35mp+ 50mm Summilux's vs APO Summicron f2 Centre

[guruguhan]

Advertisement (gone after registration)

I've seen comparisons on lower mp cameras, but I'm curious to know of any comparisons of the ASPH Summilux against the APO Summicron at f2 ignoring the corners?

 

Seeing a comparison of the APO vs ASPH Lux on an M240, from that test alone I'd be quite happy with the ASPH Lux. I'm wondering,however, that on higher mp cameras over the coming years whether the difference will become more apparent?

 

I'm primarily concerned with centre sharpness and micro contrast, as I think the APO will obviously crush the other in the corners.

 

Put another way, at what resolution will the 50mm Lux Asph be out-resolved?

Edited March 27, 2014 by guruguhan

[gpwhite]

As good as the Summilux 50 and 35 ASPH are in terms center sharpness and micro-contrast, the APO Summicron 50 is an another league... from f/2 onwards.

 

I have not used any lens with center performance (on the M9 or M240) that compares to the APO 50 Summicron. I include the Elmar 24mm and SEM 18mm, which are superlative.

 

So, IMHO, even at 18MP, the APO 50 Summicron clearly shows more bite than the Summiluxes. But, that is not to say at all that the Summiluxes do not serve with equal value depending upon your composition. The latter are more selective in DOF and portraits with the APO 50 are not "friendly."

[Advertisement]

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

[01af]

Put another way, at what resolution will the Summilux-M 50 mm Asph be out-resolved?

Never.

 

There is no such thing as a sensor out-resolving a lens. If a lens is good then it will be good on any sensor. If it's poor then it'll be poor on any sensor. As simple as that.

[Paul J]

While semantics say otherwise, real world use, yes it's true, some lenses are better kept to certain resolutions, though it's hard to tell how much will show up with an upgrade in resolution. I upgraded from a P45 to a P65 with my V Blad and later discovered I needed to swap to the H Blad because the lenses are better performing and the camera was better suited. Things like aberrations and diffraction are going to become more apparent, camera issues like sensor plane flatness, build tolerances and shutter issues etc. too, with the higher magnification that a high res sensor provides.

 

The only obvious thing I can think of with current Leica lenses, like all lenses, is diffraction. That is going to become more apparent than what it may show up as now. Given that lenses are generally limited to f16 and diffraction starts kicking in around f8 I see that being a possible issue. It also depends on current build tolerances too, which is hard to comment on at this stage, however I would take a guess things are pretty reasonable there.

 

I think up to 40MP things should be OK. It's hard to say beyond that.

[01af]

... with the higher magnification that a high-res sensor provides.

Resolution doesn't provide magnification.

 

Instead, higher magnification requires more resolution and better lenses. So: how big are you going to print? Bigger than last year?

 

 

The only obvious thing I can think of with current Leica lenses, like all lenses, is diffraction. That is going to become more apparent than what it may show up as now.

The laws of physics regarding diffraction are not going to change anytime soon—so diffraction, and hence diffraction blur, will remain just the way they always were.

[Paul J]

I'm not getting into one of these silly arguments about semantics with you.

 

Yes diffraction will stay the same and that's why it will eventually become an issue with the current design.

[pgk]

Advertisement (gone after registration)

Yes diffraction will stay the same and that's why it will eventually become an issue with the current design.

Diffraction has always been and will always be an issue depending on what you are trying to achieve.

[NightSun]

The laws of physics regarding diffraction are not going to change anytime soon—so diffraction, and hence diffraction blur, will remain just the way they always were.

 

True, but with increasing sensor resolution (about what the discussion was about), that is, with smaller pixels, the blurring effect of diffraction kicks in at larger lens apertures. So with very high resolution (>50 Mpixels) on a full-frame sensor it would be perceptible already at f/5.6.

 

If anyone in doubt, it's useful to read this article:

 

Diffraction Limited Photography: Pixel Size, Aperture and Airy Disks

[jaapv]

Still, 01AF has it right, as is shown by the superlative performance of older, theoretically less-resolving lenses on cameras like the Monochrom and M240. It is always the interaction between the lens and the sensor that determines the end result. Trying to separate the two leads to unfortunate and demonstrably false conclusions.

[BJDrew]

Alas, a 36MP sensor with no AA filter that loves the Leica 50mm lenses does exist today. I have been using it extensively for 4 months. (HINT. It is a Sony A7r)

 

Ignoring the corners is a wonderful idea as MP count increases.

 

If the only thing you really care about is resolution (I hope it is not), you will be happy to know that technique and subject are almost always the limiting factor when using the LUX at 36MP. Yes, when the stars align, the lens still has more to offer. The ability to hang on to detail and provide good contrast in bad light is of particular note (if you happen to care about more than resolution)

 

Looking at center crops from the NEX 7 is another way to predict how things might look at the center of a very high MP sensor (24MP crammed into an APSC sensor size). At base ISO and incredibly good lighting, the lens keeps up. Of course, the AA filter, early diffraction and pixel level noise get a little irritating... but at f2.8 to f4, you can really go wild with resolving power.

 

Forgetting theoretical engineering banter for a moment, when paired with current higher MP (or higher pixel density) sensors, it becomes clear that the lens has more to give in the center frame than a 24mp M240 can extract.

[01af]

... with increasing sensor resolution [...], that is, with smaller pixels, the blurring effect of diffraction kicks in at larger lens apertures.

 

No, it doesn't.

 

 

... it's useful to read this article:

 

Diffraction Limited Photography: Pixel Size, Aperture and Airy Disks

That article promotes the usual nonsense.

[NightSun]

That article promotes the usual nonsense.

 

Then perhaps you could express your argumentation in detail (strictly with respect to the underlying physics) what is that nonsense, as your brief answers do not seem to be justified by anything.

[01af]

Then perhaps you could express your argumentation in detail (strictly with respect to the underlying physics) ....

This is what I did soo many times before; I'm tired of it. Use the search function.

[Paul J]

Still, 01AF has it right, as is shown by the superlative performance of older, theoretically less-resolving lenses on cameras like the Monochrom and M240. It is always the interaction between the lens and the sensor that determines the end result. Trying to separate the two leads to unfortunate and demonstrably false conclusions.

 

While the Monochrom resolves well, it is still only an 18MP sensor. And the difference between the M9 and 240 is marginal at best.

 

This is what I did soo many times before; I'm tired of it. Use the search function.

 

Tell me again, which 60MP system do you own exactly? Because your findings are very different from mine.

[Jared]

I've seen comparisons on lower mp cameras, but I'm curious to know of any comparisons of the ASPH Summilux against the APO Summicron at f2 ignoring the corners?

 

Seeing a comparison of the APO vs ASPH Lux on an M240, from that test alone I'd be quite happy with the ASPH Lux. I'm wondering,however, that on higher mp cameras over the coming years whether the difference will become more apparent?

 

I'm primarily concerned with centre sharpness and micro contrast, as I think the APO will obviously crush the other in the corners.

 

Put another way, at what resolution will the 50mm Lux Asph be out-resolved?

 

That's not how I would recommend looking at it. A higher resolution camera will not make the Summilux a worse lens. It's just that you may not be able to benefit much from the smaller pixels. However, you have to ask yourself whether you really need those extra megapixels anyway--whether your intended use requires the extra resolution that you might be able to capture with a different sensor and a better lens. Do you plan on heavily cropping your image? Do you make large prints? 20 inches or larger? Are you shooting with a tripod? At an aperture where you aren't limited by diffraction? If the answer to all of these is yes, then you may well find with a higher resolution sensor that you can benefit from a better lens than the current Summilux 50. If the answer to any of these questions is 'No,' though, I'd be shocked if you find your pictures are better with a higher megapixel camera than the current M(240) regardless of whether you are using a 50 'lux or a 50 APO. Let's face it, at web resolutions with a hand held shot the limiting factor in image quality is virtually never the camera's megapixel count or the MTF of the lens. Frankly, for anything up to an 8x10 (A4) print size I think a 6mp camera is indistinguishable from even medium format, let alone the current crop of 18 to 36mp bodies.

 

Basically, if you like the results you are getting today with the 'lux, you will continue to like the results with higher megapixel cameras. A better sensor will only show the limitations of the lens under ideal circumstances, and what constitutes 'ideal' gets narrower and narrower as the megapixels go up. Again, if you aren't shooting with a tripod AND making large prints the current cameras are already overkill.

 

I'd recommend trying an experiment for yourself. Take an image of a test target of some sort with your camera tripod mounted and the f/stop set just short of the diffraction limit--say f/8 with the current M(240). Now make a good sized print of that image--optimal output sharpening. Now down-sample your image in Photoshop by 25% and re-print. Notice a difference? You might be surprised how far you have to go before the differences are visible in an 8x10. Now ask yourself, how large do I really print? So how much resolution do I really need? Most of us passed that point a long time ago.

 

- Jared

[guruguhan]

For the record. I want 24x36 300dpi (no enlargement). A7r does meet the spec, but I don't like the camera.

 

Jared - thanks for the tripod tip...that's the one with three legs right? Not at all talking about 8x10s - but thanks for the pro tips!

Edited April 1, 2014 by guruguhan

[Nordvik]

If you want to print 24 x 36 on Epson printers you will have to resample the image even with a Sony A7r. You don't need 36MP to print a fine detailed picture at 24 x 36. With careful post processing an 18MP M9 image will be good enough. 150 ppi is the lowest I will go on a fine detailed picture (without fine details you can print as big as you want). It's important to use as little sharpening as possible in Develop in LR (= before resampling), I set Amount to 25 and Radius to 0,5. Oversharpening will give artifacts. In the Print Module I set Print Resolution to 360ppi (Epson) and Print Sharpening to High.

I am shortsighted and can go very close to the print. I do not see any artifacts using this method.

[Jared]

For the record. I want 24x36 300dpi (no enlargement). A7r does meet the spec, but I don't like the camera.

 

Jared - thanks for the tripod tip...that's the one with three legs right? Not at all talking about 8x10s - but thanks for the pro tips!

 

24x36 is definitely a good sized print. Personally, I would want something that can handle 180 dpi for that size. That means a 36mp camera (to allow a little cropping). For 300dpi with that size print and no resampling you would need something in the 80 megapixel range. I think you can forget a handheld 35mm format for that kind of resolution--not because it's impossible to make the chips, but because it won't take better pictures. Diffraction would limit resolution at around f/4. Even minute amounts of motion blur would defeat the purpose of all those pixels, so you would HAVE to use a tripod. And all that assumes you have nailed focus. Frankly, if these are your requirements you are better off with a larger format than 35mm.

 

Lots of people seem to think that more megapixels result in higher resolution. I'm not saying you are one of them, but there is a general misunderstanding about just how much real resolution is lost from technique. Frankly, for most hand held use (meaning most Leica uses), 18mp is already overkill. Having more megapixels and better lenses will not make any difference whatsoever. The additional theoretical resolution has already been lost to motion blur, depth of field, and focus inaccuracies.

 

I know of a fine art photographer who was renowned for his large format wall-sized prints of fall colors, especially Aspens. The resolution he got was just incredible. The secret? Finding the day with the least wind in order to minimize motion blur. A tripod wasn't enough--he needed an absolutely still subject as well. For 24x36 prints that are critically sharp I think you would benefit much more from a tripod than from more pixels or better lenses. Tripod not practical for your style of photography? That's fine, but you'll never achieve the resolution you seem to be looking for.

 

- Jared

[jmahto]

If you want to be cheap then stitch multiple shots of a good medium telephoto lens (90?) on a cheap sensor. You can get 100+ mega pixel of still objects covering 35mm FOV. Another reason to carry a 90mm in your kit. Its lighter than your tripod.

Edited April 1, 2014 by jmahto

[Advertisement]

Hallo,

Look here: Leica M

[01af]

24 × 36 [millimeters? inches? centimeters? feet?] is definitely a good-sized print. Personally, I would want something that can handle 180 dpi [sic!] for that size. That means a 36-MP camera (to allow a little cropping). For 300 dpi [sic!] with that size print and no resampling you would need something in the 80-megapixel range. I think you can forget a hand-held 35-mm format for that kind of resolution ...

So you actually believe that using a 35-mm-format 36-MP hand-held camera to produce a 180-ppi 24×36-whatever print is fine but using a 35-mm-format 80-MP hand-held camera to produce a 300-ppi 24×36-whatever print is not?

 

You cannot be serious, can you? After all, it's April fool's day ...