35mm f/2.4 or 2.5 Summarit

[Guest]

Advertisement (gone after registration)

Hi all,

My current kit is the M246 Mono paired with the 35mm f/1.4 Summilux FLE. I mainly shoot people photography in reportage style.

I have been shooting M (previously M262 before switching to M246) with 35mm exclusively for the last 2 years.

The more I shoot with the M, the less I shoot at f/1.4 and more stopped down to get more context in my images.

Hence, I have lately been considering switching to the 35mm Summarit for it’s lower weight and lighter/faster focus ring compared to the 35FLE.

How does the 35mm Summarit compare to the Summilux FLE in the aperture range f/2.4-5.6? How is rendering between the two - almost identical or will the Lux give more sparkle to the images?

I do love the 35FLE, but the size, weight and ergonomics of the Summarit is tempting.

Am I crazy? 

Thanks,

Mads

[Advertisement]

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

[lct]

Cannot say for the 35/2.4 but the 35/2.5 is softer at edges and corners below f/8. It has more CA there too. Its bokeh is rather softer though.

[Stuart Richardson]

I think that sums up my experience as well. The 35mm Summilux is the better lens, but the Summarit is very good and usable in most circumstances. CA and bokeh fringing are worse, however. I only had the 2.5, not the 2.4. I eventually sold the Summarit as I did not mind as much putting up with the weight. Another more appropriate consideration might be the 35mm Summicron ASPH? I have not used that one, but it is perhaps a bit closer in performance to the FLE, since it is also an aspherical design.

If you are shooting black and white only, I think the CA and bokeh fringing will not be relevant, so if you can give up a bit of crispness wide open, you will likely not notice much difference.  I compared them side to side at the time and finding that there was not so much between them, but it has been awhile...

Edited January 4, 2020 by Stuart Richardson

[adan]

"Sparkle" is not something Leica charts for their lens performance.

However Leica's own MTF graphs show that:

1) the f/2.4 version of the Summarit has slight lower center MTF (60%/40lpmm), than the f/2.5, but less of a sag or valley at about 12-15mm out from the center (30% vs 20%). The corner MTF is virtually identical (low, at 10% at f/2.5, 30% at f/5.6).

2) At the apertures you mention (2.8-5.6) the Summilux FLE well ahead at the center (68%) and the corners (40-70%), but again a bit weaker at 12-15mm out from the center (28% - that would be at the top and bottom of a horizontal picture, and the equivalent-radius ring around the center).

Really good corner performance at wide apertures is one of the "signatures" of the 35 Summilux ASPH designs, even from 1994 - Leica knocked down the blurring/streaking effects of coma and astigmatism to virtually zero (pinpoint "Xmas lights" image as dots, and not cones or butterflies), even at f/1.4. Extremely impressive.

My subjective experience with the Summarit 35 f/2.5 is that it has "gross" sparkle (global high contrast) but without the Summilux's "delicate" sparkle of fine details (edge contrast and clarity). I could quote from a friend's description of Nikkors 20 years ago - "Lots of contrast - but in all the wrong places."

[Guest]

2 hours ago, adan said:

"Sparkle" is not something Leica charts for their lens performance.

However Leica's own MTF graphs show that:

1) the f/2.4 version of the Summarit has slight lower center MTF (60%/40lpmm), than the f/2.5, but less of a sag or valley at about 12-15mm out from the center (30% vs 20%). The corner MTF is virtually identical (low, at 10% at f/2.5, 30% at f/5.6).

2) At the apertures you mention (2.8-5.6) the Summilux FLE well ahead at the center (68%) and the corners (40-70%), but again a bit weaker at 12-15mm out from the center (28% - that would be at the top and bottom of a horizontal picture, and the equivalent-radius ring around the center).

Really good corner performance at wide apertures is one of the "signatures" of the 35 Summilux ASPH designs, even from 1994 - Leica knocked down the blurring/streaking effects of coma and astigmatism to virtually zero (pinpoint "Xmas lights" image as dots, and not cones or butterflies), even at f/1.4. Extremely impressive.

My subjective experience with the Summarit 35 f/2.5 is that it has "gross" sparkle (global high contrast) but without the Summilux's "delicate" sparkle of fine details (edge contrast and clarity). I could quote from a friend's description of Nikkors 20 years ago - "Lots of contrast - but in all the wrong places."

Thank you all for your informative answers.
 

Adan, your detailed explanation helped a lot. Sounds like, I should stick with the 35FLE, which is a lens I truly love. And considering it is a 35mm f/1.4, the weight and size is still quite small compared to other systems. The summarit would be too much of a compromise and I would loose the option of f/1.4 when needed.

I pressume, you also would recommend the more “delicate” nature of the Summilux’s contrast and clarity for pairing with the mono sensor of my M246, compared to the higher contrast nature of the summarit?

Thanks once again,

Mads

[tjv]

What about the Zeiss ZM 35mm f2.8? Great lens, and very sharp!

[easy_action]

Advertisement (gone after registration)

6 hours ago, adan said:

1) the f/2.4 version of the Summarit has slight lower center MTF (60%/40lpmm), than the f/2.5, but less of a sag or valley at about 12-15mm out from the center (30% vs 20%). The corner MTF is virtually identical (low, at 10% at f/2.5, 30% at f/5.6).

Interesting.  Having followed some of the discussions I understood that they are the same optical design.

Is this possible given the above?  (Could the different MTF results be down to sample variation?)

[easy_action]

8 hours ago, mcpallesen said:

I do love the 35FLE, but the size, weight and ergonomics of the Summarit is tempting.

Am I crazy? 

I used the 35 FLE almost exclusively for a couple of years, but was never completely happy with the size - with the hood it felt like quite a long and 'fat' lens.  Very subjective, of course.

I now have a 35 summarit (the f2.5), and don't use the hood.  It feels significantly smaller than the FLE.  No doubt it's technically not as good, but honestly I just don't notice.

f/1.4 is nice to have creatively, but for me the small size is more valuable.

[Jeff S]

7 hours ago, Stuart Richardson said:

 Another more appropriate consideration might be the 35mm Summicron ASPH? I have not used that one, but it is perhaps a bit closer in performance to the FLE, since it is also an aspherical design.

 

The 35 Summarits (both) are also aspherical designs, just not well known as such.  Puts praises them, in part, for exhibiting less flare, and with less tendency to focus shift, than the 35 Summicron ASPH, which is also a fine lens.

Jeff

Edited January 5, 2020 by Jeff S

[adan]

4 hours ago, easy_action said:

Interesting.  Having followed some of the discussions I understood that they are the same optical design.

Is this possible given the above?  (Could the different MTF results be down to sample variation?)

I belive Leica's published MTF charts are based on the computer simulations that create the design. I.E. they show the theoretical performance the computer predicts.

Unlike (possibly) Zeiss or third-party testers, who may do tests of an actual prototype sample on an optical bench.

 But I could be wrong - or have that reversed.

You are correct that "the word on the street" is that the two Summarit versions are the same glass in the same configuration, and the fact that Leica's drawings in their "Technical Data" sheets are identical (but not especially high-res, given the wavelengths of visible light) tends to support that.

But that doesn't mean Leica didn't make some tiny change to the curvature of, say, the molded ASPH surface. That would not show up on the drawings as a visible difference - but might show up in the simulated or tested results.

Here are Leica's ~f/2.8 graphs for the three current 35mm lenses

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Geek Alert: I've also included a thumbnail of "how to read a Leica MTF chart" for those who haven't tried it yet. Just keep in mind that MTF is not the be-all and end-all - just a fairly objective measurement of apparent clarity and "sharpness" for quick comparisons:

Basically, the charts graph the contrast (X axis) of line pairs of specified spacing at various distances (Y' axis) from the center of the image (0mm) out to the corners (21mm).

The solid lines are for lines/patterns running out from the center of the image radially, the dashed lines are for lines/patterns at 90° to those (tangential to the edge of the image circle - useful for catching streaking or smearing in the corners).

Leica uses 4 sets of solid/dashed lines, for resolutions of 5, 10, 20, and 40 black/white line pairs per mm (lpmm) as focused on the film or sensor. Generally, the bottom two sets of lines (20 and 40) tell more about resolution (but not everything), while the top two sets of lines (5 and 10 lpmm) reflect overall lens contrast and "punch."

Thus we might note that the 35 Summicron ASPH will image 40 lpmm, in radial orientation, at about 66% contrast in the center of the image (0mm), and then the MTF will go up and down and up and down moving towards the corner of the picture, ending up at about 27% at the outermost corners (21mm from the center). And that the deep valley at about 12-14mm with all three lenses may mean "rule of thirds" portrait compositions at f/2.8 could be disappointingly - soft.

The higher the line, the better. For 40 lpmm, >50% contrast is very good, >70% is very very good, anything below 30-40% is getting a bit "iffy" - although that depends on whether one is pixel-peeping, or looking at a smallish print. And some beloved Leica lenses like the 50 f/1.0 or pre-ASPH 50 f/1.4 are shocking bottom-feeders, with 40lpmm contrasts down below 10% over half of the image, at their maximum aperture.

Ultimate resolution has to be interpolated, since we know that good lenses from any manufacturer may resolve up to 80 to 100 lpmm or more - but those resolutions are not charted. And there are a host of things (color fringing, coma, field curvature, etc.) that go into the final contrast, but are not broken out as specific contributors - we don't know WHY the Summicron has 28% contrast at the corners, or the Summarit has 10%, without other tests. Although to the educated eye, the divergence of the solid and dashed lines may give clues.

Edited January 5, 2020 by adan

[poli]

3 hours ago, adan said:

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Geek Alert: I've also included a thumbnail of "how to read a Leica MTF chart" for those who haven't tried it yet. Just keep in mind that MTF is not the be-all and end-all - just a fairly objective measurement of apparent clarity and "sharpness" for quick comparisons:

 

Thanks for this explanation. Very educational. Finally understand how to read those graphs now. 
 

one question, the summariy graph you posted seems to be a f2.4 graph? Would that make much difference in result? 

[wlaidlaw]

I have never been a big fan of the Summarit range. I would rather go for the tiny and excellent 35/f2 ASPH Summicron. If the budget does not stretch to a new one, there are plenty of S/H good examples to choose from. The other alternative but physically slightly larger is the Zeiss 35/2 Biogon. It is not as good wide open as the Summicron, as the Biogon design is near or maybe beyond its limit at f2. However even half a stop down it is very close or equal to the Summicron and for flare resistance, it probably beats the Summicron. They are available for about the same price new as the Summarit 2.4. The Voigtlander lenses are another possibility but my experience is that QC is not great and there is quite a range of sample variation on their lenses. Since the Zeiss ZM lenses are made in the same factory and seem to have much less variation, I can only assume that Zeiss' inspection standards are rather stricter than CV's.

I keep thinking about selling my 35 ASPH Summilux as I rarely use it since I got the 35 Summicron but as I was lucky enough to find a good one of these lenses (one out the final chrome batch from September 2006), with very moderate aperture shift, I may keep it. Many examples of this lens have horrendous aperture shift. I  usually regret selling any of my Leica equipment after the act.

I have just acquired yet another Leica 35mm lens, a single cam Elmarit-R Mk.1 on a Leicaflex SL2 but it needs a clean to remove a small amount of fogging. 

Wilson

PS Correction. Now that I have looked at the photo of the back of my 35 Elmarit a bit more closely and checked from the serial number that it is a late Mk.1 lens from 1971, it is the 2 cam version of this lens

 

 

Edited January 5, 2020 by wlaidlaw

[otto.f]

14 hours ago, adan said:

Leica's own MTF graphs

Although at Karbe’s last lecture in Wetzlar it seemed to turn out that Leica’s MTF’s are calculated and not measured. I’m baffled, don’t know what to think about that.

[Guest]

On 1/5/2020 at 6:13 AM, adan said:

I belive Leica's published MTF charts are based on the computer simulations that create the design. I.E. they show the theoretical performance the computer predicts.

Unlike (possibly) Zeiss or third-party testers, who may do tests of an actual prototype sample on an optical bench.

 But I could be wrong - or have that reversed.

You are correct that "the word on the street" is that the two Summarit versions are the same glass in the same configuration, and the fact that Leica's drawings in their "Technical Data" sheets are identical (but not especially high-res, given the wavelengths of visible light) tends to support that.

But that doesn't mean Leica didn't make some tiny change to the curvature of, say, the molded ASPH surface. That would not show up on the drawings as a visible difference - but might show up in the simulated or tested results.

Here are Leica's ~f/2.8 graphs for the three current 35mm lenses

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Geek Alert: I've also included a thumbnail of "how to read a Leica MTF chart" for those who haven't tried it yet. Just keep in mind that MTF is not the be-all and end-all - just a fairly objective measurement of apparent clarity and "sharpness" for quick comparisons:

Basically, the charts graph the contrast (X axis) of line pairs of specified spacing at various distances (Y' axis) from the center of the image (0mm) out to the corners (21mm).

The solid lines are for lines/patterns running out from the center of the image radially, the dashed lines are for lines/patterns at 90° to those (tangential to the edge of the image circle - useful for catching streaking or smearing in the corners).

Leica uses 4 sets of solid/dashed lines, for resolutions of 5, 10, 20, and 40 black/white line pairs per mm (lpmm) as focused on the film or sensor. Generally, the bottom two sets of lines (20 and 40) tell more about resolution (but not everything), while the top two sets of lines (5 and 10 lpmm) reflect overall lens contrast and "punch."

Thus we might note that the 35 Summicron ASPH will image 40 lpmm, in radial orientation, at about 66% contrast in the center of the image (0mm), and then the MTF will go up and down and up and down moving towards the corner of the picture, ending up at about 27% at the outermost corners (21mm from the center). And that the deep valley at about 12-14mm with all three lenses may mean "rule of thirds" portrait compositions at f/2.8 could be disappointingly - soft.

The higher the line, the better. For 40 lpmm, >50% contrast is very good, >70% is very very good, anything below 30-40% is getting a bit "iffy" - although that depends on whether one is pixel-peeping, or looking at a smallish print. And some beloved Leica lenses like the 50 f/1.0 or pre-ASPH 50 f/1.4 are shocking bottom-feeders, with 40lpmm contrasts down below 10% over half of the image, at their maximum aperture.

Ultimate resolution has to be interpolated, since we know that good lenses from any manufacturer may resolve up to 80 to 100 lpmm or more - but those resolutions are not charted. And there are a host of things (color fringing, coma, field curvature, etc.) that go into the final contrast, but are not broken out as specific contributors - we don't know WHY the Summicron has 28% contrast at the corners, or the Summarit has 10%, without other tests. Although to the educated eye, the divergence of the solid and dashed lines may give clues.

Adan, this was very insightful. I have read quite a bit on these forums about the 35FLE’s “mid zone dip”/wavy MTF curves. Some wax a lot about it, and how it changes with different apertures. And that the 35FLE is inconsistent in the images it renders at different apertures. Almost to the point where it sounds like the lens is impossible to shoot. 😉

My question is, how much should the photographer think about it/adjust for it in real life shooting? Or should one just shoot at the aperture required for the image and DoF, and not stress further about this mid zone dip? I.e. just shoot and focus on the creative part of the image making process and not stress about these lens technicalities. I would be inclined to think so, as I would think the 35FLE was designed with reportage/photojournalism in mind.

Seems like all 3 of Leica’s current 35mm lenses has this behaviour according to the MTF charts. Would a more smooth drop off from center to edge (like the Zeiss 35mm 2.8) lead to a more predictable lens compared to this more wavy pattern? Or is it chosen in the design process, because it is deemed insignificant for the naked eye in real world shooting?

Hope my questions make sense.

Thanks,

Mads

[adan]

2 hours ago, mcpallesen said:

Seems like all 3 of Leica’s current 35mm lenses has this behaviour according to the MTF charts. Would a more smooth drop off from center to edge (like the Zeiss 35mm 2.8) lead to a more predictable lens compared to this more wavy pattern? Or is it chosen in the design process, because it is deemed insignificant for the naked eye in real world shooting?

The really short answer is  - sort of.

The not-so-short answer:

Lens designers have to corral a significant number of different aberrations, all of which reduce image quality at some, but not other parts, of the image area. Every change they may calculate in any part of the lens may improve one aberration, but degrade another. In trying different glass types and curvatures, they may come up with a design with more contrast/resolution in the "valley" - but produce less contrast/resolution in the center or the corners, or add color fringing or distortion, and so on.

Or, conversely, they can get better center performance (only?) by letting the valley sag.

Size can also be a factor. As can costs - glass X for a certain element may produce 5% less sag than Glass Y - but if glass X costs 10 times as much per lens.....is it worth it? (Keep in mind that relatively low costs are a "feature" of both the Summarit and the Zeiss ZM f/2.8.) Rumor has it that just one of the elements in the 50mm Summilux f/1.4 ASPH costs more than the other seven elements combined.

The computer can tell the designer what will happen with various changes to the design - it is up to the individual designer make a creative decision as to which combination of strengths and weaknesses is "the best."

It is not so much that a weakness is "insignificant to the naked eye," but that it is - even if visible - less important, either to the designer, or to his/her perception of users' tastes or expectations.

I'd note that the Zeiss 35 f/2.8 is larger and and has more vignetting than, say, the original Summarit f/2.5 (even with its extra 1/2 stop of speed). The Zeiss has slight pincushion distortion, the Summarit has slightly larger barrel distortion. There may be differences in color fringing that are not charted, as well (I don't know).

The Zeiss and Summarit designers obviously placed differing values on: MTF at 12mm radius, vignetting, compactness, and distortion of this or that type. Sacrifice one to improve the other.

Anyone can disagree with the designer's creative decisions, and prefer one lens (or designer's "philosophy") over the other.

Further reading:

https://en.wikipedia.org/wiki/Optical_aberration

Pages 4-10 in particular: https://www.overgaard.dk/pdf/Leica-M-Lenses-Their-Soul-and-Secrets_en.pdf

[Stuart Richardson]

Another thing to consider is that MTF is the measure of sharpness at one plane of focus: the focused distance (presumably the center).  You may have areas of the image in that part of the field that are very sharp, but they might be a bit closer than the point of focus at the center of the frame. Lenses do not always have a flat field...focus curves naturally, especially with wide angles. As an example, let's say you have a 35mm lens with field curvature and you want to take an environmental portrait. You are photographing a person standing in front of a desk in an office. You focus on their eye, and the eye is sharp. A lens with a flat field will render the eye and everything at that distance as critical sharp. A lens with field curvature might render the eye critical sharp, but then the focus curves inwards, so as you get to the edge of the frame, the point of critical focus gets a little closer. In an lot of compositions, this can actually be helpful...most photographs taken out in the world with 35mm cameras are of three dimensional subject matter, so you might not even notice in use...it may be covered in depth of field, or even benefit your composition. Flat field lenses are more straightforward to use in some cases, particularly for photographing flat subject matter or landscapes at infinity, but not all lenses with field curvature are bad...

This page had the best explanation I have seen. https://tashley1.zenfolio.com/blog/2013/5/field-curvature---a-practical-guide

[Guest]

Thank you so much, Adan and Stuart. Really helpful explanations and makes a lot of sense.

I will continue to enjoy my 35FLE 😊 It’s a great lens for my photography.

Best,

Mads