Why the coding?

[hammam]

Advertisement (gone after registration)

Sorry if this has been already explained, but there are so many posts about the M8 that I probably missed a lot of them. And since it seems now that the coding will be mandatory if one wants to use the IR filter (I'd like to know exactly why as well), here goes...

 

What exactly does the lens coding do with the M8? I know it's there to send infos to the camera as to the focal length and max aperture of the lens in place. Okay. Good. But what does THAT do? There's no other metering mode than the M8's kind of heavily-center-weighted strip. No matrix metering here. And does the coding send some other kind of information as to the optical specifications or indiosyncrasies of the lens? Does it have an influence on the in-camera processing of the picture? In what way? Then, what about when you shoot DNG?

 

Second, is the coding really necessary if you shoot WITHOUT IR filters, and why?

 

Third, is the coding necessary ONLY with IR filters, to correct some CA in the corners, as has been suggested elsewhere?

 

If somebody can answer these questions, thanks a lot.

[wide.angle]

I was wondering what the coding provides additional since, for e.g., my MP already knows which lens I put on it and adjusts the viewfinder accordingly.

 

I hear the codes are painted on, so it must only be telling the M8 the same thing it should already know.

 

And if they are painted on, instead of sending lenses into the factory for painting, I would hope someone would post a DIY way to code one's own lenses.

[puigvert]

I was wondering what the coding provides additional since, for e.g., my MP already knows which lens I put on it and adjusts the viewfinder accordingly.

 

I hear the codes are painted on, so it must only be telling the M8 the same thing it should already know.

 

And if they are painted on, instead of sending lenses into the factory for painting, I would hope someone would post a DIY way to code one's own lenses.

 

I believe that the analogical M's know the optics inserted by to mechanical action on to levy.

 

For equal Of way the M8.

 

Focal Also I believe that the codification with the painted marks makes reference to the length of the lens.

It is used for the flash with automatic zoom lens-angle. For the vinyeting i do not believe that it serves, since the M8 does not receive information of the lens on the applied opening.

 

In order to of sees able to decipher the table codes there am is simple ace ace the following thing:

each to user who there are to codified optics specifies his focal length, with codification B (black) W (white) of the marks painted in the sense of left to right. I say that is to binary codification B = 1, W = 0.

 

Eleven it is had an example of each focal length, the table will sees completes!

[marknorton]

The code consists of 6 pits milled into the bayonet ring and filled with black or white paint. This gives 63 combinations plus one for "uncoded" (all white). The code is unique to the specific lens formulation so that the 35mm f1.4 ASPH has a different code from the 35mm f2 ASPH.

 

With the knowledge of the actual lens mounted, the firmware designer will know the geometry of how the image is projected onto the sensor surface and it was always said, before the M8 was launched, that this would lead to image optimisation. We thought at the time that this was all to do with vignetting.

 

In addition, because the camera now knows the focal length of the lens mounted, it can put the focal length in the file header and (if it ever appears) adjust the reflector of a dedicated (Metz) flash unit.

 

As it turned out, the micro-lens structure worked so well that the lens coding initially seemed a waste of time; Sean Reid however has shown that there is some correction and it works well.

 

Then came the Magenta Problem, and the hope is that the lens coding will allow the camera firmware to compensate for one of the side-effects of mounting an IR cut filter on a wide-angle lens - cyan colouring into the image corners. How well it works remains to be seen.

 

Discussion here has suggested the lens type should be manually selectable if not coded. We will have to wait and see whether Leica does this (in which case the lens coding becomes nothing more than a convenience feature) or whether Leica will require lenses to be coded

 

On the one hand, I can see users wanting this facility. On the other, Leica will be smarting because of the number of bad M8 images floating around and "corrected" images with an incorrectly selected lens type might just make that worse.

[kschwarz]

If I were designing the lens "coding" system, I would use a combination of the mechanical lens ID that sets the appropriate frame lines on all M cameras, and the M8 menu. I realize that different versions of the same lens might need individual coding, so if for example a 28 mm lens is installed on the camera, then the menu selection choices would automatically be restricted to the various 28 mm lenses. In this way, the user would not need to wade through a hundred or more lens choices whenever the lens coding needs to be changed. The advantage of this system, is that existing lenses would not need to be converted, and yet full lens-specific post-processing could be preformed. Seems logical to me...

[JoelH]

It would seem possible to couple the existing focal-length frameline indication, so that uncoded lenses could at least restrict the menu choices. However, that existing mechanical system has only three cam grinding options, that is why the lens framelines come up in pairs on a given camera. The three options are distributed among the film-M focal lengths of 21, 28, 35, 50, 75, 90 and 135, leaving out the new 24mm of Tri-Elmar, the old 15mm Hologon, and the 40mm of the CL. Now there are even more focal lengths with the latest lens introductions.

 

Compounding the problem further is that some of the aftermarket M-lenses have "missed" the correct grind which are needed to cause M bodies to bring up the appropriate framelines - I've seen discussions about having to machine the mounts a little to get things working properly.

 

So even the "menu restriction" idea might not be very specific or very reliable, and it would require electric contacts. optical recognition or something else to augment the present mechanical-only coupling of that particular mechanism.

 

I also wonder why they didn't add one or two extra 'bits" to the coding while they were at it. Sixty-four choices might not be enough someday. Leica reflex users have to contend with the different generations of cam and electronic coupling, not unlike Nikon AI/non-AI issues etc.

 

If they are going to allow menu-based lens information to be entered at all, opening up the correction/EXIF firmware system to non-Leica brand lenses, then they might as well encourage a system that opens up the coding also. I can see both sides of the "open" vs. "closed" system from Leica's point of view, but having it "open" in only a limited or cumbersome way (menu input required) strikes me as kind of unsatisfying to everyone. Like a committee compromise or a short-sighted management decision which will dog them in the future.