Leica vs Canon - a surprise find

[skinnfell]

Advertisement (gone after registration)

My two favorite lenses are the Leica Summicron-M 35mm ASPH and the Canon Ef 35mm 1.4L mounted respectively on a M9 and a 5D2.

 

I have used them extensively, and shot thousands and thousands of frames with them, probably well over a hundred thousand with the canon (I am a professional photojournalist).

 

Yet I never thought of doing a direct comparison. Until now.

 

So for fun I set up my gitzo carbon tripod in front of my bookshelf, aligned the cameras as good as I could without going all scientific. Shot the images with self timer.

 

Set both cameras to ISO 160, shot comparison frames at F2 and F8. You know, one for action and one for landscapes and architecture.

I gave the Canon a 1 stop aperture ahead start just because its well under half the price of the leica. Besides I very rarely shoot the canon under F2 (but thats another story). Focus distance was about 1 metre. Canon had a UV filter that is dented so I cant remove it. Leica had no filter. I have lost the sunshade for the canon, but used the original sunshade on the leica.

 

In other words I tried to do it as best as I would be able to do it in the field, under normal working conditions.

 

So here are my findings (which I am somewhat surprised with):

Image 1 - At F2 in the center, canon is significantly sharper than leica

Image 2 - At F2 in the corner, canon is significantly sharper than leica

Image 3 - At F8 in the center, leica is slightly sharper than canon

Image 4 - At F8 in the corner, leica is slightly sharper than canon

Image 5 - At F8 in the corner, canon has much more distortion, fall-off and CA than leica.

 

Other findings, not easily visible in these shots:

-Canon has much more light-fall off than leica at all apertures

-Canon has much more colored fringes than leica particularly in the corners. In fact leica has hardly any at any aperture.

- Canon has much more barrel distortion than leica at all apertures. Leica does have some distortion but it is hardly visible in any of my test shots.

- Leica has better color, contrast and 3D, rendition at all apertures. This could be the sensor but still.

- Canon has smoother bokeh, particularly at mid-wide apertures like F4. Here the leica is quite busy in the background.

- Leica weighs 250 grams, canon 580 grams.

 

I have actually owned both lenses twice, and comparing the results from them shows no difference. Also, both lenses are pretty much perfectly focus-calibrated on my cameras, so no issues with shift as far as I can see.

 

I guess this kind of took me by surprise. I repeated the test and got the exact same results. I think this is more a tip of the hat towards the Canon 35 1.4, than detriment to the leica. I have never gotten a picture with the leica I wasnt happy to blow up because of issues with the lens. The CA and distortion in the canon has irritated me at numerous occasions. I suppose this created my belief that it was a far inferior lens. I now see then more as equals (but still prefer the leica hehe)

 

Around the interwebs there are rumors of both lenses being updated soon, so I might have to replace them both and repeat the fun...

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!

[IWC Doppel]

Perhaps a surprise, interesting to see 115 views and no comments

 

It's how the Leica paints that sells Leica to me, pixel peeping is a valid reference but tells only part of the story. As I write I feel like I might be seen as defending a seemingly lower performance solution. But sharpness and resolution don't show the complete picture so to speak

[Guest Ornello]

Perhaps a surprise, interesting to see 115 views and no comments

 

It's how the Leica paints that sells Leica to me, pixel peeping is a valid reference but tells only part of the story. As I write I feel like I might be seen as defending a seemingly lower performance solution. But sharpness and resolution don't show the complete picture so to speak

 

I think you're seeing the effects of (camera-based) sharpening more than anything else. Do this with film and I bet you'll see quite a different result.

[tobey bilek]

Is the lens made for such such focus distance. A proper test would be further.

[lct]

Just to check the number of variables:

1. Did you focus bracket your M9?

2. Did you shoot raw?

3. If so, what raw converter(s) did you use?

4. What sharpening setting(s) did you use?

[NZDavid]

From this test it would appear this Canon lens beats this Leica lens, sometimes. Moire is also a woire. But it is not really solely a test of lens vs. lens, but also sensor vs. sensor and software vs. software. With the extra variables it has become much harder to test any lens by itself any more in the digital era -- something Erwin Putts acknowledges on his site.

[Planistigmat]

Advertisement (gone after registration)

Really fine test. Thank you !

 

Leica weighs 250 grams, canon 580 grams.

Difference in weight is because lens for Leica M is for camera without mirror. Lens for Canon EF is lens for camera with mirror which means that distance between rear optic element and sensor/film should be longer than in case when we don't have mirror (valid for wideangles only). To have this distance longer retrofocus construction is needed. We need front big "minus element" in lens. So weight is bigger.

[pgk]

Can I suggest that you go back to the files and 'optimise' them by dealing with CA and adjusting colour and sharpening to see just what similarities and differences there are?

[wattsy]

Love the moire in one of the later shots - classic M9.

 

Yangtze - The Long River. Very fine book by a great photographer.

[Studio58]

As posters have said, and I agree, there are so many variables it is difficult to conduct a test that yields definitive results... Another test with a different set of variables would yield another result again. A slippery slope indeed.

[Planistigmat]

From one hand it is hard to compare lenses and cameras between two sets (Lieca M9 + 35mm vs. C5D + 35 mm). From the other hand test is good when you go to take photos and you have to choose 1st set or 2nd one. You don't think about sensors, CA, software ..... Just choice and than results.

[skinnfell]

Thanks for comments.

 

I wanted to do the test, replicating best possible field conditions - in my workflow.

 

So that meaans canon RAW and uncompressed DNG.

All other settings were standard lightroom settings, with one exception - I have my own color profile for the M9.

 

The results we are seeing are of course the end result of the combination of lens sensor and camera (and also lightrooms raw processor).

 

the sharpness issue at F2 isnt really a problem to me since that aperture mostly only is used for action shots. If the light permits I will usually always shoot between F4 and F8 with both lenses.

In the olden days I used to shoot everything at max aperture, but got kinda tired of the "dreamy" look, so nowadays I always try to make compositions that include the background and sometimes foreground. It also allows me to shoot more "from the hip" to grab moments.

 

At the middle apertures, the leica totally blows the canon out of the water in all regards. Flare, distortion, ca, sharpness, contrast, threedimensionality, etc are all much better.

 

Also, its a big deal to me to make images in the camera that a minimum, if any, postprocessing is needed. (again because I do this for a living and sometimes I dont have time to process my own pics)

 

All I wanted to prove was that they are very different lenses but very good in their own regard. After all that is why I own and use both.

 

Included is a picture I made yesterday. Completely out of the camera (although DNG), no cropping, no extra nothing. The leicas ability to produce pictures that are ready to go is a great lifesaver for me. Very rarely can I spend an hour toucing up pictures.

Get the exposure and WB right, and its ready to go.

 

(35 cron, F2, iso1250, 1/60, Wb is tungsten preset)

The picture shows a bunch of politicians being shown museum under construction.

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!

[lct]

Would you mind to answer my questions above?

[Studio58]

Get the exposure and WB right, and its ready to go.

 

 

Yes that is what blew me out when I first shot with the Leica. The images were very usable with no PP at all. After years of shooting with Canon's top models, I was not expecting a result of that nature.

[skinnfell]

Just to check the number of variables:

1. Did you focus bracket your M9?

2. Did you shoot raw?

3. If so, what raw converter(s) did you use?

4. What sharpening setting(s) did you use?

 

1.I did not focus bracket, because I have found my 35 cron to be very well calibrated. I did however repeat the test and got identical results.

I would never focus bracket in real life either, because If im using a tripod I am usually not shooting at F2.

I used 10x live view to focus the canon. In real life (w/o live view) i have found my canon to be so-so on focus accuracy, which is why I dont shoot under F2.

 

2. yes i shot uncompressed DNG and standard CR2

 

3. I used LR 3.6 which is what I use for work. Other converters might work better, but LR is what I am stuck with. I also have CS5 creative pack but as far as I can understand it uses the same raw converter.

 

4. Sharpening settings were standard for LR 3.6 (I never add sharpness unless I resize an image). No "clarity" or "knoll" sharpening added.

In other words straight from camera. In real life I will have to work more on the canon files than leica files.

[lincoln_m]

Little sad to see the M9 performance. Is that noise at ISO160 in the shadows of the M9 shots? Chroma noise on the sack colour cotton/hemp spines too from the Leica M9. I'd hope for better noise performance at min ISO from a £5K camera/sensor.

 

Zeiss do lenses with M-mounts or Canon EF but that could allow the same lens( ah no RF and SLR are different) to be tested on different cameras. Mr Putts says Zeiss M-mounts aren't beating Leica ASPHs

 

Do we have to go to the Leica S2 to get better performance? But with 37Mpixel one could go for Nikon D800 at 1/10th the price.

Regards, Lincoln

[wattsy]

1.I did not focus bracket, because I have found my 35 cron to be very well calibrated. I did however repeat the test and got identical results.

I would never focus bracket in real life either, because If im using a tripod I am usually not shooting at F2.

I used 10x live view to focus the canon. In real life (w/o live view) i have found my canon to be so-so on focus accuracy, which is why I dont shoot under F2.

 

The trouble with not focus bracketing is that your test isn't comparing the lenses but the difference in focus accuracy between your particular M9 RF and the Canon live view. I've owned three different 35 Summicron ASPH lenses and whilst I think the sharpness at F2 is often overrated in this forum, it is usually better than shown in your wide-open shot. It does look like a little tweak of the focus would have improved the resolution at F2. I'm sure the calibration of your M9 RF is fine for real world use (how often do we photograph flat objects?) but I suspect it is not bang on to be trusted for this kind of test.

[Guest Ornello]

The trouble with not focus bracketing is that your test isn't comparing the lenses but the difference in focus accuracy between your particular M9 RF and the Canon live view. I've owned three different 35 Summicron ASPH lenses and whilst I think the sharpness at F2 is often overrated in this forum, it is usually better than shown in your wide-open shot. It does look like a little tweak of the focus would have improved the resolution at F2. I'm sure the calibration of your M9 RF is fine for real world use (how often do we photograph flat objects?) but I suspect it is not bang on to be trusted for this kind of test.

 

I doubt that the focussing is off. Probably processing differences.

[skinnfell]

The trouble with not focus bracketing is that your test isn't comparing the lenses but the difference in focus accuracy between your particular M9 RF and the Canon live view. .

 

I see where you are getting at here. There is of course a slight chance that my RF is off every so slightly - not enough to show up in real world photography but enough to lose the edge in very higly detailed flat field objects like the textured books. And if it was, then the discrepancy would be the same from shot to shot.

 

I am going to repeat the test tonight and see if focus bracketing makes a difference.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Jamie Roberts]

I see where you are getting at here. There is of course a slight chance that my RF is off every so slightly - not enough to show up in real world photography but enough to lose the edge in very higly detailed flat field objects like the textured books. And if it was, then the discrepancy would be the same from shot to shot.

 

I am going to repeat the test tonight and see if focus bracketing makes a difference.

 

Can you please try a bit faster shutter too? Surely you can add some decent light to forgo such slow shutter times? How are you releasing the M9?

 

I can't speak for the 35 Summicron, but, when focused properly, my 35 Summilux ASPH at f2 is noticeably sharper at all distances, corner and center, than my EF 35 1.4L