Proof of concept regarding self coding of lenses

[stunsworth]

Advertisement (gone after registration)

With the Leica solution the six elements that make up the binary code are milled into the lens mount and then painted are they not? Simply painting codes on the lens mount will surely be a temporary solution as the paint/ink will wear off.

[mike prevette]

I'm sorry but the thought that this is akin to music piracy is inane. This is just a proof of concept and nothing more. Of course Leica's solution is the only one you should pay for, and is certainly more durable than a couple of marks from a sharpie. If a company feels that user modification to their products is patent infringement then they are going to have a long and painful PR battle ahead of them.

 

I love Leica, and want them to succeed, but this system is just to damn dumb for me to spend $1000 getting my lenses coded by them, not to mention the weeks I'd be without my lenses. I have very little intention of even using their coding system until they give me some variable control over it.

 

I'm a photographer not a 'system' guy, all I want to do is shoot with the equipment I'm comfortable with, and that I know delivers.

 

_mike

[Joseph S. Wisniewski]

Coding for Leica lenses (without sending them in) or coding non-Leica lenses seems like a great idea, BUT Leica definitely has a copyright/patent issue here that should protect their technology.

Steve, can you reference a patent number? I can't see any way they could get a solid patent, one that would hold up against a serious challenge, on this coding. It seems to be to be an obvious extension of existing coding schemes, like DX film speed coding (a pattern of paint patches on the film cassette) or the Nikon AI-s notch that indicated whether a lens was above or below 135mm to shift between a "normal" and "telephoto" program.

It's easiest to see in a case where, for example, Zeiss started to code their lenses, in competition with Leica lenses. I believe M mounts from other manufacturers were only possible once Leica's patent on the mounts expired.

Actually, there's tremendous precedent supporting reverse engineering for interoperabaility. Sigma, Tamron, Vivitar, etc. brought out Canon EOS mount lenses within two years of Canon's launch of that highly patent protected mount. The patents protected Canon against other camera makers making similar mounts for their own systems, not against other lens manufacturers making lenses to operate with the Canon system.

 

Even attempts to invoke newer, more aggressive intellectual property laws have failed, such as HP's recent attempt to use the draconian DMCA to protect the little chips on their inkjet cartridges.

So self-code at your own risk. It seems akin to downloading music without paying for it. It's done, but it's not legal.

In this case, I'm betting the courts would say it is legal.

 

And I'm also betting that Leica won't push the issue, because they don't need the negative publicity of a move that could be perceived as "closing" their system. Especially now, after some bad publicity from the M8 issues, and Reichgate. Look at Adobe: they took a big PR double hit with the closing of the PhotoShop SDK and the arrest of Dmitry Sklyarov, so they've been running around trying to get into the spotlight with their "open standards" advocacy of DNG and their condemnation of Nikon's encrypting of raw file parameters (which pales before Adobe's encryption of, well, just about everything).

[sean_reid]

Now this would make me buy an M8, if I can bypass the closed leica system and use my non-Leica lenses! Great idea!

 

Hi Lloyd,

 

I hope that Leica sees this comment because you are one of many who will buy the M8 if it is not part of a closed system. Whether they realize it yet or not, this self-coding is going to help, not hurt, their bottom line. People who want, and can afford, to buy coded lenses or have them coded will still do that but this opens the camera up to a wider group of buyers.

 

Cheers,

 

Sean

[sean_reid]

Coding for Leica lenses (without sending them in) or coding non-Leica lenses seems like a great idea, BUT Leica definitely has a copyright/patent issue here that should protect their technology. It's easiest to see in a case where, for example, Zeiss started to code their lenses, in competition with Leica lenses. I believe M mounts from other manufacturers were only possible once Leica's patent on the mounts expired.

 

So self-code at your own risk. It seems akin to downloading music without paying for it. It's done, but it's not legal.

 

Steve

 

 

Hi Steve,

 

I must disagree with you and I'm a strong supporter of intellectual property. No one here is proposing using Leica's system for another camera. Rather, we're finding ways to make the M8 (which we paid Leica to buy) work well with our existing lenses. Coding was meant to be an optional choice (see Leica's official comments about this). With the filters and cyan drift problem, coding has now become necessary for correct color rendition with lenses of 35 mm or wider. Until and unless Leica provides manual coding options via the menu, this is an excellent alternative that will serve to increase, not decrease, M8 sales. See Lloyd's post for example, he's one of many I've heard from who will not buy the M8 if it's part of a closed system. I can tell you that I have heard from *a lot* of prospective M8 owners about this issue.

 

Cheers,

 

Sean

[rvaubel]

Hi Lloyd,

 

I hope that Leica sees this comment because you are one of many who will buy the M8 if it is not part of a closed system. Whether they realize it yet or not, this self-coding is going to help, not hurt, their bottom line. People who want, and can afford, to buy coded lenses or have them coded will still do that but this opens the camera up to a wider group of buyers.

 

Cheers,

 

Sean

 

Sean

 

I'd be curious if the 16-18-21 TE lens coding would work well with the 15mm Heliar. It might make a pretty good match especially with regards the radial cyan issue.

You have both of those lenses, don't you? And a sharpies?

 

Rex

[rfleica]

Advertisement (gone after registration)

Whether they realize it yet or not, this self-coding is going to help, not hurt, their bottom line. People who want, and can afford, to buy coded lenses or have them coded will still do that but this opens the camera up to a wider group of buyers.

Sean

 

Agreed. I will pay a premium when i have to, and will save money where i can. It seems pretty obvious that with the m mount there are extremely interesting, modestly priced lenses available from Zeiss...and some bargains from CV. I will be much more attracted to the M system if the m8 is "open". In fact, it is a factor in considering whther i stay with the m system and go for the m8...or one day drop the lot.

[sean_reid]

I love Leica, and want them to succeed, but this system is just to damn dumb for me to spend $1000 getting my lenses coded by them...

_mike

 

Bite your tongue..<G> Thank goodness it is a simple coding system. Let's not suggest anything more elaborate.

 

Cheers,

 

Sean

[sean_reid]

Sean

 

I'd be curious if the 16-18-21 TE lens coding would work well with the 15mm Heliar. It might make a pretty good match especially with regards the radial cyan issue.

You have both of those lenses, don't you? And a sharpies?

 

Rex

 

Hi Rex,

 

I don't have the TE here and it's a very different design from the CV. I'm not sure how well the TE's correction at 16 will work for the 15 CV. Time will tell.

 

Cheers,

 

Sean

[Stan Raucher]

Coding for Leica lenses (without sending them in) or coding non-Leica lenses seems like a great idea, BUT Leica definitely has a copyright/patent issue here that should protect their technology. Steve

 

Actually, it's Leica that is violating copyright/patent issues by using a binary code.

 

Microsoft patented/copyrighted the numbers "0" and "1" many years ago.

 

Stan

[marknorton]

I searched the US PTO a few weeks ago and found nothing. Even if Leica has obtained a patent on it, I don't think they would go after individuals marking their lenses. They might go after another manufacturer but the real issue is that the code is not being used as an option to deliver top image quality, it's being used as a necessity to deliver (hopefully) as expected image quality.

 

Easy to say with 20/20, but most people, if tasked with deciding how to identify the lens to the camera would look at their boxes of Cheerios and say, hey, this stripey code thing tells the checker what I'm buying, couldn't we do something similar. The prior art is well established so I'd be surprised if it's even patentable.

 

What we can be certain of is that Zeiss, if they are developing a digital Zeiss-Ikon, are watching this whole saga and thinking: "perhaps we need to revisit our assumptions".

[monza]

Heh. IBM would probably argue that they have said copyright, not Microsoft.

 

This is a very simple 'reverse engineering', in effect. There are no legal consequences.

[chrism]

I wonder if one of those cheap little power engraving tools - used to mark your name on items in case of theft - could make some pits in the lens mount so the Sharpie ink won't wear off. I believe my mother has one, so I'll have to borrow it and find out if it leaves any burrs that stick up, which I assume would be a Bad Thing. If it works it would be akin to tattooing the lens!

 

Chris

[stevme]

Hi Steve,

 

I must disagree with you and I'm a strong supporter of intellectual property. No one here is proposing using Leica's system for another camera.

 

Sean --

 

This is not a forum on legal issues, and it is for Leica to decide what to do about the apparent ease with which its system for correction of lenses used on its sensor can be copied by others. Obviously, a prohibition on private pen marking would be practically unenforceable. If other manufacturers trot out coded lenses, or began to counsel purchasers on how to self-adapt their lenses for use on the M8 in order to promote sale of their non-conforming lenses, it is another matter. Leica would be in court within days seeking an injunction, or should be. If Leica did not patent its system for vignetting and cyan shift correction, an integral system essential to the Leica standard of image quality, well, maybe they should hurry on down to the patent office, right quick. Assuming a valid patent, Leica should be willing to license other manufacturers to use their coding system, for a fee, in order to increase usage of the M8.

 

I, for one, would be reluctant to champion this apparent appropriation of their technology, especially in commercial contexts, howsoever desirable it may be from a user context. Of course, what I do with magic marker pens in the privacy of my own home for non-commercial purposes, is my own business. I just might be clever enough to search the internet for new uses of magic markers, especially for LTM Leica lenses for which coding will not be made available. Even so, I plan on sending my 35 mm 1.4 up to New Jersey for coding as soon as things die down a bit. I don't want to risk gumming up the coding sensor on the M8.

 

The point stands, however. Leica should not be required to give away its technology nor let others take it from them without compensation. As a matter of common sense, it might be well to let users mark up their own lenses, but that is Leica's decision to make -- one they will be making as soon as they read this thread.

 

I know revert from legal to photographic mode . . . .

 

Steve

[sean_reid]

O-Tay,

 

I'm going to add a whole section on this to the review of the 28 mm lenses but what I'd like to refer people to is a page with pictures of the various coded lenses (obviously looking down at the mount). I think those are even more useful (and possibly more fool-proof) than the number lists. Does this exist already or does it need to be made? If the latter, who is willing about to contribute pictures (let's say 300 x 300 pixels each)?

 

Cheers,

 

Sean

[rvaubel]

Hi Rex,

 

I don't have the TE here and it's a very different design from the CV. I'm not sure how well the TE's correction at 16 will work for the 15 CV. Time will tell.

 

Cheers,

 

Sean

 

Sean

Actually, you don't need the TE you just need the code and a sharpie. Thats the fun of the self coding... you can try it and no harm done. Probably the Leica lens with the shortest exit pupil would work the best, but trial and error would be my approach. If I had the camera

 

Rex

waiting for the next wave

[marknorton]

One of the questions Leica will be asking internally is: "Do we do this user lens selection thing?". Fortunately, Mike's coding method has provided them with an ideal solution: they don't need to do it, and providing they take a relaxed view of it, they will not be accused of producing a closed system. I think they'll be happy at this progress.

[wparsonsgisnet]

There has been discussion of the 6-bit code only generating 64 values, but, of course, the body COULD know from the cam what the focal length of the lens is. Therefore, X00X00 could have different meanings for different lens lengths.

 

It appears from the list of codes that has been gathered, that Leica has not yet used all 64 codes, but they certainly can expand in the future if they wish.

 

Let's see, software version 7.6.2.4, it might be.

 

I am sticking by my guns -- I want the cyan and vignetting correction in software. I think it would be terrific for the M8 to record the focal length of the lens and put it in the exif file, but even the same lens can have a different cyan or vignetting pattern at different apertures.

 

I agree with Sean that the differences are not great, and probably not easily detectable for a single lens at different apertures, but the differences are apparent across brands in his review. So, an automated post-process workflow to correct for these aberrations is a more general solution.

 

Software is also a lot cheaper than (1) paying for lenses to be coded, and (2) both risking our lenses to the mail and waiting while they're at the spa.

 

Regards,

[rvaubel]

Someone correct me if I'm wrong, but I asume the main reason for lens coding (other than EXIF info) is to correct for the radial cyan issue and cos vignetting. As I understand it both of these correction are governed by the same Cos4 formula. So that would mean that the amount of cyan correction and the amount of vignetting correction would be roughly the same. My final assumption is that the exit pupil of the lens is the determining factor for the Cos4 formula. So doesn't this all mean that only one variable (exit pupil) would suffice to determine the lens coding?

 

Rex

[Advertisement]

Hallo,

Look here: Leica M

[marknorton]

I've been trying to code for the new T-E, apparently code 16 but I've also tried 47, nothing. I'm wondering whether the current firmware knows about this lens. Anyone else tried it?