Current 50 Summilux due for an update?

[M11 for me]

Advertisement (gone after registration)

This pretty comprehensive comparison should reassure you that the current Summilux can still perform at a decent level;

 

https://wordpress.lensrentals.com/blog/2014/06/comparing-rangefinder-and-slr-50mm-lenses-version-0-7/

 

Pip.

Oh, great article. Good that the Otus is part of the test.

[pippy]

I'm glad you found it interesting, Alex.

 

I, too, liked the fact that the Otus was included because it shows how different philosophies are at work here.

 

Going back, perhaps, to the 'Barnack' ethos; the pre-ASPH Summicron (V4 for instance) weighs 195g; the Otus weighs a staggering 1030g; The pre-ASPH Summicron takes 39mm filters. The Otus has a 77mm filter thread.

 

The brains behind the Otus have, unquestionably, made a wonderful lens. But the brains behind the Summilux andSummicrons have done something, IMO, much more impressive.

 

Pip.

Edited November 22, 2017 by pippy

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[hoppyman]

Mine was purchased 11 years ago and one of my two favourite M lenses. I have used it on a number of film and digital bodies over those years for thousands of photographs..
I think that the APO Summicron M 50 represents their state of the art.
I think, regarding the design and manufacturing process any hypothetical new Summilux might differ from the original approach. 
My lens has actually been back to the factory 3 times, once because staff looked at it during a Forum members visit (2009) and noticed that the focus action had become too loose (at least partly my fault trying to lighten the movement with many many operations), they serviced it to perfection and for free. 
Three years later it appeared to have got out of adjustment and was back focusing quite a bit after those years of continuous use. Again came back serviced perfectly but was well out of warranty.
Then I dropped it a year later. It and my camera needed to be repaired. It was again perfect on return. Four years on from there it is perfectly fine now still, when I use it.

What I can comment on  there is that anecdotally, this design means that the lens is essentially rebuilt for servicing. Perhaps a new design might use different processes for assembly for example.
As far as optical performance, for me it would be very difficult to improve its very high level of performance. Having said that, I very seldom used that lens wide open. If I had to replace it and had the budget, personally I would get that new APO Summicron.
I hardly use my M system now though, shooting almost entirely with the S these days.

[Jeff S]

If I had to replace it and had the budget, personally I would get that new APO Summicron.

.

Then you could be sending it in for a very loose aperture ring (both samples I tested). And the compact design sacrifices the fully knurled focus ring, and places it tight against the aperture ring. I loved it optically when I rented it, but preferred my Summilux ASPH ergonomically.

 

There will be a limited edition 50 APO LHSA version, which might address my concerns, but it sacrifices the nifty built-in hood and will likely cost near $10k.

 

State of the art doesn't mean perfect or issue free.

 

Jeff

[mdemeyer]

Had not seen that one before, thanks!

 

This pretty comprehensive comparison should reassure you that the current Summilux can still perform at a decent level;

 

https://wordpress.lensrentals.com/blog/2014/06/comparing-rangefinder-and-slr-50mm-lenses-version-0-7/

 

Pip.

[Steve Caddy]

Unlike most other camera systems the performance of M lenses on M camera bodies is limited by the data transfer between body and camera (6-bit only) ...

 

 

M lens coding doesn't transfer any data between the lens and camera body, it only serves to identify which lens is mounted. 6 bits is all that is required to encode the ID of every Leica lens. It's not a data pipe or control mechanism as the electronic contacts might be in more advanced systems.

[6bit]

Advertisement (gone after registration)

I wonder what the maximum resolving power for the lens is? With 24MP sensors being it for Leica cameras, will be interesting to see how the future unfolds as higher MP Leica’s come out.

[jmahto]

I wonder what the maximum resolving power for the lens is? With 24MP sensors being it for Leica cameras, will be interesting to see how the future unfolds as higher MP Leica’s come out.

It is certainly more than 24MP. TL (and now CL) certainly resolves it which has smaller pixel pitch than FF 24mp sensor.

[pgk]

M lens coding doesn't transfer any data between the lens and camera body, it only serves to identify which lens is mounted. 6 bits is all that is required to encode the ID of every Leica lens. It's not a data pipe or control mechanism as the electronic contacts might be in more advanced systems.

 

But 6-bit data is a transfer of information from the lens to the camera. You got the point though.

[Steve Caddy]

But 6-bit data is a transfer of information from the lens to the camera. You got the point though.

That’s true, but it doesn’t limit lens performance. The only limitation is on the number of lenses that can be identified: 64.

 

 

Sent from my iPhone using Tapatalk

[wattsy]

That’s true, but it doesn’t limit lens performance. 

 

 

It limits lens performance (on a digital platform) in the sense that the lens can't be tuned to the camera in the same way that 100% digital system lenses like the S and TL can. When the camera knows exactly what aperture is used, what distance the lens is focussed to, etc. more accurate and meaningful corrections can be made in-camera. The 50 Summilux-SL lens is a successful example of this approach and that lens design wouldn't work if the data connection was limited to a basic 6-bit identifier.

[mls1483]

The Summilux 50mm is a modern optical design with very good performance. Stopped down, it is sharp and aberrations are well controlled. Field curvature at f/1.4 is even appreciated (producing the "glow").

 

Of course, newer calculations allow for even better control. Good examples are the APO Summicron 50mm f/2 or the Voigtländer APO Macro Lanthar 65mm f/2 (E mount). However, the required effort at f/1.4 is far greater than it is at f/2, resulting in quite a number of elements and groups, usage of XA and ED glass, etc. In the end, the lenses become very large, bulky and especially heavy. The incorporation of asperical elements into fast lenses also caused a lot of CA with digital sensors (which unofrtunately plagues the Summiluxes like the 28mm or 21mm) and the bokeh deteriorated in addition.

 

Fast update cycles of other manufactures are a race for optimal image quality, fast autofocus, reduced production cost - and especially marketing to keep sales up.

 

I am convinced there will be an update of the Summilux 50mm at some point of time. The new lens will be better in terms of otpical performance and it will be more expensive. But Leica has also always kept their lenses smaller and lighter than the competitor's offers - and that is the biggest challenge in my opinion. The M lives from being small, light and portable. 

[lct]

[...] Field curvature at f/1.4 is even appreciated (producing the "glow"). [...]

 

Would you mind to elaborate? 

[pop]

That’s true, but it doesn’t limit lens performance. The only limitation is on the number of lenses that can be identified: 64. ....

Not quite. The frame lines selected by the lens multiply the number of recognizable lenses by a factor of about 3.

[M11 for me]

M lens coding doesn't transfer any data between the lens and camera body, it only serves to identify which lens is mounted. 6 bits is all that is required to encode the ID of every Leica lens. It's not a data pipe or control mechanism as the electronic contacts might be in more advanced systems.

Just the white dots on the bottom of the lens are „read“ by the sensors on the camera body. And with the possibility to arrange 1 to 6 white dots in any of the six fields on your lens you get 64 possibilities. Each of these possibilities are unique and assigned to a specific lens. And as there are far less Leica M lenses that have to be identified that way the 64 options are plenty.

Edited November 23, 2017 by Alex U.

[pgk]

The 50 Summilux-SL lens is a successful example of this approach and that lens design wouldn't work if the data connection was limited to a basic 6-bit identifier.

 

And in the future where will we be? Internet connected cameras which operate on a cloud-like basis, loading highly detailed profiles for lens fitted and optimising files for all 'correctable' or 'adjustable' parameters based on information relayed by the lens and camera? Purists may not like such an approach but it frees optical designers up to concentrate on the physical requirements of lens design which can't be corrected for utilising data and software. The traditional M approach is on actual optical finesse, I doubt very much that this will continue in the vast majority cameras in the future which, depending on your point of view, maybe a good or bad thing. Personally I like the M approach but I do see the advantages of a fully integrated data transfer approach.

Edited November 23, 2017 by pgk

[Steve Caddy]

It limits lens performance (on a digital platform) in the sense that the lens can't be tuned to the camera in the same way that 100% digital system lenses like the S and TL can. When the camera knows exactly what aperture is used, what distance the lens is focussed to, etc. more accurate and meaningful corrections can be made in-camera. The 50 Summilux-SL lens is a successful example of this approach and that lens design wouldn't work if the data connection was limited to a basic 6-bit identifier.

Ok, sure. Thanks for expanding.

 

 

Sent from my iPhone using Tapatalk

[mls1483]

Would you mind to elaborate? 

 

DOF at f/1.4 is very shallow, in fact 1.5cm at minimum distance of 0.7m. The quite significant field curvature of the Summilux 50mm thus results in a certain loss of sharpness around the object (blur), which is recognized as object sepration ("pop").

 

Zeiss lenses are optimized for contrast, Leica lenses for sharpness.

 

The transition between sharp and unsharp areas is rather gentle for Leica lenses.

[pgk]

DOF at f/1.4 is very shallow, in fact 1.5cm at minimum distance of 0.7m. The quite significant field curvature of the Summilux 50mm thus results in a certain loss of sharpness around the object (blur), which is recognized as object sepration ("pop").

 

Zeiss lenses are optimized for contrast, Leica lenses for sharpness.

 

The transition between sharp and unsharp areas is rather gentle for Leica lenses.

 

I think that you will find that Zeiss actually produced a paper on 'bokeh' and transitions from 'sharp to unsharp' and are aware of the maths which control its appearance/effect. Field curvature isn't a part of it. Contrast is down to the ability to control and minimise straying light within a lens (usually very small now). Lastly, DOF is shallow but mathematically calculable providing you define your requirements so any figure is only valid if the relevant caveats are described.

[Advertisement]

Hallo,

Look here: Leica M

[mls1483]

I think that you will find that Zeiss actually produced a paper on 'bokeh' and transitions from 'sharp to unsharp' and are aware of the maths which control its appearance/effect. Field curvature isn't a part of it. Contrast is down to the ability to control and minimise straying light within a lens (usually very small now). Lastly, DOF is shallow but mathematically calculable providing you define your requirements so any figure is only valid if the relevant caveats are described.

 

I was not talking about sharpness within the focal plane (resolution). Field curvature means the focal plane is bent. The mathematical description of this "Petzval area" was published by Petzval in 1843: Neglecting astigmatism, a field ist flat if the sum of the reciprocal refractive indexes times the focal lengths (for each lens) equals zero. In case of astigmatism, it is the mean of the reciprocal saggital and tangential values. Astigmatism occurs, if saggital and tangential values in the MTF differentiate. The MTF show the sum of sharpness and contrast. Thus: Field curvature results in unsharp areas at the edges of the picture, whereas the center is sharp (or vice versa). The marvellous CV Ultron 21mm f/1.8 for instance is plagued by field curvature.

 

Sharpness (resolution) is measured as lp/mm at a given contrast (usually 50%).