Rumor - Three new Leica M lenses tomorrow?

[jaapv]

Advertisement (gone after registration)

Yes, but that was for a lens/sensor unit.

Maybe the designs have been made more telecentric, assuming the optical formula was indeed changed.

[CheshireCat]

Does 'optimized for digital' correspond to : on film, not as good as the old lens  ?

 

It usually means (as jaapv says) that the lens is "more telecentric" (i.e. less steep angle of incidence) in order to work around digital sensor issues not present in film.

This does not mean that on film it will be worse.

 

But in the case of the three lenses discussed here, the design seems to be basically the same, so they may just be referring to rear element coatings minimizing sensor reflections. A "more telecentric" lens would require a re-design and usually an increased size.

Some elements (asphericals ?) may have been modified to compensate for the sensor I/R filter thickness. This is just a guess, but if so, the performance on film may actually be different.

 

In any case, anything that improves performance on a digital sensor is used by the marketing monkeys to brag about some BS. In this case it is clearly counterproductive, as digital users obviously assume a new lens is "optimized for digital" while film users are scared by the word "digital".

[jaapv]

Well, it might even mean that they have been twiddling the machining tolerances in the focusing mechanism. Which would make sense, as there have been some complaints about alleged "focus shift" -not that I have seen it on my lens- of the Summicron 35 asph. Again, it would not affect film use.

[CheshireCat]

Well, it might even mean that they have been twiddling the machining tolerances in the focusing mechanism. Which would make sense, as there have been some complaints about alleged "focus shift" -not that I have seen it on my lens- of the Summicron 35 asph. Again, it would not affect film use.

 

"Focus shift" usually indicates the focus shifting as the aperture is changed, and it is only a function of lens design, not machining tolerances.

You probably mean "focus precision" as far as the rangefinder is concerned.

[jaapv]

Not me that said it... Focus shift was the claim. I have the same near-certain suspicion that the poster meant focus precision, reason I put it in quotation marks and called it alleged. The error of confusing the two is a common one on this forum.

 Bringing us back to the content of my post, which indeed addressed focus precision.

[wlaidlaw]

Well, it might even mean that they have been twiddling the machining tolerances in the focusing mechanism. Which would make sense, as there have been some complaints about alleged "focus shift" -not that I have seen it on my lens- of the Summicron 35 asph. Again, it would not affect film use.

 

A few years ago, I bought a nearly new 11819 50mm Summicron. As I always do when I get an RF lens, I ran focus and infinity tests on it. It was in perfect focus at infinity but went progressively more out of focus as the distance reduced. I thought I must be doing something wrong so took it to Protech to put on their optical bench. Kelvin confirmed that I was correct. He had never seen anything like this so we jointly phoned up Malcolm Taylor. He said he had seen it once before where the RF cam had been wrongly ground or ground for another lens. He said as the cam was part of the sleeve which holds the elements of the optical cell, it would need a new housing and was close to being an economic write off. I set it back to Collectible Cameras in Phoenix, AZ, with the Protech report, who being nice guys, immediately gave me a refund with no quibbling. That was definitely focus shift. 

 

Wilson

[jaapv]

Advertisement (gone after registration)

Not to be pedantic, -and to be ahead of Cat - but that was shifting focus. Focus shift specifically refers to the shifting of the focal plane as the lens is stopped down.

 

But your experience is indeed what I am referring to. In a small way the tolerances of the focus mechanism will always produce this effect, which may be virtually invisible, especially on film.

As digital is more precise, machining to a more narrow tolerance can be beneficial.

[pop]

..... It was in perfect focus at infinity but went progressively more out of focus as the distance reduced. ....

I had a 90mm Elmarit-M which did that. When I had it coded, I was told that they adjusted it for infinity and that it was off at closer distances. I was also told that Leica could not repair it as they had no more parts.

[scott kirkpatrick]

Let's eliminate one source of confusion, which fooled me for a bit.  Leica has distributed product pictures of the new models which are cropped at the mount surface, leaving out the lens barrel which moves in and out to focus and sticks 4-6 mm into the camera body in the old models and also in the new ones.  I've seen the cropped pictures in online sales catalogs and even an engineering cutaway picture in Leica's docs showed only seven pieces of glass because the eighth and its barrel were cropped.  (Haven't looked today, but that one should not last too long.)  The new MTFs show that an optical design update has improved some things that show up in MTF's (agreement between tangential and saggital lines as well as better values for the high res lines), plus probably some other things that do not show up there.  But the changes in the entrance and exit node positions are probably a mm or less.  Does anyone know why Leica never publishes an exit node position?

 

scott 

[wlaidlaw]

I had a 90mm Elmarit-M which did that. When I had it coded, I was told that they adjusted it for infinity and that it was off at closer distances. I was also told that Leica could not repair it as they had no more parts.

 

My Summicron was more than a bit out at closest focus. It would not have been any use for film as well as digital. Malcolm Taylor's opinion was "just get rid of it"

[edwardkaraa]

Optimized for digital means that the optical formula has been tweaked to neutralize the astigmatism in the image periphery caused by the cover glass. This also means the lens will not perform as well on film as it will have astigmatism in the periphery. That is basically what Zeiss did with the loxia line for the Sony A7. But Sony doesn't have film cameras so no issues there

 

PS. Another example is the summilux 28, which is tweaked for the thicker cover glass of the SL and performs slightly worse on the M. I would expect its film performance to be proportionately degraded.

[Jeff S]

My 35 Summicron ASPH definitely exhibits mild focus shift at f4-5.6, which coincidentally is exactly the same as Sean Reid reported (back when I still subscribed) on the sample he tested.  However, I find it of no practical consequence in real life shooting and in making prints.  It's a favorite lens, one that I see no reason to replace.

 

Jeff

[sblitz]

Thanks Ed, perhaps this will drive down older lens prices as digital owners swap into the new lenses

 

I have a 75mm sum micron latest version that Leica USA noted was out of alignment. Took them 3 months to get it back but it did make a huge difference on digital, still had my M9 then, and film.

[semi-ambivalent]

Optimized for digital means that the optical formula has been tweaked to neutralize the astigmatism in the image periphery caused by the cover glass. This also means the lens will not perform as well on film as it will have astigmatism in the periphery. That is basically what Zeiss did with the loxia line for the Sony A7. But Sony doesn't have film cameras so no issues there

 

PS. Another example is the summilux 28, which is tweaked for the thicker cover glass of the SL and performs slightly worse on the M. I would expect its film performance to be proportionately degraded.

Edward,

 

Thank you for this. Being fairly ignorant of optics, I had been under the impression the film/digital schism was caused by a lens having a non-planar focused aerial image at the film/sensor. Much like a lens might have curvature of field on 'the other side' of things. This irregularity being accounted for by the thickness of the emulsion, something the sensor, having a thickness of zero, can not do. It never occurred to me that Leica's own design (needing a cover glass of some thickness) would interfere with the performance of its (prior) lenses.

 

Since film cameras are a sideline for Leica these days I suppose all future lenses will be designed with the cover glass's effect on the light path taken into account. Fine. But with your remark about degradation of performance with film, does this exist only as a lab-measurable effect, or is it something that would be visible when using modern fine-grained or tabular emulsions? I'm certainly not in the market for the 28mm Summilux but f/1.4 is of greater use to a film camera than to a digital camera, (DoF, bokeh, and such aside). Tempest in a teapot?

 

Regards,

s-a

[edwardkaraa]

Hello S-A,

 

I think the difference is there but not dramatic, since Leica uses a very thin cover glass on the M (around 1mm) and slightly thicker on the SL. The M works well with film lenses, and SL somewhat less so. I would expect the opposite to be true, SL optimized lenses to be less good on film. How big of an issue it is depends on each user's tolerance and expectations imo.

 

Edward,

 

Thank you for this. Being fairly ignorant of optics, I had been under the impression the film/digital schism was caused by a lens having a non-planar focused aerial image at the film/sensor. Much like a lens might have curvature of field on 'the other side' of things. This irregularity being accounted for by the thickness of the emulsion, something the sensor, having a thickness of zero, can not do. It never occurred to me that Leica's own design (needing a cover glass of some thickness) would interfere with the performance of its (prior) lenses.

 

Since film cameras are a sideline for Leica these days I suppose all future lenses will be designed with the cover glass's effect on the light path taken into account. Fine. But with your remark about degradation of performance with film, does this exist only as a lab-measurable effect, or is it something that would be visible when using modern fine-grained or tabular emulsions? I'm certainly not in the market for the 28mm Summilux but f/1.4 is of greater use to a film camera than to a digital camera, (DoF, bokeh, and such aside). Tempest in a teapot?

 

Regards,

s-a

[gpwhite]

.....  The new MTFs show that an optical design update has improved some things that show up in MTF's (agreement between tangential and saggital lines as well as better values for the high res lines), plus probably some other things that do not show up there....

 

scott 

 

I have not found the MTF estimates for the new lenses on Leica's site (or the rumor sites either). Can you send a source or link? Thx.

 

Curious about any difference in these "estimates" between original 28/2 and the "technologically different" model just released.

[FrozenInTime]

Just how much can we trust new MTF graphs or compare them to older plots ?  

 

Wether they are measured by an instrument or computer simulated - do they now include a cover glass layer equivalent to that on a sensor ?

 

Do we need three set of numbers ; corresponding to:  film ( ideal - insensitive to incident angle  ? ) , M.240 sensor and SL sensor ?

[otto.f]

Two more: M8 and M9 also have thinner cover glasses, M8 the thinnest

[jaapv]

I would expect the opposite to be true, SL optimized lenses to be less good on film.

Why?? I cannot think of a single reason. In fact a more telecentric design can be corrected better nowadays.

[Advertisement]

Hallo,

Look here: Leica M

[edwardkaraa]

Why?? I cannot think of a single reason. In fact a more telecentric design can be corrected better nowadays.

A more telecentric design still has to be large, and the lens designer still has to make a choice whether to include a cover glass in the formula or not, and at what thickness. A lens designed for the SL cover glass will have to compensate for the astigmatism in the periphery, but whenever there is no cover glass, there will be astigmatism. Tests have shown that even telecentric lenses suffer from the same issue.