Here is the official Canon EOS R system white paper

Just read through the 43 pages, and, based on the MTF, the 28-70/2 at f/2 is sharper than both 24-70/2.8 & 24-70/4 at both ends of the zoom, the RF 50/1.2 is waaayyy sharper than the EF 50/1.2 and 50/1.8, the RF 24-105/4 is slightly worse in some parts of the frame than the EF 24-105/4 Mk.II, and roughly equal to the 24-105/3.5-5.6 in others. The RF 35/1.8 IS Macro is slightly worse than the EF 35/1.4L but better than EF 35/2 IS USM overall.

I reached the same conclusion for the 24-105 after comparing the MTF charts. It is somewhat unfortunate that the resolving power isn't a full step up from the EF mk II version, but it seems that the intention was to reduce size and weight. If not looking for specialty lenses, these benefits will likely be what most people starting out in FF mirrorless photography will appreciate the most.
 

Attachments

  • Screen Shot 2018-09-05 at 19.25.55.png
    Screen Shot 2018-09-05 at 19.25.55.png
    207 KB · Views: 171
Upvote 0
Negatives are lack of tactile feedback, no hard stop at infinity or the MFD (for macro and astro, its common to just set to one end of the focal range), and that the lens needs to be mounted to a powered-on body to change the focus. The last one really only matters with lenses where the front element extends with focusing (like the 85L), where you have to remember to retract the front element before unmounting the lens. I don’t know of the RF 50/1.2L has an extending front element, but the focusing group comprises the front 2/3 of the elements (the front 11 of 15 elements).

Positives are that you can change the direction of focus (presumably via a camera setting), not sure why you’d want to.

Mixed is the focus throw can be variable (it is on STM lenses, I presume it will be/can be for RF lenses). I’ll just copy this bit from my review of the M18-150: With a fast manual rotation speed, the focus ring must be rotated 150° to rack the lens through the full focusing range, while a slow rotation speed requires a rotation of 240° to cover the same range. The idea behind this is to allow faster manual focusing with a shorter focus throw when moving between close and distant subjects, while also allowing precise fine-tuning of manual focus with a longer focus throw. I call it mixed because while it sounds nice in theory, in practice some find it annoying and difficult to get used to.


The new lenses are focus by wire, BUT are linear focus, so throw does not change with speed (I assume a firmware update could add the option to do so though)

Jordan Drake confirmed:
Also, to those shooting strobes, the camera focuses with aperture wide open ALWAYS. No AF hunting when shooting in a low lit studio at F11.

Again Jordan confirmed this:
 
Upvote 0
Jul 21, 2010
31,096
12,857
The new lenses are focus by wire, BUT are linear focus, so throw does not change with speed (I assume a firmware update could add the option to do so though)
Linear focus in that context probably refers to the actual focus motor. Canon indicates in the white paper that the 24-105 uses NanoUSM with a linear motor. That says nothing about how the lens is programmed to interpret the movement of the focus ring encoder and translate that to driving the focus motor.
 
  • Like
Reactions: 1 user
Upvote 0

CafferyPhoto

Commercial photog
CR Pro
Mar 28, 2017
22
21
Washington State
cafferyphoto.com
The new RF mount makes possible greater lens design flexibilities:
1. Large diameter rear lens elements that are much closer to the full frame image sensor — enhancing overall optical performance (in particular, tighter control over optical aberrations at image extremities)
2. Lenses having the same specifications for focal length and maximum aperture as current EF mount lenses—but having significantly higher image quality — within the same size and weight
3. High optical performance, large aperture (F1.2) prime lenses for full frame cameras
4. Zoom lenses of higher brightness with constant aperture over their focal ranges — while still modest in size and weight
 
  • Like
Reactions: 1 user
Upvote 0
RF lenses – even the L-series USM lenses – are all focus-by-wire. Ugh.
It appears to me that this new system release, reveals why in last years some tech appeared in only few releases from Canon. I think to Nano USM used in only 2 lenses ... that is kind of hybrid between USM and STM. Many asked why not used in more lenses.
 
Upvote 0

CafferyPhoto

Commercial photog
CR Pro
Mar 28, 2017
22
21
Washington State
cafferyphoto.com
In the new EOS R system the lens embodies new technologies that combine with the IS system in the camera to implement an augmented control over the image blurring that can be caused by shaking and vibration of the lens-camera system. This is empowered by an interactive data communication between the two. Within the lens a dual gyro sensor system detects any inadvertent physical movements of the system and this data is reported across lens-camera communication to the DIGIC 8 processor. At the same time the image sensor is “seeing” any blur stimulated by these same movements and it also reports this image data to the DIGIC 8 processor. These two data reports are algorithmically processed at very high speed and a compensation control signal is generated and sent back at high speed to the lens to actuate the IS optical element that counteracts the disturbance.
 
Upvote 0

CafferyPhoto

Commercial photog
CR Pro
Mar 28, 2017
22
21
Washington State
cafferyphoto.com
"This paper has discussed details of the new RF lens mount. The advantages of this new mount design are immediately evident in the designs of the first four lenses of the RF lens family — as outlined in this paper."

Ha! I would have written that conclusion sentence as a middle schooler.
 
Upvote 0

BeenThere

CR Pro
Sep 4, 2012
1,242
672
Eastern Shore
The new RF mount makes possible greater lens design flexibilities:
1. Large diameter rear lens elements that are much closer to the full frame image sensor — enhancing overall optical performance (in particular, tighter control over optical aberrations at image extremities)
2. Lenses having the same specifications for focal length and maximum aperture as current EF mount lenses—but having significantly higher image quality — within the same size and weight
3. High optical performance, large aperture (F1.2) prime lenses for full frame cameras
4. Zoom lenses of higher brightness with constant aperture over their focal ranges — while still modest in size and weight
#2 is implying we will be seeing significantly Better lenses coming for the R mount. Should I be dumping the old punk lenses now?
 
Upvote 0
Apr 23, 2018
1,088
153
Linear focus in that context probably refers to the actual focus motor. Canon indicates in the white paper that the 24-105 uses NanoUSM with a linear motor.

agree. no more rotational drive, no more gears needed, including manual focus, thanks to focus by wire. so almost no mechanical moving parts. good!

focus group just slides directly driven by a "Ultrasonic/vibration device" in linear motion along a guiding rod (in optical axis , Z dimension).

not sure from the white paper/ illustrations whether there is only one or more guiding rods/drive "actuators" in the rf 24-105. really good lenses gave 2, 3 or 4 rods.

personally i would think 3 spaced at 120 degrees to be most stable solution against any decentering of focus group and ultra-fast operation. will also depend on size/weight of glass to be moved.

i'd expect future rf 2.8 "pro" zooms to come with such 3-rods/3 actuator LEM drives.

Sony and Fuji are doing it already (in some lenses). more info in roger cicala article on LEM drive lenses.
 
Upvote 0
May 4, 2011
1,175
251
Interesting, thanks for sharing.

I was hoping the RF 24-105 would (also) be optically superior to its EF brother, but clearly the optical potential touted didn’t quite make it to that one. Still, it might be good enough for travel use if the size/weight is enough of a reduction, we’ll see.

On the other hand, the 50 1.2 looks awesome. Can’t wait to see actual images from it.
 
Upvote 0
Linear focus in that context probably refers to the actual focus motor. Canon indicates in the white paper that the 24-105 uses NanoUSM with a linear motor. That says nothing about how the lens is programmed to interpret the movement of the focus ring encoder and translate that to driving the focus motor.


Fair enough. I hope this clears the matter further for you then:

 
Last edited:
  • Like
Reactions: 1 user
Upvote 0
Dec 8, 2012
166
28
I reached the same conclusion for the 24-105 after comparing the MTF charts. It is somewhat unfortunate that the resolving power isn't a full step up from the EF mk II version, but it seems that the intention was to reduce size and weight. If not looking for specialty lenses, these benefits will likely be what most people starting out in FF mirrorless photography will appreciate the most.

Be careful when looking at and comparing MTF charts. There was a discussion on the new 600 f/4 that the new charts may be using a different standard. The new 600 III MTF looks bad when compared to the MTF charts released with the 600 II. The RF 24-105 has the new "Look", it only has 4 lines instead of the 8 used in the old charts for the EF 24-105. Not sure what it all means or what's changed. Canon Japan had the new charts for both the 600 II and III and they showed an improvement in the corners for the 600 III.
 
Last edited:
Upvote 0
May 4, 2011
1,175
251
Be careful when looking at and comparing MTF charts. There was a discussion on the new 600 f/4 that the new charts may be using a different standard. The new 600 III MTF looks bad when compared to the MTF charts released with the 600 II. The RF 24-105 has the new "Look", it only has 4 lines instead of the 8 used in the old charts for the EF 24-105.

Going by pages 24-26 in the "white paper", it appears the "new" charts are used in all of the comparisons. By Canon's own words, there are aspects where the RF version is worse than the EF one (and likewise, others where the RF is a little better). We'll have to wait for the tests to see how that translates to actual images, but personally, not too encouraging TBH.
 
  • Like
Reactions: 1 user
Upvote 0
Jan 21, 2011
523
1
I wonder if this provides more of an indication that EF-M is a dead end? The fact that the electronic interface has been updated significantly is an important note. My understanding is that EF-M uses the same electronic interface as EF, with the same pins. Assuming that is the case, EF-M will never support the data transfer that RF is capable of. That means that EF-M will get left behind.
Canon must have acted in the knowledge that people will interpret the EOS R and RF launch as the death knell of the EOS M. - Unless I am reading things totally wrong.
 
Upvote 0
CanonRumors Guy: Thank you for providing that paper/link to the paper.

I was a little bit tired waiting for the release of a FF mirrorless. But it was maybe the quiet before a storm.

Now I am confident that Canon has not left the route of holistically optimized camera SYSTEMS:
- I always hoped that Canon will exploit the freedom to place lens elements in mirrorless systems
and I am satisfied now. I really expect e.g. a compact RF 2.8 16mm or similar with stellar IQ at
a reasonable price of ~ 800 EUR/$ within the next two years!
- EF-adapter with ring to set parameters: Bravo! The left hand has not much to do if you use primes
and AF so very good move.
- Seamless integration of EF lenses is promised and it still works with the EOS M system
- The section about lens optimization parameters shows that Canon wants to offer different mixes
of properties and that the size of a lens is ONE optimization criterion - the RF 1.8 35 seems to be a
great lens (what a pity that it will not work on M50).
- Bringing out a medium specced camera and two high end & two medium end lenses is a good move
providing solutions to different photographic priorities.
- The high quality adapters with sealing - the simple one at just ~100 EUR/$ - is a commitment to the
EF lenses and if I think about a pro photographer who wants to use RF 28-70, RF 1.2 50 +
EF 2.8 70-200 and EF 2.8 300 and use only one mount: Buy two adapters for the EF lenses and you
have effectively one lens mount to work with. 200 bucks is nothing compared to the ~12 000 bucks
for the lenses. And maybe this will be the standard route because there is no mirrorless advantage for
telephoto lenses.

If Canon can put its new products into the market in the next months in non-homeopathic doses they win a large part of the FF mirrorless market.
 
  • Like
Reactions: 1 user
Upvote 0
I wonder if this provides more of an indication that EF-M is a dead end? The fact that the electronic interface has been updated significantly is an important note. My understanding is that EF-M uses the same electronic interface as EF, with the same pins. Assuming that is the case, EF-M will never support the data transfer that RF is capable of. That means that EF-M will get left behind.
Canon must have acted in the knowledge that people will interpret the EOS R and RF launch as the death knell of the EOS M. - Unless I am reading things totally wrong.

On paper, it looks less than ideal. In reality though, the impact is not significant. No one has huge investment in EF-M lenses. Even if they just stay as is, those cameras/lenses remain excellent tool as back up (for its size and quality, and they don't cost an arm and a leg).
 
Upvote 0