Canon Says it’s up to Sigma to Make Full-Frame RF Lenses

[Canon Rumors]

In a recent published interview with French publication Phototrend.fr, a pretty big bit of information came from Canon themselves. It doesn't sound like the lack of third-party autofocus lenses for the RF mount has anything to do with Canon. Full Frame Sigma RF Lenses While I touched on it in the interview, I think the […]

See full article...

 

[InternalRevenue]

All of Sigma’s current RF APS-C lenses DO have control rings: for the older ones it doubles up as the manual focus ring (same as all of Canon’s RF-S lenses) and the two newest primes and 17-40 have a separate control ring (which is used as an aperture ring on E and X mount versions). They also work with the in-camera focus scale, digital corrections, full time electronic manual focus options, and more. Older EF lenses don’t work well with IBIS either, so I don’t buy this idea that Sigma and other third parties are having to reverse engineer the RF mount.

None of the Sigma RF lenses offer autofocusing for the entire viewfinder however; anyone know if the other mount also suffer from this limitation?

 

[Canon Rumors]

I didn't notice what day it was.

All of Sigma’s current RF APS-C lenses DO have control rings: for the older ones it doubles up as the manual focus ring (same as all of Canon’s RF-S lenses) and the two newest primes and 17-40 have a separate control ring (which is used as an aperture ring on E and X mount versions). They also work with the in-camera focus scale, digital corrections, full time electronic manual focus options, and more. Older EF lenses don’t work well with IBIS either, so I don’t buy this idea that Sigma and other third parties are having to reverse engineer the RF mount.

None of the Sigma RF lenses offer autofocusing for the entire viewfinder however; anyone know if the other mount also suffer from this limitation?

Thanks. I left out the word dedicated. I didn't touch on the rest of it because none of that is hard. I think it all has more to do with Sigma not being able to manufacture to the scale required. People grossly overestimate how many lenses SIgma sells.

We now know that it has nothing to do with Canon. There will still be the naysayers, but it doesn't matter what you tell some people.. the vacuum can be strong.

 

[DocInfoSci]

Makes sense. I agree, it’s not the contract (license) that’s the issue here but some other business constraint. I was thinking more in terms of limited lens types (doesn’t exclude them from offering a lens, just not all possibilities) but manufacturing capacity is a real consideration.

Still, if just capacity then at worst case just add a control ring and the basic EF instruction set and call it a day for gen 1 so that the issue is more a lens mount swap like prior; go fancy for gen 2. People try too hard sometimes with all at once offerings.

 

[Bob Howland]

I didn't notice what day it was.


Thanks. I left out the word dedicated. I didn't touch on the rest of it because none of that is hard. I think it all has more to do with Sigma not being able to manufacture to the scale required. People grossly overestimate how many lenses SIgma sells.

We now know that it has nothing to do with Canon. There will still be the naysayers, but it doesn't matter what you tell some people.. the vacuum can be strong.

I disagree with your conclusion. Sigma already makes FF lenses that could be ported to RF. They have a mount conversion service costing about $250. I wanted them to convert two APS-C EF-M lenses to RF. They replaced the two lenses that I sent them with brand new RF lenses. They said they had RF M-mount lenses but hadn't received the conversion kits from Japan. As I recall, I got the new lenses on the days that the RF mount version became available in the U.S.

Does anybody doubt that Sigma would be overjoyed if the demand for existing FF lenses (plus a few parts) suddenly increased by, wild guess, 30%? They already know how to make lenses for the RF mount. They do it every day.

 

[Canon Rumors]

I disagree with your conclusion. Sigma already makes FF lenses that could be ported to RF. They have a mount conversion service costing about $250. I wanted them to convert two APS-C EF-M lenses to RF. They replaced the two lenses that I sent them with brand new RF lenses. They said they had M-mount lenses but hadn't received the conversion kits from Japan. As I recall, I got the new lenses on the days that the RF mount version became available in the U.S.

Does anybody doubt that Sigma would be overjoyed if the demand for existing FF lenses (plus a few parts) suddenly increased by, wild guess, 30%? They already know how to make lenses for the RF mount. They do it every day.

"Ported" and being able to actually make enough are different things. A lens like the 135 f/1.4 would be a very difficult lens to make within tight tolerances. Sigma made nothing worth owning for most of the EF days. When they did start making things that were sort of nice there were already 100+ million EF lenses. Having worked directly with Gentec, I know how few lenses they sold in the grand scheme in at least one market (maybe 2). It was peanuts.

Demand increasing 30% is great, now you have to scale everything else 30%. There's a point in lots of businesses where the next step costs exponentially more than the previous one.

Take a company like Leica (yes, they do it different). They make 40 cameras a day. Major products usually take about a year to meet the demand. How much would it cost them to increase production to 60 a day? A boatload. Sigma is closer to Leica than they are to Canon. Leica's revenue is actually higher than Sigma's.

 

[[email protected]]

Canon's ways of limiting Sigma are many and complex, not least because the intellectual property levers they have operate differently in different countries. In the US, reverse-engineering a standard is usually fine. In other countries it is sometimes not. One of the things that some of the Chinese vendors got into trouble over was claiming "RF Mount" branding on their packaging. Viltrox and Samyang may have had more trouble with the copyright/trademark (different mechanism) problem than with the an engineering/mount/patent issue. In the past, we've heard narrowly-stated comments from European Canon heads that turned out to be technically true, but not the whole truth. I recall a German executive talking about this issue back in the early RF days, and everyone got needlessly excited.

Then there's the business relationship issue. Most OEM camera companies use Tamron and Sigma to provide elements of lenses, or even whole lenses. Those OEM relationships are quite important to the companies, and it's not surprising that the little firms like Viltrox were the ones making RF lense while the ones with OEM contracts aren't.

Sigma's CEO has spoken around the issue in several interviews, most notably the recent one where the Petapixel crew visited him in Japan. My impression was that he wasn't able to speak because of an active provision that prevented him from speaking. He's a stand-up guy, and I'd put more faith in what he says (or indicates he can't say) versus a whole sack of OEM company executives.

 

[Canon Rumors]

Sigma's CEO has spoken around the issue in several interviews, most notably the recent one where the Petapixel crew visited him in Japan. My impression was that he wasn't able to speak because of an active provision that prevented him from speaking. He's a stand-up guy, and I'd put more faith in what he says (or indicates he can't say) versus a whole sack of OEM company executives.

An impression is a lot different than what he actually said. He has never directly addressed anything on the topic. Sure he's a nice guy, why does he get more credit for not saying anything? Insinuating that an executive at Canon would lie when he could have just said the usual "we don't comment" that they're quite good at.

Canon says, there's no differentiation between APS-C/FF lenses, they don't collaborate at all on products and they have no idea what Sigma's roadmap would be. They said three direct things on the topic, but Mr Sigma is a nice guy.... so....

Someone at Canon Europe says something, it's probably more a guess than actual knowledge. Canon Inc. isn't opening up about their business to subsidiaries. They buy gear from Canon Inc, and then they go sell it. That's it.

OEM relationships wouldn't play a role at all, unless there was transparent exclusivity, which there obviously isn't. Tamron does stuff for Nikon, Tamron also does a bunch of other things. Nikon is closer to exclusivity with an eyewear company than anyone in imaging. Cosina/Voigtlander make elements (and entire lenses) for a pile of companies, they don't seem to be restricted from doing whatever they want. They compete directly against a brand they make lenses for, in the exact same space. No business would sign on to restrict themselves.

Look at the L-Mount alliance, none of those companies are restricted from doing whatever the heck they want. They are pooling some resources like is done in tons of industries, and then directly competing with each other.

 

[Kingsix88]

If I remember correctly, when Canon first came out with the RF mount, these was something strange with the image on third party lens. Unexpectedly strong vignetting and purple hue. I believe it has something to do with canon sensor technology and mirrorless cameras having much shorter distance between rear element and sensor. So light rays hit the sensor at much sharper angle, compared to dslr. Sony didn’t have that issue. It is possible canon mount requires different lens design in certain focal lengths for adequate performance. This is not worth it for a third party lens manufacturer to do. Maybe crop doesn’t have the same issue.

 

[neuroanatomist]

If I remember correctly, when Canon first came out with the RF mount, these was something strange with the image on third party lens. Unexpectedly strong vignetting and purple hue. I believe it has something to do with canon sensor technology and mirrorless cameras having much shorter distance between rear element and sensor. So light rays hit the sensor at much sharper angle, compared to dslr. Sony didn’t have that issue. It is possible canon mount requires different lens design in certain focal lengths for adequate performance. This is not worth it for a third party lens manufacturer to do. Maybe crop doesn’t have the same issue.

I don’t think so. When the RF mount first came out, there were no third-party lenses for it. That is a recent development, and only for crop lenses.

What you are referring to are issues with peripheral image correction on third-party EF lenses, those were caused by the camera incorrectly identifying the lens. Such lenses spoof Canon lens ID numbers, and the RF mount made that problematic.

Since those were EF lenses, they required the mount adapter, and thus the distance from the lens to the sensor was the same as that on a DSLR. Optics was not the problem, electronics/software was the issue.

 

[Dragon]

So a question. The RF mount clearly provides for extensive geometric correction (and I believe even includes correction for focus shift) in-camera and the data to support that correction clearly resides in the lens, since I don't have to upgrade my camera firmware every time Canon comes out with a new lens. The question is, do the Sigma and Tamron APS-c lenses supply that correction info the camera or just depend on being good enough optically to not perceivably need in-camera correction.? Lack of that tech would be a much bigger deal for FF than for APS-c. It is pretty clear that Even Canon has not yet exercised all the features of RF mount, so until there is a very large body of Canon bodies and lenses to evaluate, it would be very risky for 3rd parties to release lenses that could be rendered obsolete with the next camera release.

I think most of the folks complaining have been assuming that a licensing deal includes access to the technology and the comments you refer to clearly refute that. That said, figuring out the protocol may be much harder that some think, since the communications could well be encrypted. That pin that switches back to EF protocol may be more complex than many think. It is notable that the folks who got shut down (Samyang, et al) were simply using EF protocol with an RF mount and Canon would have been perfectly justified in zapping them for misrepresentation.

 

[Kiton]

I am onboard for a
200-400 f4 no built in converter
2nd place a 100-500 5.6 constant aperture.
3rd place their 300-600 f4, a bit overkill for NHL but would be good for football and soccer!

 

[zapper]

I did some searching for lens sales numbers. An estimate for Sigma is ~1 million/year, across all mounts, including legacy DSLR. For Canon, ~7 million/year. So your hypothesis that Sigma's production capacity is a limiting factor seems plausible. In that scenario, the Sigma RF-S lenses make sense, as Sigma stands to sell many without competition from Canon. I've two of their RF-S lens, including the 18-50mm f/2.8, but I wouldn't have bought it if Canon had an APS-C 16-50mm f/2.8 as Nikon does.

 

[Martin0815]

"Ported" and being able to actually make enough are different things. A lens like the 135 f/1.4 would be a very difficult lens to make within tight tolerances. Sigma made nothing worth owning for most of the EF days. When they did start making things that were sort of nice there were already 100+ million EF lenses. Having worked directly with Gentec, I know how few lenses they sold in the grand scheme in at least one market (maybe 2). It was peanuts.

Demand increasing 30% is great, now you have to scale everything else 30%. There's a point in lots of businesses where the next step costs exponentially more than the previous one.

Take a company like Leica (yes, they do it different). They make 40 cameras a day. Major products usually take about a year to meet the demand. How much would it cost them to increase production to 60 a day? A boatload. Sigma is closer to Leica than they are to Canon. Leica's revenue is actually higher than Sigma's.

Sorry, I have the completely opposite opinion.
Why should Sigma (or any other 3rd party lens manufacturer) be 'required' to make enough RF lenses to fulfill a very high demand in short time?
Sigma would simply sell as many lenses as they can make, be very happy with the revenue, and later they can decide to increase production capacity if the demand justifies it.

In a free market, any supplier can offer as many product units as he wants, while there is no obligation to produce a very large quantity, even if an initial short supply might disappoint some prospective buyers. After a while production capacities will adjust to the demand.

Additionally, all RF and RF-S cameras use the same RF mount, only the image circle of an RF-S (APS-C) lens is smaller than FF.
Consequently, there is no additional 'technical difficulty' to overcome or any 'reverse engineering' needed for FF RF compared to APS-C RF-S. Note that there are already several 3rd party autofocus RF-S lenses available, e.g. Sigma 15mm f/1.4 DC, so 3rd party manufacturers have already solved the 'technical difficulty'.

 

[zapper]

So a question. The RF mount clearly provides for extensive geometric correction (and I believe even includes correction for focus shift) in-camera and the data to support that correction clearly resides in the lens, since I don't have to upgrade my camera firmware every time Canon comes out with a new lens. The question is, do the Sigma and Tamron APS-c lenses supply that correction info the camera or just depend on being good enough optically to not perceivably need in-camera correction.? Lack of that tech would be a much bigger deal for FF than for APS-c. It is pretty clear that Even Canon has not yet exercised all the features of RF mount, so until there is a very large body of Canon bodies and lenses to evaluate, it would be very risky for 3rd parties to release lenses that could be rendered obsolete with the next camera release.

My Sigma RF-S 10-18 f2.8 and 18-50 f2.8, when mounted on my R7, show "Not available with the attached lens" for distortion correction, and "Cannot correct - no data" for Digital Lens Optimizer, in the Lens aberration correction menu. This doesn't bother me much, as I shoot raw and process in Lightroom, which does apply the corrections.

 

[Ron888]

Has no one noticed what day it is? lol

 

[Dragon]

My Sigma RF-S 10-18 f2.8 and 18-50 f2.8, when mounted on my R7, show "Not available with the attached lens" for distortion correction, and "Cannot correct - no data" for Digital Lens Optimizer, in the Lens aberration correction menu. This doesn't bother me much, as I shoot raw and process in Lightroom, which does apply the corrections.

I think that answers my question. Unless the lens is relatively distortion free, that missing bit, while not much of an issue for a RAW still shooter, is a big deal for video. Simply due to lens size implications, this would be a bigger issue for FF lenses than for APS-c. A number of Sigma's EF lenses were quite good, and relatively distortion free, they were also big and heavy. Given that Canon is taking full advantage of electronic correction, that gives them both a size and cost advantage over a third party that is obliged to do all correction optically due to lack of access to the electronic correction capabilities of the camera. As an aside, I have a fair number of 3rd party EF lenses that will not work correctly on either RF or EOS-M mount cameras and all my Canon EF lenses work normally on both mounts, so even the EF protocol had some undiscovered features that the 3rd parties didn't catch and now have no interest in fixing. DPAF definitely messed with the AF algorithms that both Sigma and Tamron were using. FD lenses work fine on RF .

 

[Dragon]

Has no one noticed what day it is? lol

True, but the relevant conversation with the Canon exec was reported well before April 1, so the article is legit.

 

[mclaren777]

I would gladly buy the Sigma 300-600mm if they released it in either EF or RF mounts.

 

[AJ]

I'm still scratching my head.
1) Sigma already has spent the R&D money to develop full-frame lenses like 300-600/4
2) Sigma already has spent the R&D money to figure out the RF mount, as evidenced with their RF-S lenses.

So why aren't they putting 1 and 2 together? I don't see this being a Sigma issue. It seems logical the issue lies elsewhere.