It's the same as it ever was. There will always be third-party companies making things on their own, for their own benefit. They've done it with camera lenses, camera flashes, and camera bodies. And they've always been stealing away purchases that would otherwise have gone to the primary camera manufacturer. To think that all of a sudden, the camera manufacturers will rise up and squash the third-party companies is non-sense. Things are going to continue to operate as they always have. And I don't think the big camera companies have any other recourse other than what they have always done in the marketplace...work to convince buyers to stick to name-brand products.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Canon mount Foveon
- Thread starter N/A
- Start date
Jimmy Chen
Senior Member
That patent has been around for quite awhile. I understand how it works, but I just cannot see how it is going to work in real life. What is that sensor going to do when you want to take an 1/10 second motion shot? 1/10 second exposure total, which means each layer only gets 1/20 a second, or 1/10 second to each layer which means the camera requires 2/10 seconds to take the shot.
Either case, you will not get the motion you have within 1/10 second. Unless someone can figure a way to have a double exposure within the same 1/10 second, I just cannot see how it is going to work.
--
jc
Jimmy Chen
Senior Member
Jimmy Chen
Senior Member
It has been said that S produces chips for N because N just happens to sell S their chip fabrication machines.
--
jc
You can ask Canon yourself. They'll tell you that they had absolutely no part in the Kodak SLR/c. That includes any authorization, assistance, parts, or royalty considerations. It's a completely independent Kodak product. They'll also tell you that Canon can't and won't take responsibility in any problems your Canon lenses may have with the SLR/c because they have no involvement with it. And they'll tell you the same about non-Canon lenses, too, because they have no involvement in, and offer no approval of, those products either. It's all a product of independent reverse engineering outside of Canon, and with no payment to Canon.
Jimmy Chen
Senior Member
So what this means is Kodak did a cleanroom development, like AMD with their "Intel" chip. This then also means even tho Sigma built the body, Sigma have no rights to use the Kodak Canon system unless Kodak agrees.
Then the question becomes, will the profit be there for Sigma to recover the R&D cost for a Canon mount body?
It most likely do not as I bring back what I have said earlier, Sigma currently expressed no interest in offering any Canon mount bodies, as they intend to stay with the SA mount for their furture SD bodies.
--
jc
Joseph S Wisniewski
Forum Pro
Sorry to start the conversation this way, but I find this a little frustrating.
First, you have still not presented a business case in which Sigma's production of a Canon mount camera will actually hurt Sigam's business.
Second, you continue to ignore the fact that Sigma is currently producing Canon mount bodies, a fact which pretty much negates all your intellectual property arguments.
Now, on to the rest of the discussion.
Second, focusing protocols are rather generic and are not what differentiates mounts. I ennumerated several characteristics that do make a difference.
--
Salvage troll posts! When you see a thread started by a troll, post something useful to it. It will drive the trolls up the wall.
Ciao!
Joe
http://www.swissarmyfork.com
First, you have still not presented a business case in which Sigma's production of a Canon mount camera will actually hurt Sigam's business.
Second, you continue to ignore the fact that Sigma is currently producing Canon mount bodies, a fact which pretty much negates all your intellectual property arguments.
Now, on to the rest of the discussion.
AMD had a cross licensing agreement with Intel, which got them up to the 486. For the Pentium procecssors, AMD used an entirely different interface. You cannot plug an AMD Athlon into a motherboard built for an Intel Pentium, nor can you use a Pentium on a board built for Athlon. The processors even differ in instruction sets (SMM vs 3Dnow, and AMD's "Hammer" 64 bit extensions).
Intel accounts for a very small percentage of motherboards.
I'm sure Dell, IBM, and HP are going to be very surprised to hear this, as they mistakenly believe they're occupying the first three places. Although Dell at least is using Intel chipsets along with Intel processors.
Would you like some reading suggestions?
Actually, they did. Intel moterboards were much more common in servers back in the good old days than they are today.
You're missing several things. First, as I said, lens design is complicated by the fact that there are an awful lot of lens design patents out there.
Second, focusing protocols are rather generic and are not what differentiates mounts. I ennumerated several characteristics that do make a difference.
--
Salvage troll posts! When you see a thread started by a troll, post something useful to it. It will drive the trolls up the wall.
Ciao!
Joe
http://www.swissarmyfork.com
Joseph S Wisniewski
Forum Pro
He said Foveon worked with the Canon mount. This does not imply that Canon worked with Foveon.
Lots of people work with the Canon mount. It's popular, reverse engineered protocol documentation floats from lab to lab. Folks like Foveon use it for prototype cameras. College students use it for their optics, astronomy, and robotics projects.
Michigan's ROTSE uses an array of four Canon 180mm f1.8 telephotos, with reverse engineered protocol to sync the EOS lenses.
Canon didn't work with any of thses people.
Not unless Foveon could give Canon something with low light and long exposure capabilties equivelant to existing Canon sensors, something larger than 1.7x, and something cost competative with Canon's internal production (which is running over 125,000 chips per month).
--
Salvage troll posts! When you see a thread started by a troll, post something useful to it. It will drive the trolls up the wall.
Ciao!
Joe
http://www.swissarmyfork.com
Joseph S Wisniewski
Forum Pro
Well, they just got the Canon mount body running for Kodak a couple of months ago?
For their own product, Sigma appears to have bery limited development resources. They couldn't even get SD9 running off a single battery like Nikon and Canon could. Even the simple things take time when your budget is small...
I love a good conspiracy theory. This one can be disolved rather easily with a little "daylight".
Sony happily provides Canon, Nikon, Olympus, and Kodak with sensors for all their point and shoot cameras, despite Canon's use of their own sensors in most of their DSLRs (1D used a Matsushita), Nikon's attempt to emulate Canon and use a in-house chip for D2h, Oly's use of Kodak CCDs in their DSLRs, and Kodak's use of FillFactory sensors in their DSLR.
Nikon and Canon have shown nothing but tolerance to the specialty cameras nibbling at the edges of the market. A Canon mount Sigma would be ignored, just like a Canon mount Kodak, a Nikon mount Fuji, or a Nikon mount Kodak.
Got any references to that, because it sounds totally off the wall.
You're right, it is entertaining.
--
Salvage troll posts! When you see a thread started by a troll, post something useful to it. It will drive the trolls up the wall.
Ciao!
Joe
http://www.swissarmyfork.com
Yi Ding
Senior Member
Joe,
Do you know if Canon charges a license fee for the 3rd-party lens makers to make Canon mount lenses?
When the Kodak SLR/c was annouced, I read a post suspecting that Kodak's license of making Canon mount bodies back from the early DCS days hadn't expired. If a license from Canon is required to make a Canon mount body, then it might be Canon who is blocking the production of a Canon mount Foveon camera by Sigma.
Yi
http://www.pbase.com/yiding/galleries
http://www.pbase.com/sigmasd9/yi_ding
Do you know if Canon charges a license fee for the 3rd-party lens makers to make Canon mount lenses?
When the Kodak SLR/c was annouced, I read a post suspecting that Kodak's license of making Canon mount bodies back from the early DCS days hadn't expired. If a license from Canon is required to make a Canon mount body, then it might be Canon who is blocking the production of a Canon mount Foveon camera by Sigma.
Yi
--Actually, he's not. You're taking one issue, and trying to make it
into two...
This case isn't similar, at all. First, in the case of PC's using
Intel processors, the interface is so complex, and Intel is so
agressive, that no one can build either chip or PC without
licensing IP from Intel. Since Intel doesn't particularly want to
be in the motherboard business, they license the interface to
several chipset vendors.
In the case of the camera and lens, the patents (you should look at
them) cover the interface between camera and lens. There's no
difference, from a legal standpoint, in building a camera or a
lens. In fact, building lenses is harder, because each lens design
has its own individual patent, so you have to make sure your lens
design is unique.
The issue at hand is legal precident. Canon has set the precident
by allowing Sigma, Tokina, Tamron, and Cosina to manufacturer
lenses relatively unhindered for almost 2 decades. Granted, those
companies have to reverse engineer the communications protocol (the
patents are incomplete in their specification of message sets. This
is often done deliberatly). But they can't turn around now and say
"stop" because suddenly people are manufacturing cameras. That's
why Sigma can get away with building Canon mount cameras under
contract for Kodak. Because the Genii has been out of the bottle
for almost 20 years, and tehy're noit getting it back in.
Intel has set their own precident: come to us for permission, build
the interface, pay us royalties, and pay us to certify it so that
your interface doesn't sully Intel's rep. Someone tries to build a
processor with the latest Intel interface, Intel is either going to
tell them to buzz off, or demand royalty payments so high that
Intel will make as much money as if they had sold the chips
themselves.
--
Salvage troll posts! When you see a thread started by a troll, post
something useful to it. It will drive the trolls up the wall.
Ciao!
Joe
http://www.swissarmyfork.com
http://www.pbase.com/yiding/galleries
http://www.pbase.com/sigmasd9/yi_ding
Bluhighway
Member
- Messages
- 28
- Reaction score
- 0
D
David Martin
Guest
If my memory serves me right, that is, it was postede here that the license to the Canon lens mount has expired - not surprising as it is what? 20 years old now?
DaveMart
Please see profile for equipment
Regards,
DaveMart
Please see profile for equipment
D
David Martin
Guest
I agree, it sounds tricky - my info that it was being worked on came froma post by Paul Pope, who has good access, that Canon were hoping to intor such a camera in the 1DIII, for release in 2006 - but that post was around a year ago now, and I have heard nothing more, so it is more than possible that the research has not been successful, or that Canon has simply changed it's mind.
As an aside though, the new 1DsII is putting great strain upon the lenses, and presumably the fewer pixels in a full-colour at every photosite sensor would be more forgiving on lenses, so it seems to me that that is a possible incentive for Canon to try do make this technology work.
Regards,
DaveMart
Please see profile for equipment
As an aside though, the new 1DsII is putting great strain upon the lenses, and presumably the fewer pixels in a full-colour at every photosite sensor would be more forgiving on lenses, so it seems to me that that is a possible incentive for Canon to try do make this technology work.
--
Regards,
DaveMart
Please see profile for equipment
Cryptnotic
Senior Member
Sigma has a very simple strategy:
1. Sell as much stuff as possible.
2. Make as much money as possible.
A lens mount is a simple thing. It's a piece of machined metal with a couple of cut-outs and a locking pin. The communications protocol is probably relatively simple as well. There are some number of pins (I'm too lazy to look right now). There's ground, digital power, motor power (probably separate). Then there are probably separate transmit and receive pins, or maybe they use a clocked serial communications scheme like I2C. In any case, it takes an engineer a couple of hours to decipher what the signals are doing. Then he either builds a monitoring tool attached to a PC or perhaps a stand-alone logic analyzer. He assumes that the Canon engineers designed their system logically. He looks at what the aperture says it is set to and compares it with what gets communicated on the wires. He looks at what gets communicated when he turns on the camera and when he presses the shutter button. He compiles all of this information. He then makes an emulator, a device that hooks up to the camera and pretends that it's a lens. Then, he gets that emulator to work with the Sigma lens and do what it is supposed to. Then, of course, there's the making sure that it works with every EOS camera... That's the hard part...
Anyway, this is something that should only take a single engineer a month or two to complete. It's NOT rocket science. Also, it's not illegal to sell without paying Canon royalties unless there are patents (unlikely). Just don't call it "EF" or "EOS", because those are trademarks of Canon. You're usually clear to say it's "compatible with Canon cameras" though.
Yes, I am an engineer at a company that does both reverse-engineering and contract engineering.
1. Sell as much stuff as possible.
2. Make as much money as possible.
A lens mount is a simple thing. It's a piece of machined metal with a couple of cut-outs and a locking pin. The communications protocol is probably relatively simple as well. There are some number of pins (I'm too lazy to look right now). There's ground, digital power, motor power (probably separate). Then there are probably separate transmit and receive pins, or maybe they use a clocked serial communications scheme like I2C. In any case, it takes an engineer a couple of hours to decipher what the signals are doing. Then he either builds a monitoring tool attached to a PC or perhaps a stand-alone logic analyzer. He assumes that the Canon engineers designed their system logically. He looks at what the aperture says it is set to and compares it with what gets communicated on the wires. He looks at what gets communicated when he turns on the camera and when he presses the shutter button. He compiles all of this information. He then makes an emulator, a device that hooks up to the camera and pretends that it's a lens. Then, he gets that emulator to work with the Sigma lens and do what it is supposed to. Then, of course, there's the making sure that it works with every EOS camera... That's the hard part...
Anyway, this is something that should only take a single engineer a month or two to complete. It's NOT rocket science. Also, it's not illegal to sell without paying Canon royalties unless there are patents (unlikely). Just don't call it "EF" or "EOS", because those are trademarks of Canon. You're usually clear to say it's "compatible with Canon cameras" though.
This might be the case based on the private contract between Kodak and Sigma. Perhaps Kodak paid Sigma to do the SLR/c body with the stipulation that they could only sell it to Kodak (very common practice with contract engineering). Maybe Kodak said, "Hey, we want you to do a Canon-lens compatible body for our new camera. We don't need it to be exclusive. Our sensors are exclusive enough. So you can do what you want with it on other projects, but we're not paying you extra." Either case is possible.
Yes, I am an engineer at a company that does both reverse-engineering and contract engineering.
Dominic Groß
Veteran Member
But this technology is pretty much worth nothing for a Camera like a 1Ds, temporal aliasing will imho badly spoil it for daily use, except for really still subjects, just like multishoot Cameras. Some other Companies offered technology like this (variable filter in each pixel) before (for example silicon vision) and went out of buisness pretty quick.
--
http://www.pbase.com/dgross (work in progress)
http://www.pbase.com/sigmasd9/dominic_gross_sd10
--
http://www.pbase.com/dgross (work in progress)
http://www.pbase.com/sigmasd9/dominic_gross_sd10
Well, that's certainly up to them. But that would make about as much sense as them only offering SA-mount lenses (and not Canon, Nikon, Pentax, and Minolta-mount lenses). And it would make even less sense considering that they already have a Canon-mount body on their hands (with what they've been able to provide to Kodak). I think it's only a matter of time before they utilize that body as a Sigma-marketed EF-mount body. After all, how many SD bodies have they been able to sell by limiting it to their dead-end, niche-market SA mount? And would they even be in business if they likewise limited their lenses to only SA mounts?
Would you buy such a camera?
How much would you be willing to pay for it if it were available now?
Would you buy it as your main camera or as a secondary one, to complement the Canon body you already have?
I, for one, would buy one, as a secondary body, if it were no more than $1600.
--
Smile! Tomorrow will be worse!
Jimmy Chen
Senior Member
And Kodak is the only vendor that has a production camera using that mount. This is espeically interesting considering how free the Canon mount engineering is as you indicated.
--
jc
Jimmy Chen
Senior Member
Sure I have. You just do not want to accept what I have said can be a possibility. You keep saying Canon mount lenses equate to Canon mount bodies. Then why Kodak is the only 3rd party that has a Canon mount body, while there are plenty of others companies that offers Canon mount lenses?
And you keep ignoring those bodies are for Kodak. No different then Fill Factory is making FF chips for Kodak. Do you see Fill Factory selling FF chips to anyone else? What indicated that Sigma developed the Canon body w/o Kodak's help, and not under a Kodak contract?
And for the P chip, AMD did a cleanroom development and not some sort of reverse engineering. Only cleanroom development under the court of law you arent liable to pay any royality fee.
I am sorry, you are incorrect here. Here are some hints ... both MS and Intel where tiny companies subcontacted by IBM. Intel did not have the man power nor know how to offer any MB back then. Dell didnt even exist and HP wasnt even in the PC business when all this took place. You need to research a lot further back in history on this one. Your scope on the computer history is too short.
--
jc
Because the Genii has been out of the bottle
Patents normally EXPIRE at the end of 20 years.
The EOS system is around 17 years old. The patents may also predate the EOS system.
People like me buy Canon because the EOS lens system is very comprehensive. Canon's advantage is not merely aging IP.
Keyboard shortcuts
- f
- Forum