MF Zeiss lenses for Sony?

[Jonathan K]

I just read this on photographyblog.com:

"Carl Zeiss are running a teaser advert for a brand new SLR lens mount. Already available for Nikon, Pentax and Leica M42, the un-named mount will be unveiled on September 15th. Speculation is growing that the Sony/Minolta mount is the most likely candidate, possibly coinciding with the Sony A900 pro DSLR… "

http://www.photographyblog.com/index.ph ... _zx_mount/

Is that something new or has it been discussed before? Would it make sense to have a AF Zeiss line-up next to a MF Zeiss line-up?

Jonathan

[David Kilpatrick]

Zeiss distributors have pointed out the value of producing MF, chipped lenses in the Alpha mount to them. It's easily done and it is just possible that the M-AF mount will be 25 years old this year although the camera was only previewed at photokina 1984 - that could mean the patent is 1983, and if I remember correctly, 25 years is the normal term before Japanese dedicated lens mounts become free instead of licensable. I imagine that Sony has new patents referring to SSM and other protocols in the new lenses. Even so, Sigma's decision to put HSM into the Alpha mount could also relate to the age of it, and licence issues.

But - they might also go for the 4/3rds system where their lenses would be appreciated and Leica is getting cold feet, leaving just Olympus and Panasonic to create the micro 4/3rds standard.

David

[PhotoTraveler]

I did a longer response over a Dyxum.

There is only a couple real options. The main one to me, is ZR for Leica R. Leica is bringing an R10, the ZM lenses have hit just right due to the M8. The R10 will really increase demand there. And the simple fact an R lens doesn't require a ROM.

I don't think 4/3rds would pan out. We haven't heard of Zeiss being let in that group. And they would have to figure out how to make the lenses all electronic mount, which is no simple change from what the ZF/ZK/ZS line is. Also the lenses are much bigger than needed. It would be a bit of a hack, but if they were let into the 4/3rds frat, they could get access to all they need.

Minotla has a good chance. I wanted this from the start. The ROM is an issue. Zeiss responded to me saying the ZA lenses are Sony's, Zeiss can't sell A mount lenses not through Sony. You need a ROM for it to work. Now if Sony was smart, they would cut a deal with Zeiss, licenses them the ROM to sell them on their own, but get a cut of the money. And it would mean A mount just gains a whole bunch of lenses. Since those are MF, and the Sony Zeiss's are Manual focus, it would work fine. Also, lets say there is a lens that would work well with the system, but Sony isn't to sure on. This would be a perfect path. Just off the top of my head, lets say Zeiss is well aware of the need for PC/TS lenses. Those are manual focus anyways. They would do great by making some TS and PC lenses for F, K, S, A mount. Sony could license them the ROM for this. It fills a hole for Sony they might not be ready to invest in. I think Sony should do this on anything they aren't ready to do yet, or ever. Just have Zeiss make a manual focus versions with 1st party mount blessed by them. Minolta mount is probably the only way we will see a ROM get in those lenses. A900+SSS+Sweet Zeiss lens, oh yes. And since they could have a rom, we get working SSS, EXIF data. ADI might be the one part the ditch for simplicity. To make a Prime work nicely, they only need a Cira 1985 ROM. Anyone who has ever taken apart an old Minolta prime knows what this means. It's just a chip on a board with 5 pin contacts, no encoders, flex circuits or anything. Same part for all lenses, just burn in the right Lens data/ID. This would be heaven.

The situation that won't happen unless Hell Freezes is Canon. I've been told multiple times by Zeiss point blank there will be no EOS mount lenses because Canon won't license them the ROM. A EOS mount lens cannot function without a ROM, it's all electronic. You also need to adapt an electronic aperture in it to make it work, which would make the lenses very different from the other models. Canon licensing to Zeiss has one very small chance, as Canon would loose all sub 100mm prime sales over night. Plus the PR image for them to suddenly allow that would look bad "Canon admits their wides suck, begs Zeiss for help". Additionally, if Canon didn't Licenses them the ROM, I give no chance of Zeiss reverse engineering it. From there previous statements, and simply if they were to do it, the Nikon and Pentax communities would freak out. As it means they could have had ROM'd lenses too, but Zeiss didn't do it for them. Would be a PR nightmare on those fronts. Zeiss has taken the high road to date and done no reverse engineering, I can't see them starting.

ROMs aren't something that has patent issues. They don't get patented. It's IP hidden in a chip, they don't patent it, which means the info to make it work never gets published. So companies reverse engineer it. So that's not a hold up. Also even if it was open, you run into issues like Lens ID's. A 3rd party can't demand to be issued any. So they high jack ones that exist. That would continue even if everything was Open. We could be 100 years from now and Zeiss wouldn't have the ability to make an EOS lens without reverse engineering it. And so far, from their actions and statements, they don't plan to start.

In the end, I expect it to be ZR, but I really hope it's Z-Minolta (there is no letter available that really makes sense, other than E for electronic or D for Dynax, maybe go with secondary letters. ZI, ZN, ZO...)

The other factor would be their contract with Sony. I would expect Sony struck and exclusive deal on Zeiss AF SLR lenses. And maybe even ROM'd lenses. Sony is probably fine with Zeiss making manual focus stuff, as it's no threat. But soon as Zeiss starts ROM'ing enemy mounts, that could go sour.

[PhotoTraveler]

On the sigma, I don't think it has anything to do with Licenses and such, Sigma is the company apposed to licensing even if offered.

I think the HSM was as simple as lack of market. They had pulled out of the mount. They would have developed their A mount interface kit a long time ago. Updating this would have taken money and for a limited increase in sales. Same for Pentax.

I think the tripping point for them was when Nikon started removing body motors, forcing them to put motors in all their Nikon lenses, and then Sony coming into the mount offering future sales increases, and then finally Pentax going to SDM. With pentax going there. It meant all the lenses they make could go to in-lens motors. Thus they could massively simplify their interfacing for different mounts. Now they only need to support the aperture systems. All the rest is a mount, contacts and a ROM. This means they could design new lenses without the burden of screw drive and in-lens motors. I think when they began to stop making new designs for Pentax and A mount, it was not just that the market was small, but that Sigma had gone to a new generation of architecture in their lenses where they had removed Screw drive capability. When Pentax added SDM, it signaled a good time to make a new interface for A and K mounts. All new Sigma models coming out are HSM. So they have just transitioned for manufacturing simplicity. If all mounts ever get electronic aperture (F, EOS, 4/3rds have it), then that will make Tamron and Sigmas life much simipler. Only K and A lack confirmed support for this. At that time, a 3rd party lens can be identical for all mounts except for the rear mounting flange, contact location and ROM. You would see 3rd parties start doing "adaptall" style systems again, where you could buy the electronic mount for various mounts and screw it onto the lens for what ever you want to use it on.

It might have taken Sigma some time to crack the protocol too. With SSM coming out late, I'm surprised there wasn't any encryption in it or anything else to block 3rd parties.

[Dr. Harout]

Wow, thanks Phototraveler, great explanation.
So in conclusion we MIGHT expect on Sept. 14-15 a tilt-shift lens from Zeiss along with A900.
Few days left...

[David Kilpatrick]

On the sigma, I don't think it has anything to do with Licenses and such, Sigma is the company apposed to licensing even if offered.

I think the HSM was as simple as lack of market. They had pulled out of the mount. They would have developed their A mount interface kit a long time ago. Updating this would have taken money and for a limited increase in sales. Same for Pentax.

Actually, I interviewed Sigma's team at photokina 2006 - and they were completely unaware that Minolta/Sony had ever used SSM. They sent for the Japanese and got together in a meeting while I was there, and then came over with an interpreter and quizzed me about it. Sounds hard to believe, but they had no idea SSM existed, and they tried to argue with me that I was wrong. So their reasons for not putting HSM motors in Alpha mount lenses were not as complex as you imagine - they simply never tried!

David

[Javelin]

Funny thing aboit Sigma. when they were at the Toronto show they had no idea the sony ring light wasn't a flash.

[Jonathan K]

I just read this on photoscala.de:

Nachtrag (2.9.2008; 12:55 Uhr): Eine Designagentur aus dem indischen Bengaluru (einst: Bangalore) will es bereits genau wissen: Demnach wird Zeiss ZE-Objektive vorstellen, sprich: Objektive fürs Canon-Bajonett. Und zwar, so wollen die Designer wissen, mit den für die Datenübermittlung notwendigen elektronischen Kontakten.

Somebody (a design agency from from Bengalore) claims it would be ZE (for Canon) lenses... even with electronic connections for data transfer...
That would be too bad

So could it be that those 25 years are running out for Canon mount as well? Photo Traveler pointet out that Zeiss lenses for Canon mount are unlikely to be released, but isn't the story about licenses being freed after 25 years applicable for EOS mount? They started around the same time, with the A-mount having a slight edge, if I remember.

Jonathan

[PhotoTraveler]

David, I remember your comments on Sigma and SSM, but I don't think that is part of it. Sigma UK may have been without a clue, but I really doubt the folks back in Japan didn't know.

Jonathan, On the licensing, 25 years, 50 years, 1000 years, it doesn't matter. There is no system that forces companies to hand over IP. Unless Minolta or Canon actual filed complete Protocol information for a patent, it will be locked away forever. You only give up what you patent, anything else is yours for ever. That's why manufacturing processes are almost never patented. Age plays no part. And to date, I've never seen anything showing Canon or Minolta Patented their protocols. If they did, that was pretty dumb on their parts. Also if they had it would mean 3rd party lenses for both would work perfect, but they don't. Further companies like Sigma would be in massive Patent infringement on stuff like their SD mount as it uses Canon's protocol (with a modified Pentax K mount).

[Jonathan K]

Hi Photo Traveler,

Then why do you believe it is unlikely that Zeiss will bring out lenses for Canon mount?
Personally I believe it could take away some of the convincing power of ine of the very few asset Sony has to attract professionals: Their Zeiss line-up...

Jonathan

[David Kilpatrick]

David, I remember your comments on Sigma and SSM, but I don't think that is part of it. Sigma UK may have been without a clue, but I really doubt the folks back in Japan didn't know.

Jonathan, On the licensing, 25 years, 50 years, 1000 years, it doesn't matter. There is no system that forces companies to hand over IP. Unless Minolta or Canon actual filed complete Protocol information for a patent, it will be locked away forever. You only give up what you patent, anything else is yours for ever. That's why manufacturing processes are almost never patented. Age plays no part. And to date, I've never seen anything showing Canon or Minolta Patented their protocols. If they did, that was pretty dumb on their parts. Also if they had it would mean 3rd party lenses for both would work perfect, but they don't. Further companies like Sigma would be in massive Patent infringement on stuff like their SD mount as it uses Canon's protocol (with a modified Pentax K mount).

Interesting view on patents requiring publication but protocols relying on their obscurity - I can believe that.

Although the three Sigma people I was talking to initially (looking at the DP-1) were from the UK, they brought the Japanese technical/development guys over because of what I said about SSM, and unless there was some complete misunderstanding, they argued with me that 'Sony does not have HSM!' - and they went away to visit the Sony stand and see the lenses I was talking about. So I really believe they did not know that Min/KM/Sony had a supersonic motor capability in the cameras.

David

[pakodominguez]

Although the three Sigma people I was talking to initially (looking at the DP-1) were from the UK, they brought the Japanese technical/development guys over because of what I said about SSM, and unless there was some complete misunderstanding, they argued with me that 'Sony does not have HSM!' - and they went away to visit the Sony stand and see the lenses I was talking about. So I really believe they did not know that Min/KM/Sony had a supersonic motor capability in the cameras.

David

Last year at the PhotoPlus Expo I asked the Tokina guys about their lenses for A-mount and, after a couple of minutes debating (all in Japanese...) they just said "no". No further explanation, no hope for a "when Sony get a beeter market share". After the look they gave me, I thought they didn't even knew that Sony was making DSRLs!

I think Japanese reps are there to follow the official speach, and any thing that they can not find in their manual, just doesn't exist -that way they avoid "problems".

[PhotoTraveler]

Some more on the mount.

It's unlikely Minolta would have had anything patentable in the protocol to go though the effort there to show that to the public, since mostly what is going on there is just typical communication stuff. There might not be anything patentable to it. Thus all the more reason to just keep it there IP.

Another thing is the mechanical mount itself. People often talk of patents there, but even there, there is limited things that can be patented. The mount is a Bayonet mount, those had existed for decades, so they can't have a patent on it. To get a patent, you have to have something truly new in there, changing some dimensions doesn't cut it to get a unique patent. Though if someone has something patented and you want to do similar, you can often dodge someone's patent by making a minor change.

What Minolta may have patented is the AF drive coupling. That would have been new, but still it has to be found to be different enough from other such mechanism all ready existing. The aperture ring less mount might also have had something new there. But for the most part, there wasn't a lot of truly new stuff there, so it would be pretty hard to patent stuff. They could potential have a patent for "a mechanism for a AF SLR with interchangeable lenses" which does some describing of the basic design. I think Olympus has some such patents on 4/3rds. Minolta would have also filled for trade marks, such as "A mount", "Alpha Mount" and such. Those would still hold today.

Now holding a patent on the bits doesn't fully block 3rd parties. Since the dawn of patents, it's been perfectly legal for 3rd parties to make products that are compatible with that patent. So a company like Sigma can make a lens that mates to A mount without issue. Sigma isn't selling the mount to other, they aren't claiming it as theres. They are making their products conform to what you have, that's perfectly legal, and that's what allows many things to function, say auto parts industry. Third parts also won't call their lenses Alpha lenses or A mount lenses, they will say For Alpha Mount, or For Minolta, stuff like that, so no trade mark infringement.

Now patents don't always require detailed dimensions. Strong patents never do, as the key to a good patent is covering as much as you can which is done via generalization. Thus why patents read so funny with "1+n features to attach to 1+n matching features connected via a device" That ensure you didn't say "attaches via 4 screws", since a person could dodge your patent by using 3 or 5 screws. So they are general. Which means dimensions are often left off, or are just nominal. Which makes it very difficult to reverse engineer, and technically impossible to reverse engineer. I can grab the mounts off all 15 of my lenses and take measurements. But I will never be able to make an A mount bayonet that is true. To make a correct one, I need the original drawings, with nominal dimensions, tolerances, GD&T. Just through measuring, I have no way to know the limits, so I can't make something that is guaranteed to always fit. This is the problem for 3rd party lenses and often why there lenses fit funny, or stick. They have a sample of mounts from minolta lenses, they guess the tolerance from the statistical data they measured. But they never know the real limits. So if you get a body that push the extreme, or you get a 3rd party lens that pushed the limits, you can have a fit issue. This can and does happen. I can get you a mount that will probably work most the time, but it's never certain. Most people don't understand this part, and it's pretty important. You can hand me a part, but I will never be able to reproduce it and have confidence in it, I need the drawing it was made from for that. And companies like minolta never share that. Just like they don't share protocols.

But as the case of Compaq reverse engineering the IBM Bios to make "IBM Compatible" PCs, the courts have long said reverse engineering something to make something work is legal. As long as you follow the right procedures (different group has to do the reversing from the people who do the designing). So 3rd parties have been able to reverse engineer how to make lenses work. But they don't have the official specs, so nothing is ever certain with them. And thus we see Sigma lenses that don't work on new bodies. We see sigma lenses whos gear strip out (sigma would have no way of knowing what the minimum torque capability is for an A mount lens since it's not published).

So nothing legal stops Zeiss from making a Canon lens. They can do it. But there is the public image, business moral side of things. Reverse engineering stuff is not the best public image path. It makes your company look like a Hack. To day, Zeiss has maintained a very high company image. They have designed the mounts for lenses they make. The newer ZF/ZK stuff is mechanical mounts, so they can do a "good enough" path to get the mount right. But they haven't done any ROM stuff. That's sketchier. They have access to it, they have Cosina making those lenses, Cosina knows how to make a lens work for all mounts. Zeiss could pay Cosina for that info and it would be money for Cosina, so they would be fine with it. But clearly Zeiss has had issue with doing that. If they didn't we would have seen EOS lenses long ago, and the Nikon and Pentax mount lenses would have had ROMs too. So if they suddenly put a ROM in those lenses to make Canon mount ones, it means they have had a sudden about face on this. I've talked to Zeiss people in person and through emails, and the response was always the same that they would not reverse engineer the ROM for those lenses. They even had on their website FAQ at one time that they can't make Canon mount lenses because of the ROM licensing issues. Instead they offered the ZS lenses to use with adapters. That line of lenses has faded as Zeiss figured out people would just buy the ZF lenses and use an adapter. Works the same, but can also be used on Nikons. I think they have only ever made 3 lenses in ZS. So while it's possible for them to do it, so far, everything has pointed to them not going there. I know I would be disappointed in them doing this. Canon and Sony probably won't be happy either. Canon probably can't do anything about it though.

The ZF lens sprang over to ZK, which was expected by me, because it was the next logical hop. ZR is the next one, if an R10 comes, it will have the same impact the M8 did on ZM lenses, which will be big sales as you may have noticed Leica R lenses are not cheap. This would also mean all the lenses carry the same basic manual lens construction. If you were a company like Zeiss and could have an engineer design a couple mechanical bits, and update an assembly drawing and have a whole line of new product cheap that sells to a crowed of people with lots of money to burn, and you are anything but struggling to keep up with demand, wouldn't you go for that? Canon mount make spell dollars, but it also has pitfalls.

I suppose there is on other possibility not yet mentioned. Zeiss could be plotting their own SLR. They have the IKON with M mount. Now they work close with Sony. It's not out of the question that they could strike a deal to make their own body off Sony parts. They could try to bring back one of the Contax manual mounts. They don't have the Contax name back yet, and I doubt they can ever produce N lenses. Also N system would be a waste as they would be heading in a market controlled by many companies (AF slrs), and they system flopped the first time, why try again. But they could make a run with a manual body, digital camera. So the new lens could be something there.

I think we all would love it to be for A mount, but I just don't think we are that lucky.

There are rumors the mount is called ZE, I don't know how sound that is or what the basis is. Are people just jumping to conclusions that it means EOS, or Electronic must mean Canon? Electronic could mean 4/3rds. After all, calling 4/3rds lenses Z4 would be a bit funny. It could be simply people thinking it has to be Canon by a process of elimination that was not fully vetted.

[Jonathan K]

Photo Traveler,

thanks for answering in such an elaborate way! Interesting read.
Since I am pretty new to this, let me ask you:

What stops an A-mount shooter to use the M42 (ZS) Zeiss lenses? There are M42 adaptors for the A-mount, if I got it right. And if I understood it correctly, even the ZF lenses don't transfer data, so where would be the advantage of having those lenses in an A-mount version?

Jonathan

[Greg Beetham]

You sure have the fits and tolerances aspect nailed there PT, without those critical allowable max mins a third party lens maker machining a mount is up chits ditch...without a paddle. Even with the best machinery the cutters wear over time depending on how hard the cutters are, how well they are cooled/lubed, how hard (brinell/rockwell) the substance is that is being machined and how well designed the actual machine itself is and how well it copes with repeated cycles (wear and tear in the slides, transmission and transfer lead screws etc.), so you really need those measurements too keep within the max or min allowable tolerance, and be able to identify just when it is that defective units start too appear (bin jobs), making re-calibration and/or retooling necessary.
What does occur too me though is that without access to IP or access to the patent, (there is none) there is always the possibility of scalp hunting, a 3rd party manufacturer could make some outlandish offer too the right employee, or several (one's that posess the information they need) and retain them for as long, or until they are milked of the critical info, I'm being hypothetical here a little as I'm fairly sure that such key personal would have too sign an NDA, on pain of large amounts of litigation, but still, people do move on too greener pastures over time, and then how does one prove that the NDA has been breached at some point....down the road....
I think specialists, good ones, (opticians, engineers, technicians) all over the place are in short supply, going on what I've seen and heard, so it would be a headache for HR just finding and keeping key personall, if true, (I don't think it's just a rumour) that could be one of the reasons why Sony seems to be having a hard time getting lenses out onto shelves in the shops, and also having the odd glitch with QC.
Greg