Leica lens quality/resolution vs MP count...Is this really a "thing?"

[Mikep996]

Advertisement (gone after registration)

I have recently read claims whereby current high MP cameras can out-resolve older lenses.

With a high MP Leica camera is there really a point at which a previous edition Leica lens will not be able to resolve sufficiently to take advantage of the high MP numbers for cropping purposes?  IOW, in real life, can the lens reduce the effective sensor's capability, making, say, a 60MP sensor into a 24MP sensor?  Or, to be more specific, would the most current version of a 50mm Sumimcron out-resolve the 1990-era version of the same lens?  I realize there are other criteria involved with selecting lenses but in this case I am only interested in resolution capability, not other more subjective/artistic factors.

I know a response might be, "If you're happy with the output of your current lens/camera combination, why worry about it?"  My answer is, "Yes, I'm perfectly happy...unless it could be better!"  🤔  

 

 

 

 

 

[Advertisement]

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

[frame-it]

GFX50R [51MP] + Leitz Elmar 35mm LTM [ 1946 ]

 

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!

 

 

 

 

 

[hteasley]

Yes, sensors are, for the most part, out-resolving older lenses, and have been for a while. It doesn't make a 60mpix image into a 24mpix image, and it's not to say that the resulting images are inferior. I have not yet regretted a photo taken with an older lens on my M11 (although now that I have the new 35 Summilux, that's pretty much staying on my camera...) In general, the slightest movement or jostling of the camera while taking your shot will have a greater influence.

Lens sharpness is measured in lp/mm, and it varies across the field and through the apertures. Lower priced lenses are tested to a certain level, perhaps 30 line pairs per millimeter. More expensive lenses are tested to higher resolutions, like 60-90 lp/mm. The companies don't ship lenses that don't meet their resolution criteria.

Google "Airy disk", "Rayleigh criterion", and "modular transfer function" if you want to really grapple with the topic. http://www.normankoren.com/Tutorials/MTF.html is a good page.

EDIT: A lot of sensors can out-resolve newer lenses, too, for given apertures. Older lenses also have more factors that can affect their resolution, like cleaning marks or fungus.

 

 

Edited April 13, 2023 by hteasley

[Mikep996]

I guess the only way to really determine IF there is an effect would be to directly compare the specific lenses at high crops and see if there is any difference.  I should have explained, I am not concerned about pics where no or minimal cropping is done, post #2 clearly shows that an old lens can deliver excellent, sharp photos in that situation.  I am basically looking at the maximum cropping capability of a lens when used "Q2-like" - a single lens to 'simulate' a longer lens.   Again, I am NOT concerned with depth of field differences, etc between a long lens and a shorter lens cropped to produce the same photo, just the sharpness.  FWIW, I have seen fairly large prints of Q2 photos on uTube that were cropped to an equivalent of 176 mm that looked quite good though the equivalent MP count must be down in the "hardly any" range.  😱  No idea what sort of processing they did other than cropping.  I don't do much fiddling with images other than (usually minor) cropping and occasionally bring up shadows;  I don't use any of the usual suspects for digital processing, just Apple photos.

 In my younger, M6 film days I carried a camera bag with the M6, 50 Cron, 50 Noct, 28mm Cron, 90mm Elmarit, table top tripod, flash unit, cable release, etc.  However, I've become less and less willing to carry multiple lenses/gear and my "outfit" when M shooting nowadays is the camera and two lenses - a current 28mm and a 1990s 50 'cron.  So although I think a 28 as a 176mm is too much of a stretch for me, a 50 as a 176mm sounds reasonable!  So I'm interested in any difference in that capability between the 1990's 50 I have and the current 50.

[dpitt]

I read somewhere that the DR Summicron outresolves the Summicron v5. Not sure about the Summicron 50 APO

 

[dugby]

Today’s announcement of the new Summilux-M 50mm stated the following

 

”The new Summilux-M exploits the full performance potential of new camera sensors and offers exceptionally high-contrast and detailed results even in difficult light conditions.”

[Mikep996]

Advertisement (gone after registration)

Sounds interesting!  I've always preferred the "cron" size/weight and don't feel the need for f 1.4 nowadays with digital but what the heck, might try some trade-in negotiation...

[dpitt]

1 hour ago, Mikep996 said:

Sounds interesting!  I've always preferred the "cron" size/weight and don't feel the need for f 1.4 nowadays with digital but what the heck, might try some trade-in negotiation...

I highly doubt it will be better than the Summicron 50 APO.

3 hours ago, dugby said:

Today’s announcement of the new Summilux-M 50mm stated the following

 

”The new Summilux-M exploits the full performance potential of new camera sensors and offers exceptionally high-contrast and detailed results even in difficult light conditions.”

Leica probably is referring to previous versions of the Summilux 50.

[RexGig0]

1 hour ago, Mikep996 said:

Sounds interesting!  I've always preferred the "cron" size/weight and don't feel the need for f 1.4 nowadays with digital but what the heck, might try some trade-in negotiation...

Do not be so quick to trade a much-loved lens! Pay close attention to the optical improvements, in the new ASPH II. The “pre-II” Summilux ASPH is not a flat-field lens, and is not strong in the corners, or at mid-field, if the center is perfectly focused. Any Summilux is only better, if it matches how one wants the image to look. Of course, the corners improve, when the ASPH lens is stopped-down. 

I do not pretend to be any kind of expert, at reading the MTF charts, but there are those who have already said that the ASPH II is stronger in the center, at the expense of being less strong elsewhere in the field. Do not take my word for it. Please read what the truly knowledgeable folks are saying.

A Summicron 50mm will have a flatter field, whether it is the 1979 optical formula, in the Version IV or V, or the newer APO Summicron-M 50mm ASPH. A flatter field is neither better nor worse, but a creative choice.

I started Leica M shooting, with a pre-owned Summilux-M 50mm ASPH. I added the Leica system, for THAT rendering, because none of my SLR lenses could do anything like that. That rendering was due to the optical imperfection, most apparent when the aperture is wide-open. I later added a pre-owned Summicron-M 50mm, Version IV, for the smaller size, or when I just prefer the Summicron rendering. Later, for the times that size is really important, such as wanting to sling the camera under a jacket, vest, or outer shirt, I added an amazingly sharp pre-owned Elmar-M 50mm f/2.8 collapsible lens.

Speaking of size, the ASPH II is larger than the “pre-II.” Look at the specs. It has a shape that makes it appear smaller than it really is.

i do not know whether the ASPH II would benefit my shooting. I can use my APO 75mm ASPH, at 0.7m, to achieve a near-equal magnification as the ASPH II at its MFD of 0.45m. (See Jonathan Slack’s review, just posted today.) This helps justify my acquisition of a pre-owned APO 75 ASPH, early last year, when I opted to upgrade lenses, rather than get a newer M camera. The sharper center of the ASPH II might not be fully realized, with the 24MP sensors in my original M10, and M Type 246 cameras. Only side-by-side testing would show the difference.

 

Edited April 13, 2023 by RexGig0

[01af]

vor 8 Stunden schrieb Mikep996:

I have recently read claims whereby current high-megapixel cameras can out-resolve older lenses.

Roger Ciala (www.lensrentals.com), "Why Perceptual Megapixels are Stupid":

Zitat

If you ask something like ‘is my camera going to out-resolve this lens’ you sound silly.

.

vor 8 Stunden schrieb Mikep996:

... would the most-current version of a Summicron-M 50 mm out-resolve the 1990-era version of the same lens?

Yes, it will. On any sensor. On film, too.

.

vor 8 Stunden schrieb Mikep996:

... can the lens reduce the effective sensor's capability, making, say, a 60-MP sensor into a 24-MP sensor?

No, it cannot. Lenses have no megapixels.

.

vor 8 Stunden schrieb Mikep996:

"Yes, I'm perfectly happy ... unless it could be better!"

A better lens will always be the better lens—on any sensor.

A good-enough lens will always be good enough—on any sensor.

A poor lens will always be poor—on any sensor.

 

[Mikep996]

   8 hours ago,  Mikep996 said: 

... can the lens reduce the effective sensor's capability, making, say, a 60-MP sensor into a 24-MP sensor?

"No, it cannot. Lenses have no megapixels

 

I understand that lenses don't have megapixels and that a lens can't actually turn a 60MP sensor into a 24MP sensor.  What I was wondering was that if the lens isn't sufficiently sharp, could the 60MP sensor show no increased image sharpness over the 24 MP sensor?

[01af]

No.

[UliWer]

vor 29 Minuten schrieb 01af:

vor 8 Stunden schrieb Mikep996:

... would the most-current version of a Summicron-M 50 mm out-resolve the 1990-era version of the same lens?

Yes, it will. On any sensor. On film, too.

No, it won't.

The current version of the 1:2/50mm Summicron-M has the same optics as an example of the same lens type from 1990. The current 50mm Summicron-M is in production since 1979. 

[01af]

Umm ... I interpreted the term 'most-current version' to include the Apo-Summicron-M 50 mm Asph. If you want to stick to 'Summicron' literally (no Apo) then you're right.

Anyway—generally, newer Leica lenses out-perform older Leica lenses. And my point is: they did, and they do, and they will continue to do that on any sensor and on any type of film.

[UliWer]

Gerade eben schrieb 01af:

to include the Apo-Summicron-M 50 mm Asph.

Different lens with different name (and completely different optical design). Never call an "Apo" just "Summicron".

And the general rule that newer Leica lenses will outperform their predecessors might have some exceptions - as always with Leica. Did you look at the MTF-graphs published by Leica for the new 50mm Summilux Asph.? Will it really outperform the former version if you look at the whole frame? There are other exceptions from the "newer will outperform older"-rule from the past. 

[01af]

UliWer, technically I cannot disagree with you—but umm, do you actually get my point ...?

[adan]

There is one place where a higher-Mpixel sensor may show an improvement over a lower-M-pixel sensor - moiré patterns. "Oversampling" of the lens's resolution with more pixels prevents such artifacts.

Tests with my M10 vs an M11 showed the M10 produced moiré (lens resolving a repeating texture right at the same frequency as the pixel pitch ±) while the oversampling of the M11 sensor produced no such sampling errors.

Conversely, a higher-Mpixel sensor may reveal lens flaws (such as color fringing) that do not resolve as obviously with a lower-res sensor. Just as more Mpixels may reveal camera shake that is there but not recorded with a lesser sensor.

But for any other function, it is best to memorize and repeat to oneself, "A sensor, no matter how good or bad, can never improve a lens's image - it can only degrade it less or more. And vice versa."

The MTF of the imaging system (lens MTF x sensor MTF) will always be a number less than 1.00 (100% = perfect reproduction), because there is no MTF possible higher than 1.00/100%.

99.9% x 99.9% = 99.8% (multiplying the parts results in a value less than either one alone).

As the old saying goes, you can't win, and you can't even break even.  

But 99.9% x 99.9% = 99.8% is better than 99.9% x 75% = 74.925%  

[01af]

vor 2 Stunden schrieb adan:

Conversely, a higher-megapixel sensor may reveal lens flaws (such as color fringing) that do not resolve as obviously with a lower-res sensor. Just as more megapixels may reveal camera shake that is there but not recorded with a lesser sensor.

Oh please ... don't repeat this misconception again ...!

.

vor 2 Stunden schrieb adan:

A sensor, no matter how good or bad, can never improve a lens's image—it can only degrade it less or more. And vice versa.

That's right. And I wonder why you don't see how this contradicts your previous statement.

Edited April 14, 2023 by 01af

[Advertisement]

Hallo,

Look here: Leica M

[Mikep996]

Related...is there a noticeable difference in images shot with a current 50mm (standard) Summicron and an APO Summicron that can be seen in photographs?  I have never used an APO lens; never even seen one in person.  The price difference would indicate there is a big improvement in the image using the APO lens.  Of course the price difference between a Leica and a Fuji would indicate the same thing...🤣

In the film era, unless there was some particularly artsy thing being done, we mostly strived for the clearest/sharpest pics we could get.  It's why film and lenses continued to improve.  I haven't changed my view and the idea of the "film" look that some folks strive for now is a recent, and artificial, creation.  Most current pics that demonstrate the alleged "film look" - various imperfections/much visible grain - would have been unacceptable in the film era.

So I'm still wanting the sharpest images I can get - film or digital, especially with the idea of carrying fewer lenses and using the "digital" zoom.

[jaapv]

Total IQ= lens IQ*sensor IQ. It is not a weakest link situation. Unless one or the other is crap ( when the improvement will be marginal) either a better lens or a better sensor -preferably both- will improve the final IQ.