SIGMA RF mount lens information finally coming in February 2024? [CR1]

[DhlcadR6]

Are you suggesting that eye detect has a pinch of CDAF on top of DPAF to get more precise focus and thus is sending fine tuning commands to the lens rather than a simple "go to command" that is typical of PDAF systems or just that the repetitive nature of servo AF gives better odds? For reference, the CDAF on the 5DSR in live view seems to make almost all 3rd party lenses behave, probably due to the iterative behavior of CDAF.

For RF cameras, I don't think they have CDAF? They are all PDAF-based AF system. I don't have 5Dsr and only the 80D which is the closest camera to your 5DSR's AF system. My limited finding is the dedicated AF modules on DSLRs has too much focus shift for non-1st party lenses. Once switch to mirrorless mode, regardless of CDAF or PDAF, it will be accurate.

 

[David - Sydney]

The EF-M sigma 56mm focuses just like Canon EF-M lenses. That’s the only datapoint I have for recent Sigma lenses.

EF-M are mirrorless.. uncertain how much of a difference that makes but the original claim was for DLSR/lenses

 

[Surab]

They have talked about it.
It is not as simple as people think it is but it is what they do.
Sigma designs for the L mount and E mount at the same time.

Sigma does indeed but I'd assume they design for the L mount with 20mm and just elongate the tube by 2mm for E mount. Their designs should be quite straightforward to transfer to RF, unlike Tamron.

That's the beauty of the Z mount: biggest mount and shortest flange distance.

 

[Del Paso]

Which one?

None!

 

[SNJ Ops]

Sony kind of cripples 3rd party lenses though.
To be fair, they also cripple non-GM Sony lenses.
The only Sigma lens that I would pick over a GM lens right now is the 14 f/1.4.
Although, the Sony 14 f/1.8 GM is perfect for what it is.

Its mix of Sony limiting 3rd party lenses to a max of 15fps even if they have linear motors that handle more (Sigma’s recent releases) and some lenses having older style AF motors that aren’t able to shoot very fast in terms of frame rates full stop.

Most of Sony’s bodies actually can only shoot up to 10fps (something they really need to improve)so the limitation only applies to the A9s and A1.

 

[Walrus]

Are you sure that Sigma AF for DLSRs are as good as Canon native EF lenses released relatively recently??

Let me just say "it's good enough to not feeling any limitation or difference whatsoever".

Explanation: on 6D i had Sigma 50 & 135 Art and the Tamron 24-70 2.8 VC G1.
On R6 I switched the Tamron (all G1 lenses AF is known as being sketchy on Canon R, and certainly mine was; only after selling I discovered that Tamron service could flash the firmware) for the Sigma 24-105 Art, and later I got a very lucky shot in the used market and switched the Sigma 24-105 for the Canon EF 24-70 2.8 L II, which is also the most recent EF lens I purchased and used extensively, and so it's my term of comparison for "how EF stuff should behave".
Then, later I sold the Sigma 50 and 135 Art and bought Sigma 40 & 85 Art.

So I had a decent amount of use of Sigma recent primes on both DSLR and ML, and use of recent Tamron zoom on DSLR and Sigma zoom on ML; and of course extensive use of recent (24-70 II) and less recent (70-200 non-IS) EF lenses on both DSLR and ML, and use of TON of stuff I had and used over the last 24 years, both Canon (24 STM, 35 f2 40 STM, 50 1.4, 50 1.8, 85 1.8, 100 f2, 135 f2, etc etc etc) and third party (Tokina 11-16, 16-28, various old Sigma non-Art)

From my personal experience, and I mean "personal", so let's not get to absolutes because I've no interest in getting into cat fights, I can surely say that, excluding the Tamron 24-70 on ML, any recent third party lens I used (which, excluding that Tamron on ML, basically means just Sigma Art and the Tokina's on DSLR) showed no difference of AF precision, speed and general reliability (lost contacts between camera and lens, IS issues, mech issues, etc) compared to my reference EF 24-70 II.
In a blind test, in my experience, and for what I do and how I use camera and lenses, I can't tell apart the EF 24-70 II from any Sigma art lens I owned, and can't tell the Canon apart from that Tamron when used on DSLR.
As I can't tell apart any recent (Canon or third party) EF lens from any RF lens (16, 35, 50 and 85, all STM) I owned and used in the last 2 years. EDIT: no, wait; actually I can tell apart the 85 STM AF that was really REALLY slow and hunting, especially when backlight, that's why I resold it at the speed of light.

Of course YMMV, but frankly of ALL above mentioned, the only thing I bought new is the 40 Art (and bought it online), all the rest comes used from eBay, so I wasn't going to the shop and asking to test 3 or 4 Sigma copies and come home with the golden lens. So can't believe I'm so lucky I got all perfect lenses, and all the rest of the world got terrible lenses plagued with AF inconsistency.

Then, I do studio & location portraits, and weddings; not doing BIF's, not doing sport, etc, and I shoot at 1fps; if there's a difference in "extreme" conditions, meaning super fast subjects and/or super high framerates, where third party lenses loses their grip, that's another story, and I don't discuss that there can be a difference there.
But for what I do, no, any Sigma Art I owned is or was as good as my 24-70 L II.

 

[Walrus]

I don't think any EF-RF translation is needed. R bodies speak EF protocol when one of the additional lines is terminated in a particular way. The catch is that such a lens could not take advantage of any of the new RF features like faster communications and cooperative IS and that would put a third party lens at a serious disadvantage relative to a Canon lens.

If Canon would say to third parties "guys, you can manufacture anything you want with RF mount, electronics and working AF, but it has to be EF protocols forever and ever, we will never give you access or space for reverse engineering on RF protocols", well, I'm pretty sure all the manufacturers will be happy, and also all the customers, too.
Yes, it's limited, but it really is? Depends on what you do, speed of communications and cooperative IS may be something you really don't feel difference at all.
Then, if I buy a Sigma lens but "max fps with AF are capped to 4/5fps instead of the 10/20/30fps camera can do", then I don't find it limiting, I know I'm buying something off brand and probably cheaper, so I cannot complain if my lens has some limitations compared to genuine stuff.
But as long the AF is snappy enough and locks onto the subject, frankly speed of communication, cooperative IS and sh*t is not something I would really care at all, for 95% of people that translates into marketing claims, but I really want to see, in real life, how many people REALLY benefit from the "faster communication given by the new extra contacts on RF mount". That's (mostly) marketing BS.

 

[koenkooi]

Its mix of Sony limiting 3rd party lenses to a max of 15fps even if they have linear motors that handle more (Sigma’s recent releases) and some lenses having older style AF motors that aren’t able to shoot very fast in terms of frame rates full stop. [...]

What do AF motors have to do with shutter fps? I can get 40fps when using the MP-E65mm on my R8 and that lens has no AF motors at all. I do know that Canon limits fps on lenses with slow aperture motors (e.g. EF180L), but that actually makes some sense.

 

[koenkooi]

[...] But as long the AF is snappy enough and locks onto the subject, frankly speed of communication, cooperative IS and sh*t is not something I would really care at all, for 95% of people that translates into marketing claims, but I really want to see, in real life, how many people REALLY benefit from the "faster communication given by the new extra contacts on RF mount". That's (mostly) marketing BS.

Having used both the EF100L and EF100-400L II on EF and RF and then switched to the RF100L and RF100-500L, I can say that the coordinated IS is a big improvement over both DLSR+EF lens as well as RF-IBIS+EF lens.

 

[Walrus]

Having used both the EF100L and EF100-400L II on EF and RF and then switched to the RF100L and RF100-500L, I can say that the coordinated IS is a big improvement over both DLSR+EF lens as well as RF-IBIS+EF lens.

Coordinated IS is a thing, I agree, but "faster communication given by the new extra contacts on RF mount" is mostly marketing. Yes, more speed, but 95% of times you don't really feel it.

But my piece was different: the question was "if you could buy an AF third party lens with RF mount, with same optical quality of the Canon equivalent, missing coordinate IS, less comm speed, fps are capped to half, etc, but for half of the price, would you buy it?".

You probably wouldn't, but may others would, included me.
I'm interested in RF third parties for the optical formula, eventually size and weight, or missing focal lengths offers (35-150), and of course price.
I then would be perfectly fine with them having EF protocols, that work wonders.
For you that's limiting, from me and surely many others may not be limiting at all; in the end I'm buying off brand, I certainly expect to have limitations. I actually may DESIRE limitations, so the gear is cheaper. If Canon gives away RF protocols they'll do for a fee, and so third parties end up being more expensive then if they were using EF protocols.

 

[SNJ Ops]

What do AF motors have to do with shutter fps? I can get 40fps when using the MP-E65mm on my R8 and that lens has no AF motors at all. I do know that Canon limits fps on lenses with slow aperture motors (e.g. EF180L), but that actually makes some sense.

I could be wrong but maybe certain AF motors can’t handle above a certain number of AF calculations from the body. Not all Sony lenses are capable of shooting 20fps either so it isn’t just 3rd parties that have limitations.

Its one thing to get a high frame rate but quite another to get a high frame rate with images that are in focus. Why the stacked sensor bodies and higher end lenses cost more.

 

[EOS 4 Life]

@EOS 4 Life was saying that cleanroom reverse engineering isn't possible after 1994 which seems strange to me

A cleanroom protects from copyright.
It is no protection whatsoever from a patent.
You do not need to even be aware of a patent to violate one.
That is why patents expire sooner than copyrights.
Sorry, it was 1995.

In re Beauregard, 1995 - Malhotra Law Firm, PLLC

Beauregard was a very important case in the History of Software Patents. In re Beauregard is the case that opened things up completely for software patents. It signaled the end of the U.S. Patent and Trademark Office’s resistance to allow patent applications that contained algorithms...

www.patentsusa.com

 

[EOS 4 Life]

I'm pretty sure all the manufacturers will be happy, and also all the customers, too.

I am pretty sure manufacturers would complain and going by here customers are never happy.

 

[EOS 4 Life]

Sigma does indeed but I'd assume they design for the L mount with 20mm and just elongate the tube by 2mm for E mount. Their designs should be quite straightforward to transfer to RF, unlike Tamron.

That's the beauty of the Z mount: biggest mount and shortest flange distance.

That is a good point.
Porting E-mount lenses to Z-mount is pretty straightforward.
The RF mount would not be.
It should be easier for Sigma to bring over L-mount designs.

 

[neuroanatomist]

If Canon would say to third parties "guys, you can manufacture anything you want with RF mount, electronics and working AF, but it has to be EF protocols forever and ever, we will never give you access or space for reverse engineering on RF protocols", well, I'm pretty sure all the manufacturers will be happy, and also all the customers, too.

But my piece was different: the question was "if you could buy an AF third party lens with RF mount, with same optical quality of the Canon equivalent, missing coordinate IS, less comm speed, fps are capped to half, etc, but for half of the price, would you buy it?".

You probably wouldn't, but may others would, included me.

Why would Canon would be happy about that? Or do you think it's Canon's goal to make their competitors and customers happy? It seems that you don't know much about business.

 

[EOS 4 Life]

Tbh is the problem for all the 3rd parties, the protocol they have is not entirly compatible with late EF era DSLR or RF MILC.. With eyeAF/subject detecti AF on it solves 70~80% of the problem.

I have zero problems with newer Canon EF lenses.
The old ones were not so great on DSLRs either and are better on mirrorless.
My Sigma and Tamron lenses are not nearly as reliable.

 

[Walrus]

customers are never happy.

Unhappy customers make companies work and improve, and help competition between companies; so as a customer I feel it's a duty to be critic and pretend always more from the companies I buy stuff from, them being camera companies, car companies, or any other.

 

[robinlee]

14mm f/1.4

 

[dtaylor]

For RF cameras, I don't think they have CDAF? They are all PDAF-based AF system. I don't have 5Dsr and only the 80D which is the closest camera to your 5DSR's AF system. My limited finding is the dedicated AF modules on DSLRs has too much focus shift for non-1st party lenses. Once switch to mirrorless mode, regardless of CDAF or PDAF, it will be accurate.

That has not been my experience with either the 7D, 5Ds, or R6. The pre-ART Sigma 50mm f/1.4 had too much play in its focusing for 100% DSLR PDAF accuracy, which is one of the reasons I sold it. But my Tamron latest generation SP lenses (70-200 f/2.8 G2, 35 f/1.4, 45 f/1.8, 90 f/2.8 macro) all focus as accurately as my Canon glass using both DSLR PDAF and R6 DPAF. And the 70-200 is one of my fastest focusing lenses. The 70-200 did require MFA, but has been perfect after that. The three primes have rather long focus throw, so they can take a moment to rack from one end to the other, but they nail focus consistently.

Not sure if this is relevant or not, but only one of those Tamron lenses works on the 35mm EOS 3, the 45mm f/1.8. But the problem does not appear to be AF. They seem to have a problem reporting and/or responding to aperture commands. Not sure what would have changed between 1998 and later bodies.

 

[dtaylor]

Having used both the EF100L and EF100-400L II on EF and RF and then switched to the RF100L and RF100-500L, I can say that the coordinated IS is a big improvement over both DLSR+EF lens as well as RF-IBIS+EF lens.

Every one of my EF IS lenses, including the Tamron lenses, are more stable by 1-2 stops on the R6. And that's not just pixel peeping 20mp vs 50mp, it's quite obviously a difference in usable shutter speed, i.e. if I had a 21mp 5D3 the blur would be visible at shutter speeds that are sharp on the R6. I'm not sure what's left on the table if I'm hitting 6-7 stops combined IS already.