Leica vs Canon - a surprise find

[bocaburger]

Advertisement (gone after registration)

I've got a 5D1 and the EF 35mm f/2 (bought used for $225). At up to 11x14, IQ is not worse in any way anyone can see than my M9 and 35 Cron. The Canon gets left at home most of the time, because the M9 kit is more comfortably portable. I don't need to convince myself its superior to the Canon in every way and every lens. The M9 satisfies exactly why I bought it. More than that I don't ask and don't care.

[Advertisement]

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

[fotolebrocq]

Comparisons like this are always interesting and I do think the focus bracketing is a good idea just to confirm the findings. I also found my Summicron asph very slightly soft wide open and I started to lose some confidence in the lens. When I bought the new version 2 Voigtlander 1.2 it was a revelation - sharp and good contrast wide open and from f2 bitingly sharp with not a hint of focus shift (which I am convinced is there in the Summicron). Finally I sent the Summicron back to Solms and had it calibrated for the M9 and it is definitely the better for that. I love it because it is such a small and easy to use lens and as a walkaround combo with the M9 it's just about perfect. But I still prefer the Nokton probably because of it's low light capabilities and supreme subject isolation - The new Summilux FLE will eventually replace them both but as it's been on order for 8 months with no sign of imminent delivery that may be quite a while.

 

My experience with the Canon 5D Mkll with the 35L is the opposite of Skinfell's - mine was very soft wide open and pretty much unuseable until f2 - but even then I was never that impressed and sold it after 6 months. Quite probably it was a bad copy as others do seem to like this lens but it just goes to show.

 

Tony

[lct]

...In real life I will have to work more on the canon files than leica files.

So do i with 5D1 pics compared to M8.2's but your pics above look like LR did sharpen the Canon files only or you've got some focus issues more probably. Just a feeling as i have no experience with either M9 or LR, but my sample of the 35/2 asph appears significantly sharper than yours at f/2 (M8.2, 35/2 asph, f/2, 160 iso, 0.7m, C1 default settings, no added sharpening).

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!

[Guest Ornello]

Remember that Leica's design philosophy is intended to complement hand-held cameras. Overall brilliance, tonality, and contrast at lower frequencies are therefore paramount design goals, as these properties 'survive' the occasional slight movement that is inevitable in hand-holding, whereas very fine details (highest frequencies) may be lost. It's the overall impression that matters most, in Leica's opinion.

[lct]

I was not aware of this philosophy but the Summicron 35/2 asph is sharp handheld as well.

It's just that DoF is thin at short distance and the least misfocusing or miscalibration is visible with digital.

(M8.2, 35/2 asph, handheld, 160 iso, f/2, 0.7m, focus on the "IV" label).

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!

[Lindolfi]

Although it is obvious that many Leica lenses have a high modulation in the mid spatial frequencies (10, 20 and 40 linepairs per mm), they also show a high resolution, espcially at f/5.6 and f/8. In fact the resolutions I measured with a microscope in the image plane is often more than two to three times the frequency that can be resolved with the sensor of the M9.

 

That is certainly not lower than that found in other high quality designs that I have measured. So from that point of view Leica lenses do not differ from other designs.

 

The tolerance of handheld exposures is not set by the lenses but by the resolution of the sensor or the film used.

 

I agree with lct that (from my experience) if a Leica lens seems soft, it usually boils down to an inaccurate focus. If you take care of that, there is really no problem.

[Guest Ornello]

Advertisement (gone after registration)

Although it is obvious that many Leica lenses have a high modulation in the mid spatial frequencies (10, 20 and 40 linepairs per mm), they also show a high resolution, espcially at f/5.6 and f/8. In fact the resolutions I measured with a microscope in the image plane is often more than two to three times the frequency that can be resolved with the sensor of the M9.

 

That is certainly not lower than that found in other high quality designs that I have measured. So from that point of view Leica lenses do not differ from other designs.

 

The tolerance of handheld exposures is not set by the lenses but by the resolution of the sensor or the film used.

 

I agree with lct that (from my experience) if a Leica lens seems soft, it usually boils down to an inaccurate focus. If you take care of that, there is really no problem.

 

I am not sure it is 'soft' (it does not look 'out of focus' to me); there are numerous possible factors to account for what is seen here. What I was getting at is that looking at a tiny section is not the way to 'inspect' Leica lens quality. The overall 'look' is what matters most.

 

I have read a lot of Erwin Puts' writings on the subject, and have tested lenses myself, so I am familiar with at least some of the issues involved.

 

It seems to me that a test that used a single roll of film moved between two bodies would be the best way to test these lenses.

[Lindolfi]

I have read a lot of Erwin Puts' writings on the subject, and have tested lenses myself, so I am familiar with at least some of the issues involved.

 

It seems to me that a test that used a single roll of film moved between two bodies would be the best way to test these lenses.

 

Could you please describe how such a test works and how it solves the issue about possible lack of finest detail in images?

[Guest Ornello]

Could you please describe how such a test works and how it solves the issue about possible lack of finest detail in images?

 

Well if you used a single roll of film you would eliminate sensor variables, processing variables, etc. A nice sharp film such as Ektar 100 or T-Max 100 would be a place to start.

 

You get a Canon film body and a Leica film body and do the test. Shoot half a roll, rewind just until the tongue is sticking out, then put it in the other body, advance the film with the lens cap on till you come to the unexposed part. Simple.

[Lindolfi]

A nice sharp film such as Ektar 100 or T-Max 100 would be a place to start.

 

So that is the start, and next? How does it solve the issue of lack of finest detail?

[Guest Ornello]

So that is the start, and next? How does it solve the issue of lack of finest detail?

 

We don't know if there is a "lack of finest detail" in the lenses because there are so many other variables. Doing what I suggest will eliminate most of them.

 

The test you ran did not compare the lenses directly.

[Lindolfi]

This whole thread is about the lack of finest detail in images from the Leica lens of the OP. Some suggestions have been given. Your suggestion was that it was in the design. I gave data from measurements that disproved that. If you decide to measure sharpness (more precise MTF curves) on the same roll of film, the experiment you wish to do still needs a full protocol to make the results meaningful for the OP. So please could you describe how to proceed with a single roll of T-Max 100 in two different bodies?

[Guest Ornello]

This whole thread is about the lack of finest detail in the Leica lens of the OP. Some suggestions have been given. Your suggestion was that it was in the design. I gave data from measurements that disproved that. If you decide to measure sharpness (more precise MTF curves) on the same roll of film, the experiment you wish to do still needs a full protocol to make the results meaningful for the OP. So please could you describe how to proceed with a single roll of T-Max 100 in two different bodies?

 

Not sure what you mean. I would suggest repeating the test using a roll of film. That way no sensor or digital processing would be involved at all.

 

You simply shoot part of the roll in one body, then shoot part of the roll in another. What is difficult to understand?

 

It would of course be well to have cameras that are properly set up, with focussing errors at a minimum, recently in the hands of a competent technician..

[Lindolfi]

Not difficult to understand, just very incomplete.

 

The reason why I use a microscope instead of a film or sensor is that you look directly at the image formed by the lens. In this way you have the same situation for all lenses. Besides, most good lenses outresolve T-Max 100.

The Summicron 35 ASPH runs well beyond the resolution of the M9 sensor (and T-Max 100) according to my microscope. So the lack of finest detail is caused by something else in the images by the OP. The suggestion of Ict of lack of focus is not bad, I think. Another less likely possibility is that one of the glass elements is slightly displaced or some other quality control issue.

[Guest Ornello]

OK, thanks. Clear. Not difficult to understand, just very incomplete.

 

The reason why I use a microscope instead of a film or sensor is that you look directly at the image formed by the lens. In this way you have the same situation for all lenses. Besides, most good lenses outresolve T-Max 100.

The Summicron 35 ASPH runs well beyond the resolution of the M9 sensor (and T-Max 100) according to my microscope. So the lack of finest detail is caused by something else in the images by the OP. The suggestion of Ict of lack of focus is not bad, I think. Another less likely possibility is that one of the glass elements is slightly displaced or some other quality control issue.

 

I am not opposed to using a microscope, but not everyone has access to that sort of thing or knows how to interpret what is seen. I was suggesting something relatively easy to do. It would be easy to put the negative under a strong loupe or microscope.

[skinnfell]

Dear forum members.

I have now repeated the test, this time with focus bracketing on the M9. I did five different settings: Perfect focus (at least in viewfinder), very slight overfocus, slight overfocus, very slight underfocus, slight underfocus. I did two shots at each setting to rule out camera/tripod shake. I even had my wife look over the shots and pick the sharpest ones for me. She had a hard time deciding between two, which ended up being the two at perfect focus.

I define slight overfocus as a 1mm movement on the the barrel.

 

I do not have the files on this computer, but I hope you believe me.

Perhaps my canon 35 is a particularly fine example? In fact the results, as far as sharpness really did impress me. Of course it is pretty lousy in all other regards.

 

I guess the results can be interpreted this way: a lens that performs well but not necessarily best in ALL categories is better than one that does extremely well in one but lousy in all others.

 

I for one cant really afford one Leica lens for ultimate bokeh, one for ultimate F8 performance, one for ultimate low weight, one for ultimate focusing speed and accuracy, etc etc, etc.

 

The 35 cron asph really draws beautifully, despite certain shortcomings, and balances and focuses better than any other lens I have tried.

 

Thanks :-)

[NZDavid]

Well if you used a single roll of film you would eliminate sensor variables, processing variables, etc. A nice sharp film such as Ektar 100 or T-Max 100 would be a place to start.

 

You get a Canon film body and a Leica film body and do the test. Shoot half a roll, rewind just until the tongue is sticking out, then put it in the other body, advance the film with the lens cap on till you come to the unexposed part. Simple.

 

This would be the purest comparison of the lenses by themselves. Use slide film and a high mag loupe (or, as suggested, microscope). Otherwise both can be scanned at the same settings, preferably with a top-quality scanner like an Imacon. Otherwise, this will only ever be a test of one camera and lens combo vs. another. Still very interesting, though.

[Guest Ornello]

Dear forum members.

I have now repeated the test, this time with focus bracketing on the M9. I did five different settings: Perfect focus (at least in viewfinder), very slight overfocus, slight overfocus, very slight underfocus, slight underfocus. I did two shots at each setting to rule out camera/tripod shake. I even had my wife look over the shots and pick the sharpest ones for me. She had a hard time deciding between two, which ended up being the two at perfect focus.

I define slight overfocus as a 1mm movement on the the barrel.

 

I do not have the files on this computer, but I hope you believe me.

Perhaps my canon 35 is a particularly fine example? In fact the results, as far as sharpness really did impress me. Of course it is pretty lousy in all other regards.

 

I guess the results can be interpreted this way: a lens that performs well but not necessarily best in ALL categories is better than one that does extremely well in one but lousy in all others.

 

I for one cant really afford one Leica lens for ultimate bokeh, one for ultimate F8 performance, one for ultimate low weight, one for ultimate focusing speed and accuracy, etc etc, etc.

 

The 35 cron asph really draws beautifully, despite certain shortcomings, and balances and focuses better than any other lens I have tried.

 

Thanks :-)

 

 

Not clear here...are you saying the Leica lens shots that you posted here are in perfect focus? If so, then I was right all along.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[lct]

...I did two shots at each setting to rule out camera/tripod shake. I even had my wife look over the shots and pick the sharpest ones for me. She had a hard time deciding between two, which ended up being the two at perfect focus...

Would you mind to show those perfect focus shots?

[skinnfell]

Would you mind to show those perfect focus shots?

 

Focus was on Hatje Cantz. Steidl is slightly recessed by about 2 mm. Journal is rounded so the lettering is sligtly closer to the camera. All other constants were the sam

 

Ok, here they are:

 

1. Focus as indicated by viewfinder.

 

2. Very slight overfocus. Now Steidl becomes sharper, Hatje Cants slightly blurry with reddish (?) edges.

 

3. Very slight underfocus. Journal becomes slighly sharper, two others are unsharp, in particular steidl.

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!