Canon announced the RF 70-200 f/2.8L in February 2019. This is notable for being much smaller than the equivalent EF lens, being perhaps two thirds of the length, yet maintaining the speed of the EF version - see for example the fourth picture on this page.

My perhaps naive understanding was that while the short focal plane distance of mirrorless cameras allowed for smaller retrofocal lenses, the advantage was mostly lost for telephoto lenses. Have I misunderstood this, or have Canon just optimised the RF lens for size, thus presumably giving up something else - if so, what?

(Canon have confirmed that the lens does not use diffractive optics - "We decided not to use Diffractive Optics with this lens" - so that's not the answer this time)

  • \$\begingroup\$ Agreed. Poring over the patent claim there seem to be just a clever stacking of alternating convergent/divergent lens groups with some tight positional relationships: when zooming, 6 lens groups are moving... \$\endgroup\$
    – xenoid
    Mar 18, 2019 at 13:05
  • \$\begingroup\$ It's only shorter at 70mm. At 200mm it's actually longer. \$\endgroup\$
    – Michael C
    Mar 19, 2019 at 11:34

3 Answers 3


They are two totally different designs.

  • One is a non-extending lens that does all of the "zooming" internally.
  • The other is an extending zoom lens that is considerably shorter at 70mm than at 200mm.

In fact, the RF 70-200mm f/2.8 L IS is slightly longer than the EF 70-200mm f/2.8 L IS III when the RF lens is zoomed all the way in to 200mm.

  • 1
    \$\begingroup\$ Yep, this is the point I missed. Probably because Canon "accidentally" forgot to mention it in their press releases about the lens... have you found any photos of the lens in the fully extended state? \$\endgroup\$
    – Philip Kendall
    Mar 19, 2019 at 11:46
  • \$\begingroup\$ I haven't really looked. Every "rumor" I've seen regarding this lens says it is an extending design. No one claims that it does not extend. \$\endgroup\$
    – Michael C
    Mar 19, 2019 at 11:53

In its current design, it was achieved by using an extending lens body.

A recent interview in DPReview with Canon's executives states:

Q: In terms of making the new 70-200mm smaller, how difficult was it to decide to move to an extending zoom design?

A: We've not actually disclosed that it is going to extend or not, but we do have the extending mechanism in other lenses we make. So we do have the experience and know-how in-house: we have the capability to introduce such a feature in a new lens. To be clear, though: I have not said that it's going to be an extending design.


A new video from Canon Korea gives us a closer look at Canon’s upcoming RF L F2.8 trinity and confirms what we’ve more or less known with the upcoming RF 70-200mm F2.8 — it won’t be an internal zoom.

DPReview: "Canon Korea promo video confirms the RF 70-200mm F2.8 lens won't have internal zoom"

Lens-rumors.com claims that the US-patent #20190004296 is for the RF 70-200mm F2.8L IS USM. This patent states:

Focal length [mm]       | 72.00  | 135.00 | 194.99
Length of the lens [mm] | 172.73 | 227.50 | 242.86

Found in the patent application's p. 7

The official length for the EF 70-200mm F2.8L IS USM III is 199.0 mm and it is not extending.


enter image description here Stolen from ephotozine's hands-on article

I further browsed through the illusive patent application and I found three different tables for lens sizes:

[Numerical data 1] Zoom ratio 2.71
Focal Length     |  72.00 | 135.00 | 195.00
F-Number         |   2.87 |   2.91 |   2.96
Tot. Lens Length | 172.73 | 227.71 | 242.23

[Numerical data 2] Zoom ratio 2.71
Focal Length     |  72.00 | 135.00 | 194.99
F-Number         |   2.92 |   2.92 |   2.92
Tot. Lens Length | 172.73 | 227.50 | 242.86

[Numerical data 3] Zoom ratio 2.71
Focal Length     |  72.00 | 135.00 | 194.98
F-Number         |   2.90 |   2.90 |   2.90
Tot. Lens Length | 172.73 | 218.46 | 231.71

[Numerical data 4] Zoom ratio 2.71
Focal Length     |  72.00 | 135.00 | 194.98
F-Number         |   2.89 |   3.15 |   3.35
Tot. Lens Length | 207.73 | 207.72 | 207.71

[Numerical data 5] Zoom ratio 1.95
Focal Length     | 100.00 | 150.00 | 195.00
F-Number         |   3.93 |   4.02 |   3.96
Tot. Lens Length | 167.73 | 207.60 | 231.68

As @MichaelC states in his comments, it is probable that these are just the designs that Canon has tried out before applying the patent - they say little to nothing about the lens to come.

@JohannesD pointed out that the numbers in the patent application include the flange focal distance, i.e. they are measured from the front of the lens to the sensor. Since the FFD of the RF mount is 20mm, we can subtract them to get the lengths of the designs.

Note that the only one that is smaller than the EF lens is #4 - 199 vs. 187.7mm. Interestingly, this seems to be an (almost) non-extending design.

  • \$\begingroup\$ Simply comparing the lens diagram (global.canon/ja/c-museum/wp-content/uploads/2015/05/…) ... it's a whole different beast. I've no idea how to interpret the changes but, it sure is pretty. \$\endgroup\$
    – OnBreak.
    Mar 18, 2019 at 18:24
  • \$\begingroup\$ Products do not have to match the dimensions of the patent. they have to match the design elements that make them unique from other, previously patented, design elements. \$\endgroup\$
    – Michael C
    Mar 19, 2019 at 11:38
  • 1
    \$\begingroup\$ "Total lens length" in an optical patent text always includes the flange distance; it's the distance from the front element to the image plane. Subtract 20mm from the lengths mentioned in the patent to get more comparable measurements. \$\endgroup\$
    – JohannesD
    Mar 30, 2019 at 18:46
  • 1
    \$\begingroup\$ @MichaelC Definitely! But here the length of the EF 70-200mm that was mentioned (199mm) was its actual physical dimension. \$\endgroup\$
    – JohannesD
    Apr 5, 2019 at 14:28
  • 1
    \$\begingroup\$ @JohannesD incorporated that information in my answer. thanks! \$\endgroup\$
    – flolilo
    Apr 5, 2019 at 14:37

It is possible to image using a modest single element lens. Sorry to report, the resulting images will be second-rate. That’s because all lenses suffer from aberrations that degrade. Opticians mitigate aberrations by combining numerous lens elements. Some are positive (convex) and some are negative (concave) as to power. Additionally some are cemented together; others are air-spaced. It takes all this to mitigate aberrations. Nevertheless, residual aberrations always remain.

If the camera were to be fitted with a single element lens and focused on a distant vista, we could take a measurement from the center of the lens to the image plane. This value is the focal length. In a complex lens array, finding the point to make this measurement is more obscure. The point we need to find is called the rear nodal.

Opticians can and do shift the position of the rear nodal. Now a long lens is one that has a long focal length. The longer the focal length, the more magnification it will deliver. A long lens is very desirable if you are into sports or wildlife or the like. However, you might find a long lens to be somewhat awkward.

Opticians have a trick up their sleeve that physically shortens the lens barrel. This is accomplished by shifting the rear nodal forward. If the optician desires, a complex array of lens elements can be constructed so that the rear nodal falls in the air, forward of the front element.

Remember, the focal length is a measure taken from the rear nodal to the image plane. The advantage of such a design is a shorter, less awkward barrel length. Let me add, a true telephoto design differs from the long lens in that the telephoto is foreshortened as to barrel length.

Also, you should know that short wide-angle lenses often place the rear lens group too close to the image plane. If true, there is no room for the mirror mechanism of the SLR. The optician, desiring more room for the back-focus distance, will shift the rear nodal rearward.

  • 1
    \$\begingroup\$ Your wordy explanation might be interesting...but it doesn't really answer the question in any way. \$\endgroup\$
    – osullic
    Mar 17, 2019 at 16:57
  • \$\begingroup\$ This does not answer my question. Why could Canon design the RF version be shorter than the EF version? \$\endgroup\$
    – Philip Kendall
    Mar 17, 2019 at 16:58
  • 2
    \$\begingroup\$ The RF version sports a shorter barrel which is less awkward to use. This shorter design is accomplished by shifting the rear nodal forward away from the center point of the lens barrel. \$\endgroup\$ Mar 17, 2019 at 21:09
  • 1
    \$\begingroup\$ It sounds like you are guessing why rather actually knowing. \$\endgroup\$
    – Eric S
    Mar 18, 2019 at 1:21
  • 1
    \$\begingroup\$ I am unable to know about the operation of their mind. In other words, your guess is as good as mine. \$\endgroup\$ Mar 18, 2019 at 3:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.