Picture of new M mount ???

[MP Guy]

This is believed to be a picture of the new Leica M mount with contacts for the M8 or Digital M. This cannot be confirmed yet. But, rumor has it Leica will make an annoucement this week.

 

[RayPA]

Would this be a shot of the D Vario-Elmarit 14-50mm? It looks like those "contacts" are to be read optically (?).


.

 

[jlw]

Interesting -- looks like a simple and clever system. One bit of good news about this is that it should be pretty easy for other M-mount lens manufacturers to reproduce -- all they need to do is start forming depressions into their mount flanges, and then paint them white or black in the correct pattern for the specific lens. It should even be possible to make screw-to-bayonet adapters that you could encode yourself with paint and a peel-off stencil.

This does seem to rule out earlier speculations (including mine) that Leica might be moving to some system that would transfer aperture, distance, and other information to future bodies; a system of painted dots obviously would be capable of encoding only static, unchanging information, not quantities that varied as you adjust the lens. This also means that "vario" lenses would not be able to shift their framelines automatically, at least not without some tricky internal mechanism to change the dots.

A thought: If I'm doing my binary math correctly, a six-dot coding system would allow for 64 different values; that seems like a lot of possibilities compared to the relatively small number of lens focal lengths that Leica makes, or might make, that would be covered by the viewfinders of future M cameras. (Even if you postulate an optically unlikely viewfinder that somehow could provide framelines for, say, 12mm, 15mm, 21mm, 24mm, 28mm, 35mm, 40mm, 50mm, 75mm, 90mm and 135mm lenses, plus one uncoded position for "other," that's only 12 possibilities.)

I still wonder if they might be using only three dots to encode focal length (that would provide eight possible values for, say... 15, 21, 24, 28, 35, 50, 75 and 90) and the other three to encode the lens' maximum aperture? The remaining three dots also would provide eight possible values for this, which would be enough to cover the range from f/1 to f/3.5 in half-stop increments.

Why would they want to do this? As I've speculated before, currently there's no way for an M-mount lens to signal its current working aperture to a camera body, and adding that in the conventional way would involve more moving parts (an aperture-ring encoder) and electrical contacts.

On the other hand, if the camera body "knows" the lens' maximum aperture, it can calculate its approximate working aperture by comparing a TTL light reading to another reading from a non-TTL sensor, such as one that might be used for auto white balance. This would allow a digital M camera to record all the expected data -- shutter speed, aperture, EI and focal length -- in its EXIF headers, without requiring the addition of any moving parts to the lens.

Yeah, I know, this is speculation -- but it's a kind of creative speculation that entertains the brain...

 

[RayPA]

...

On the other hand, if the camera body "knows" the lens' maximum aperture, it can calculate its approximate working aperture by comparing a TTL light reading to another reading from a non-TTL sensor, such as one that might be used for auto white balance. This would allow a digital M camera to record all the expected data -- shutter speed, aperture, EI and focal length -- in its EXIF headers, without requiring the addition of any moving parts to the lens.

Yeah, I know, this is speculation -- but it's a kind of creative speculation that entertains the brain...

could be coding to the camera (M8) ROM information that'll populate EXIF and supply the proper framelines (?).


.

 

[Mark Norton]

Those patches are outside the lens bayonet and so open to the elements which is why they will be read optically, if they were electrical, damp could give an incorrect reading.

Looks to me like it's a simple 6 bit coding - 63 different values plue one for "None"; how they will use the coded values is anyone's guess, but I expect there will be values for the focal lengths and possibly specific values for the actual lens being used.

But, it's all speculation....

 

[Ben Z]

Interesting -- This also means that "vario" lenses would not be able to shift their framelines automatically, at least not without some tricky internal mechanism to change the dots..

Yes, so the fact both versions of the Tri-Elmar can be coded would also rule out the previous speculation that the coding will be necessary for bringing up the right framelines, so therefore uncoded lenses will bring up the appropriate ones. That would follow from Leica's statement on the other thread that someone translated from the French "As for the system Leica R, the contacts will not be essential, but they will make it possible to optimize the adjustments with the M numerical."

I worked happily for more than a year with a whole set of non-ROM lenses on an R8, as many people continue to do with the DMR, and if uncoded lenses will work no worse on the M8 then I guess a lot of people will be satisfied. Requiring retrofitting or manually keying the framelines would be a horrible misstep on Leica's part and I will be flabbergasted if that's what they've done. I mean, they went to the trouble of setting up the spatial relationships so a Visoflex and also googled lenses will work on the M8, so it just doesn't stand to reason they would screw up the frameline actuation and thereby make using older lenses a PITA.

 

[Socke]

On the other hand, if the camera body "knows" the lens' maximum aperture, it can calculate its approximate working aperture by comparing a TTL light reading to another reading from a non-TTL sensor, such as one that might be used for auto white balance.

Why so complicated? The Contax G2 Databack records an "Estimated F-Number" by applying sunny 16. Sensitivty and shutterspeed are known so it's easy to deduce the corresponding f-stop from the meter reading. Only unknown variable would be the use of filters.

 

[MP Guy]

Good point Socke. If indeed is to control frame lines, I hope there is a manual option.

 

[Bob Ross]

Good point Socke. If indeed is to control frame lines, I hope there is a manual option.

I thought that this might be a good place to make my first post to the forum 😀
I suspect that the "manual" way to pick the correct frame lines will be in the menus, as is done with the Nikon D200 when using older AI manual lenses. The 1.33 FOV must have been a design challenge, since it not only changes the field of view, but also the difference between focal lengths. For example the difference between the 50mm & 75mm is 25mm, but with 1.33 FOV it is the difference between 67mm and 100mm is 33mm. The old frameline pairs don't nicely match up with the new cropped frameline pairs. Another issue might be the digital lens change protocol: turn camera off, change lense, turn camera on and set frame lines. This will change some old habits.... 🙂

 

[MP Guy]

Here is some more information from a French website which goes into some details.

This is the link:

http://www.summilux.net/m_system/objectifs_6_bits.html

and it has a list of lenses which can be adapted.

I have run it through a translation site and came up with the following english translation from the above link:


New : From the 1er July 2006, Leica will deliver objectives M with contacts . As for the system Leica R, the contacts will not be essential, but they will make it possible to optimize the adjustments with the M numerical.

Contacts on an optics M

From the 1er July 2006, Leica will deliver objectives M equipped with contacts 6 bits.

* The code 6 bits extends the functions of the numerical LEICA M.
* Compatibility is completely ensured with the system present, passed, future.
* The public price of all the objectives will be slightly increased to the 1er July 2006.
* The current objectives, delivered before the 1er July 2006, could be updated for a tariff appreciably equal to the increase in the 1er July 2006.
* The references remain unchanged, packing of the coded objectives 6 bits will be recognizable with the labels announcing the 6 bits.
* The former objectives can also be updated (see the tables below).

Which is the object of optics M with identifying code 6 bits ?
The update of the bayonet of optics Leica M is used to ensure an optimal result of image associated the LEICA M numerical. The system of the optical sensor makes it possible the apparatus to read the code on the bayonet of the objective and to identify optics gone up on the aircraft. Of course, it is possible to use uncoded optics and to obtain good results with the digital LEICA M, and, vice versa, all coded optics can be used on silver apparatuses LEICA M without any restriction. The digital LEICA M adds the data transmitted by the objective to the metadata and uses this information to treat the data of the image optimalement. With an aim of preserving compatibility with the former and future material, the bayonet of the LEICA M remains unchanged.

How does the code function ?
With an aim of marking the type of objective, the ring of the bayonet is equipped with 6 contacts black and white. On the digital LEICA M, coding is recognized with the assistance of 6 sensors integrated into the bayonet of the apparatus.

 

[Mark Norton]

I can't see they will be allocating 3 bits to focal length and 3 to maximum aperture because they have only 6 bits to play with and there would be lots of combinations which would never be used - 21mm f1, for example. Seems more likely they will have a code for 50mm f1.4, one for 50mm f2, one for 50mm f4 and so on. Whether they will distinguish between different 50mm Summiluxes is another matter.

What we don't know yet is what use will be made of this information and that will depend on how far away they've moved from the M7 viewfinder. If it's pretty much unchanged, the lens coding will be for the EXIF header and maybe the in-camera processing. However, if they've done something radical with the viewfinder, like variable magnification or electronically generated frame lines, the coding could be used to display a single frame line instead of a pair to match the lens mounted.

I think the bayonet lug would be retained for uncoded lenses to display a pair of frame lines and if the coding said "Tri-Elmar", the lug position could be used to display a single 28, 35, 50mm frame.

Inside the firmware, there will be a mapping between code value and lens characteristics, so if they came out with a new focal length, the firmware would need to be updated to teach the camera about the new lens.

If you then add into this mix the question of aux finders (and the need for new ones to suit the crop factor) and the finder magnifier which is the subject of its very own patent - take a look on the US Patent Office Web-Site - there's an awful lot of stuff we do not know about this camera yet.

 

[MP/CLE]

Geez, guys... I just want to be able to use all my m lenses on the new digital camera, and maybe pick up a couple that will be essential for the m8. Is this unreasonable? C'mon, father Leitz... amaze me!

 

[Mark Norton]

I'm sure you'll be able to do that, you may just have to select the lens you are using from a menu if you don't go down the coding route.

 

[Gid]

The codes may be related to in camera image processing to counter vignetting etc. Olympus does similar on the 4/3 E series cameras.

 

[Ken Ford]

The codes may be related to in camera image processing to counter vignetting etc. Olympus does similar on the 4/3 E series cameras.

My feeling, too - just like Nikon Capture does with specific lenses. I also suspect the coding will be a simple numerical sequence - lens A is 01, lens B is 02 etc.

 

[Steve L]

The coding also could be used for in-camera vignetting compensation...knowing which lens is mounted would do that. And it will be an issue with the M8, for sure.

Each lens would have a profile so that the firmware could adjust.

That might explain the line in the announcement about optimizing the image....

 

[Ben Z]

To compensate for vignetting the camera would need to know not just the lens in use, but also what aperture it 's set for, since vignetting changes as you stop down.

 

[Kensey]

I was recently in Richard Caplan, a Leica dealer in London, and was told the following: M-lenses will need to be "chipped" at a cost of £75 each. Non-chipped lenses will work on the M-digital in much the same way as 3cam R lenses work on R8/R9 but to get full functionality ROM lenses are needed.

 

[anselwannab]

I think Mark Norton hit on something there with the patent issue. I guess the mount will be the same, but they dots might tell the camera what lens is attached and how to compensate for it. If that is patented, 3rd party vendors won't be able to match it?

I thougth 3rd party engineers work around and re-engineer lenses to work on Nikon and EOS cameras? Could LEica make it so that no one can make lenses with the dots?

As to vignetting and such, couldn't software (DXO Analyzer ??) do the same fixes to pics, just in post processing?

If no one else can "dot" their lenses, I can forsee alot of people "dotting" their own lenses!

Mark

 

[Ben Z]

I was recently in Richard Caplan, a Leica dealer in London, and was told the following: M-lenses will need to be "chipped" at a cost of £75 each. Non-chipped lenses will work on the M-digital in much the same way as 3cam R lenses work on R8/R9 but to get full functionality ROM lenses are needed.

If non-chipped M lenses will work as well on the M8 as non-ROM lenses work on an R8 we can all breathe a sigh of relief and save some money.