is refocused for maximum edge sharpness. The reason the edges are as soft as they are imaging a flat test chart when the lens is focused for the center is that the point of sharpest focus on the edges is well in front of the surface of the flat test chart.
 
  • Like
Reactions: ColinJR

SecureGSM

2 x 5D IV
Feb 26, 2017
1,319
313
  • Like
Reactions: jd7

mk0x55

[5DsR]
Nov 16, 2018
55
52
I say this as someone who has stuck with Canon primarily because of their tilt-shifts—I don't think they need to focus on tilt-shifts for RF, but I'd be damn curious to know what kind of improvements they might be able to bring to TS-E lenses designed for the RF mount.

As is, tilt-shifts work great on my R—the adapter makes accessing the shift knobs easier! lol
I guess they will introduce them to the RF line at some point, but surely not before they release a pro EOS R-like body with a lot of resolution for architectural and landscape shooters. Also having in mind that they just recently released a number of tilt-shift EF lenses.
 
  • Like
Reactions: ColinJR

mk0x55

[5DsR]
Nov 16, 2018
55
52
Michael Clark
Does this image(s) represent the Sigma Art entire range rendering character qualities you are referring to?

https://www.reddit.com/r/SonyAlpha/comments/91jtuh
https://www.reddit.com/r/SonyAlpha/comments/8y17wm
Overall they render the portrait beautifully, but then you open the pictures and zoom in on them, you'll see the flatness of them if you just consider the face or body parts. The 3D separation is there more through the out-of-focus areas than the tonal transitions.

You'll see a lot of difference if you compare a shot taken using a Zeiss or even a Sony FE lens and a Sigma Art lens as Tony Northrup's example also shows:
The 3D-pop of the Sony is way better than the Sigma Art.

There are two articles on microcontrast (or the lens' ability to render 3D characteristics of a scene):
There are more examples on this out there.

Modern lenses often sacrifice this 3D-rendering to resolution (there are some tradeoffs to be made due to physics and what is technologically possible to make); some of them appear to strike a better balance to my taste (e.g., Zeiss) than others (e.g., Sigma). On the one hand, I wouldn't accept poor resolution to retain top microcontrast (e.g., old prime lenses with very low element count); on the other hand, I wouldn't accept poor microcontrast to agressively optimize for the best possible resolution either (e.g., Sigma Art). Good microcontrast creates the perception of a better resolution; and if you finally view a picture, you seldom view it at 100% magnification (unless you are looking at a bird picture that has been heavily cropped, perhaps, or something similar).

EDIT:
I must confess though that there is a bit of confusion around it; and even I feel slighly confused - there are two lines of using the term microcontrast or 3D-pop. The first is what Yannick Hong describes among other in the above mentioned articles. The second is what Lloyd Chambers describes here:
While Yannick bases the notion of it on the quality/fidelity of tonal gradations; Lloyd seems to mostly base it on resolution, the lens' ability to hold contrast at high frequency of changes, which is actually a lot different from Yannick's definition.
To sort this out in my mind, I think there are two distinct things that contribute to an image rendering a scene with a 3D-look:
(1) The quality of the tonal gradations as the low-frequency intertonal detail (low-frequency MTF) - this is probably the dominant factor;
(2) The high-frequency intertonal detail (high-frequency MTF) that helps separate the interfaces between different objects, or an object from its background.

It would be cool to hear something clear and definitive on this, as there seems to be a lot of confusion around it. (The 3D-pop effect in images and the differences between different lenses' rendering are pretty clear at the same time, though.)

EDIT 2 (follow-up):
I'd actually agree with Lloyd on that the term Microcontrast shall for the sake of intuitiveness refer to the high-frequency MTF;
What Yannick refers to, and what often has a more dramatic impact, especially one that is retained when down-sampling the picture, shall be called something like tonal fidelity or tonal rendition.
Both together make up the 3D-pop of an image. The more of each, the better; but when making a lens, the design optimizations for the one compete with the other to a degree.
 
Last edited:
  • Like
Reactions: mb66energy

mb66energy

EOS 6D MK II
Dec 18, 2011
1,294
197
Germany
www.MichaelBockhorst.de
@mk0x55 thanks for providing the link to Lloyd Chambers article - the most interesting part for me was his statement about sharpening: With AA sensors some sharpening is always helpful but I always use the standard Sharpening with DPP instead of this aggressive Unsharp Mask thing. Noise reduction is usually turned off except in some situations and then chroma noise first. This gives the best 3D effect and texture reproduction for my taste.
And the EF-S 10-22 needs always 2 steps more sharpening to be comparable like better lenses loosing some of the real detail while the EF-M 32 mostly suffers from the AA filter and needs one step less compared to my other lenses.

IMO Canon does very well with lenses like EF-M 32, EF-S 60, EF 100 macro, EF 2.0 100, EF 5.6 400 and just the old EF 2.8 24. The EF 70 200 4.0 IS is close, the EF-S 10-22 is not but thanks to high contrast and great color reproduction a fine lens.
At the moment I am not too interested in the RF system but if Canon provides an interesting camera I am happy with what I have (24/100/400/70-200) except ... a 1.4 50mm lens for FF - and if it is compact & macro like the 1.8 35 I will trade in a half stop for these advantages!
 
  • Like
Reactions: mk0x55

mk0x55

[5DsR]
Nov 16, 2018
55
52
IMO Canon does very well with lenses like EF-M 32, EF-S 60, EF 100 macro, EF 2.0 100, EF 5.6 400 and just the old EF 2.8 24. The EF 70 200 4.0 IS is close, the EF-S 10-22 is not but thanks to high contrast and great color reproduction a fine lens.
At the moment I am not too interested in the RF system but if Canon provides an interesting camera I am happy with what I have (24/100/400/70-200) except ... a 1.4 50mm lens for FF - and if it is compact & macro like the 1.8 35 I will trade in a half stop for these advantages!
I think some lenses are pretty good in this respect, but the large count of optical elements in new glass is a bit worrying, actually.

One article from Yannick shows some figures as "The current state of [lens brand]" for Canon, Nikkor, Sigma, Zeiss, etc. However, I'd take it with a grain of salt because (1) there is no scientific measurement of it, as well as (2) he claims that both Milvus and Otus 1.4/85 and 1.4/50 belong below the "line of realism", which means poor tonal fidelity. I can't quite imagine that to be the case, but will gladly verify it as soon as I have at least the Milvuses at hand.

This is the article I refer to: https://yannickkhong.com/blog/2016/2/23/the-problem-with-modern-optics
 

mk0x55

[5DsR]
Nov 16, 2018
55
52
Michael Clark



Does this image(s) represent the Sigma Art entire range rendering character qualities you are referring to?

https://www.reddit.com/r/SonyAlpha/comments/91jtuh
https://www.reddit.com/r/SonyAlpha/comments/8y17wm
One more follow-up on these two portrait images. While I think they are somewhat flat by the rendering of the tonal gradation of the models; I think that the use of flash (as opposed to only ambient lighting), tends to alleviate that weakness of a lens (poor tonal rendition), as the flash beam is brief and intense; intense enough to better penetrate the hyper-corrected high-element count glass, and so rescue some of that 3D pop that would be largely lost in coatings and the glass itself in pure ambient lighting conditions.

EDIT: A few more examples on 3D-pop - pretty good article:
He makes fun of a few claims relating to causes of the 3D pop differences between classical and modern glass; nevertheless I think there is a drop of truth to it. There is concrete, independent, photographic evidence of it out there (although not for conclusions like in the 'Summary' section).
 
Last edited:

degos

EOS RP
Mar 20, 2015
201
123
It was clear that Canon had to come up with an up-to-date design with the RF version. On the other side, the RF is clean and perfect but does not have this old school character of the old EF version...
There was nothing stopping them coming up with an up-to-date EF design, other than business strategy; certianly they updated other lenses for less valid reasons. And I doubt it would have been any heavier than the 950g of the RF lens. After all the EF 50mm 1.0 was only 67g heavier than that... and despite all that 'optical correction' the RF lens still has coma out the wazoo.

One of the claimed advantages of the short flange was that it made wide lenses easier to design than on EF, yet here we are with 1kg lenses and not even with IS. It's just not making sense.
 
Last edited:

Mikehit

EOS 5D MK IV
Jul 28, 2015
3,227
415
One of the claimed advantages of the short flange was that it made wide lenses easier to design than on EF, yet here we are with 1kg lenses and not even with IS. It's just not making sense.
Why do you quote their reference to 'easier to design' then talk about weight? The two are not mutually exclusive. They seem to have done more with the RF 50mm f1.2 than they managed with the EF version.
OK, I accept we are talking about several years' technolgical development since the EF was released so they would have improved that but comments I have seen suggest the RF has taken it further.
 

ColinJR

EOS-R
Nov 27, 2018
32
26
robertsonrep.com
I guess they will introduce them to the RF line at some point, but surely not before they release a pro EOS R-like body with a lot of resolution for architectural and landscape shooters. Also having in mind that they just recently released a number of tilt-shift EF lenses.
Exactly. I also recognize that tilt-shifts are still extremely niche. It says something that neither Fujifilm nor Sony have any tilt-shifts in their line, and Nikon barely bothers to try and compete with Canon.
 

QuisUtDeus

EOS 80D
Feb 20, 2019
115
80
Released on 2012.
Ah, yes you're right. Still, another way of phrasing it is "since 2012, Canon has filled in the few gaps in their sub-$1k FF lens lineup". Seriously, they have a standard zoom, tele zoom, trio of IS primes, and a pancake. The 100L macro is barely over $1k, and the non-L is $600. For the entry-FF user, what's missing?
 

Mr Majestyk

EOS 80D
Feb 20, 2016
178
73
Why can they give us a lens roadmap but not a camera roadmap. This indicates they have no faith in themselves to deliver a truly class leading camera any more. They don't need many details just a broad outline of what to expect. Not a chance in hell I'd buy into the R system without indication they can compete camera wise.
 
  • Like
Reactions: epiieq1

Jethro

EOS R
Jul 14, 2018
241
116
Why can they give us a lens roadmap but not a camera roadmap. This indicates they have no faith in themselves to deliver a truly class leading camera any more. They don't need many details just a broad outline of what to expect. Not a chance in hell I'd buy into the R system without indication they can compete camera wise.
In terms of whether you buy in now, it depends on whether the R or RP are right for you - if not, then you presumably won't buy in. All a 'roadmap' for upcoming R series cameras would do is potentially put you off buying one of those cameras (if the upcoming one was looking like being closer to your specs). But such a roadmap would never provide definite dates, or definite specs or definite pricing - so all it would do is fuel speculation. Canon's tactic is to try and sell as many of each new R series camera as they can - so there was no real inkling of the RP for months (and a Xmas buying period) to avoid putting off sales of the R. Similarly, they won't cannibalise (or postpone) sales of the RP (or R) by encouraging people to wait until an officially flagged different version is confirmed. This is how they work - it is actually how they 'compete camera wise'.
 

epiieq1

EOS 5D III, 1DX
Aug 9, 2013
33
17
Take a look at the Panasonic S1R. I’ve been shooting it for a couple weeks. At 120 refresh rate, it’s like looking through glass.

Just waiting for ef adapter now.
I rented one for a week or so. I was thinking it might be my choice for going mirrorless...sent it back, and am now trying to decide what other direction to go as I was super disappointed (AF was atrocious).
 

mb66energy

EOS 6D MK II
Dec 18, 2011
1,294
197
Germany
www.MichaelBockhorst.de
I think some lenses are pretty good in this respect, but the large count of optical elements in new glass is a bit worrying, actually.

One article from Yannick shows some figures as "The current state of [lens brand]" for Canon, Nikkor, Sigma, Zeiss, etc. However, I'd take it with a grain of salt because (1) there is no scientific measurement of it, as well as (2) he claims that both Milvus and Otus 1.4/85 and 1.4/50 belong below the "line of realism", which means poor tonal fidelity. I can't quite imagine that to be the case, but will gladly verify it as soon as I have at least the Milvuses at hand.

This is the article I refer to: https://yannickkhong.com/blog/2016/2/23/the-problem-with-modern-optics
This article supports my observation that e.g. the EF 5.6 400 has very good rendering using only a few lens elements. The EF-M 32 is an exception while having 14 lenses it renders IMO very naturally straight from f/1.4. Just a loose unscientific observation: Canon seems to increase the lens numbers but tries to group them where possible - so the EF-M 32 has only 8 lens groups hence 16 glass-air surfaces where reflections occur. Maybe this is the secret (or one of it) of the high contrast and very good technical correction without loosing to much "character".

Like Yannickkhong I would like to see slightly modified conventional lenses with new coatings, new glass types and some aspherical surfaces (if necessary/helpful) to provide e.g. a light compact image stabilized 4.0 200 mm lens with 6 lenses in 3 groups or so. But in a world where specs are often more important than real value of things only a few are in the market for a highly corrected 6/4 lens/elements optics @ 700 EUR while other lenses have three times the lens elements, more technical sharpness at the same price ...
 

Kit.

EOS 6D MK II
Apr 25, 2011
1,360
747
I say this as someone who has stuck with Canon primarily because of their tilt-shifts—I don't think they need to focus on tilt-shifts for RF, but I'd be damn curious to know what kind of improvements they might be able to bring to TS-E lenses designed for the RF mount.
An obvious (but far from easy) one would be autofocus (including autotilt).