Sony a7R vs. Leica M9 infinity test with 20+ Leica, Zeiss, Voigtlander rangefinder lenses

As a follow-up to my overwhelmingly popular October 2013 infinity test with a preproduction Sony a7 and various rangefinder lenses, here’s the infinity shootout I really wanted to do back then… between a regular production Sony a7R and Leica M9, with a similar, though not identical, collection of a couple dozen mostly rangefinder lenses… Unfortunately, this time I’m not one of the first in the world with the results! Please refer to the a7 shootout for an explanation why infinity focus is such a ‘torture test’ for these lenses and cameras.

As with the a7 test, I’m making this information and the high resolution files freely available for personal, non-commercial use only. If you find this useful, please consider making a donation through the Paypal button below, or making your next purchase through B&H using the indicated links throughout the article. Any amount is appreciated!





The test scene is basically the same view overlooking downtown Hamilton and should be sufficient to give an idea of general edge-to-edge lens performance on these cameras. My interest is the left and right edges more so than the extreme corners.

The parameters:

- Tripod mounted.

- Self-timer release (2 seconds IIRC).

- Tried to keep the Sony’s shutter speeds in the higher range due to the reports of shutter induced camera shake.

- a7R images were all manually focused using maximum focus assist zoom on the church spire near the center of the frame.

- Lenses on the M9 were set to the infinity hard stop except for 50mm and longer lenses, which were rangefinder focused because past experience indicated that the spire would not be far enough away for true infinity focus.

- a7R vignetting correction and other in-camera corrections were turned off. Not sure if these are embedded in RAW files, therefore wanted to be sure they weren’t.

- Some lenses tested on the M9 were not 6-bit coded. If correction was not applied (and therefore embedded in the DNGs), it’s sometimes indicated in the watermark.

- Two Sony-Leica M adapters were tried. One was from Novoflex (B&H) and the other was from Metabones. The Metabones adapter would not focus to infinity with the few lenses I tried. Additionally, the Leica mount connection was very, very tight. It was to the point where I was concerned about possibly damaging some of the more delicate lenses (such as the Canon LTMs). After a few lenses I gave up and only tested with the Novoflex. None of the results are with the Metabones adapter. Contrary to my experience with the a7, this same Novoflex adapter on the a7R allowed infinity focus very close to the infinity hard stop with most lenses. One concern debated in the forums is whether a significant difference between focus at the infinity hard stop and actual infinity as seen by magnified live view might adversely affect image quality from lenses with floating element systems (such as the Leica 50 Lux ASPH). I can’t really say one way or the other, but with the difference being minimal, whatever influence this had on final image quality will likely be very minor. With only one adapter, there’s no way to really know how much influence the adapter had on edge to edge sharpness inconsistencies apparent with some lenses. For a better understanding of how adapters can make a mess of things, see this great blog post by LensRentals.

- RAW files converted in Lightroom 5 with the default parameters and some tone curve, shadow, highlight and WB tweaks. WB adjustments were generally a single setting for each camera, which therefore doesn’t compensate for lens to lens tint shifts, etc. I also didn’t spend much time trying to get the two cameras to look the same. No CA or distortion correction. Looking back on these – some of the Sony files could use a bit of exposure adjustment to tame overall brightness. Again, try to ignore exposure/color/light differences. These were shot later during a winter afternoon and the creeping shadows as the sun set certainly changed the look of the scene. Testing this many lenses is time consuming and between the two cameras, the light changed a fair amount.

- Jpegs exported at 100% from LR, then watermarked in Photo Mechanic and exported at 50% to keep file sizes reasonable. Should be good enough for this purpose.

Also…

- Sony a7R files were resized to Leica M9 dimensions of 5212 pixels wide.

Why did I do this? I’m primarily interested in a direct comparison against the M9. I wanted to see if resizing the Sony files to M9 dimensions had any influence on the problem of edge smearing – i.e. if downsizing minimized the problem. It didn’t, as you’ll see. This also revealed that the a7R’s greater resolution, even when downsized, will retain better fine details and ‘truer’ detail information that is less likely to be a sensor/software aliasing aberration. This can often be seen in building details, where the M9 creates false details (or maybe it’s Lightroom’s interpretation of the M9 files).

Why did I turn off vignetting correction for the a7R but left the M9 set to read 6-bit lens codes and make automatic vignetting/color shift corrections? Because that is how I would shoot with both cameras. The M9′s system works fairly well while the a7R doesn’t have a solution for non-native lens color shift correction, therefore it would still need to be done later in software. This gives you and me a better idea of how bad the vignetting and color shift is with each lens.

Another interesting revelation from this comparison: it appears the M9′s sensor is slightly cropped (smaller) compared to the a7R’s. In every lens set, you will see that the given lens has a slightly wider angle of view with the a7R than in the corresponding M9 images. So… I haven’t done the math, but perhaps the M9 is a 1.05x crop? I don’t know if this is also the case with the new Leica M (Typ 240). In real world use, no big deal, at least for me. What I don’t know is whether this could be a function of the RAW converter cropping the M9′s files a bit, but I suspect not because it’s the same file dimensions whether I export from Lightroom or Photo Mechanic.

Different from my a7 infinity test, I’ll break this down by lens and offer a few comments about each on both cameras. Order is by focal length and lens speed. You’ll also find the Sony RX1R (B&H) camera included. Being a Photo Mechanic junkie (possibly one of the most powerful, fastest image sorting, tagging, captioning apps available), I’ve color classed the images from each lens set so that viewing by color class will put the files from each camera at the same aperture side by side for easier flipping between the two. For those using other viewing apps, you’ll have to figure out something else, like renaming the files, etc.

This comparison is by no means perfect or exhaustive. I’m certain valid criticisms will be found. That said, it gives me the basic information I wanted: whether or not the a7R causes edge smearing with a given lens for near infinity scenes and how bad that smearing is compared to how the lens performs on my reference camera – the Leica M9. Many of the poor performers here will likely be much more acceptable in applications that place less demand on edge rendering and sharpness.

Commentary below generally refers to image quality from the a7R unless otherwise stated.

As noted above, the difference in framing between the two cameras is not only due to slight repositioning errors between camera runs, but due to a slightly wider angle of view from what is likely a slightly larger sensor in the a7R.

Click on the lens name/headline link to access the full-rez files hosted on Google Drive. When the thumbnail page opens, click on one of interest and a larger preview will appear. To download the full-rez file, click on the downward pointing arrow at the bottom right of the page. If you’d like to save time and download the full set of images in a given gallery, look for the last thumbnail with the downward arrow symbol. This is a zipped file of everything in that gallery. Similar download procedure, though you’ll probably get a warning after clicking on the thumbnail stating something has gone wrong… it just means no preview is available. Ignore it and click on the arrow at the bottom right of the page.

The Lenses:

Voigtlander Ultra Wide-Heliar 12mm f5.6 Aspherical LTM

Fairly strong vignetting with color shift. Performance is surprisingly very similar to the M9 in respect to edge sharpness. This copy or the adapter might be causing softness along the left edge which clears by f/8. Better than I expected, though the lens is slow and will need vignetting and color shift correction in post. If you look at building details in the central frame area, you can see where the higher resolution of the a7R has an advantage and translates well when downsized to M9 dimensions. The M9 files have very, very strong moire/aliasing in certain building details that is rendered much more accurately by the a7R. Coded as the Leica 21/2.8 pre-ASPH for the M9.

Buy from B&H
Manufacturer’s product page

Voigtlander Super Wide-Heliar 15mm f4.5 Aspherical M

This lens is one of the most problematic for vignetting and color shift on the M9, if used uncoded. But as you can see here, even when coded, there is color shift. Very strong vignetting with color shift on the a7R too. Performance is very similar to the M9 in respect to edge sharpness, if perhaps slightly better on the a7R. Like the 12mm, better than I expected, though it too is slow and will need vignetting and color shift correction in post. If you look at building details in the central frame area, you can see where the higher resolution of the a7R has an advantage and translates well to downsizing to M9 dimensions. Though not as bad as with the 12mm, M9 files have moire/aliasing in certain building details that is rendered more accurately by the a7R. Coded as the Leica 21/2.8 pre-ASPH for the M9.

Buy from B&H
Manufacturer’s product page

Zeiss Distagon T* 18mm f4 ZM

Some color shift and wide open vignetting is stronger than when on the M9 and set as uncoded. Edge sharpness is poor compared to the M9 and might be usable at f/8 and higher, depending on the application. At f/4 the softness creeps fairly far towards the center of the frame. Central sharpness is excellent already from wide open and better than the M9. This might be due to downsizing of the files.. Better moire control than the M9. Two sets of M9 images. One is coded as the WATE at 16mm, the other is uncoded because I wasn’t sure if the lens’s owner’s choice of coding was the best. In hindsight, color shift is evident and a different lens code might have worked better. Perhaps the 21/2.8 pre-ASPH?

Buy from B&H
Manufacturer’s product page

Leica Summilux-M 21mm f1.4 ASPH.

This one is interesting. My lens might not be properly infinity optimized at the hard stop for the M9. On the a7R, a fairly large central area looks great already at f/1.4, though the edges definitely smear compared to the M9 and take until at least f/5.6 to be comparable. As I pointed out in my 21mm lens shootout, the Lux has a mid-zone drop in sharpness apparent with the M9, but interestingly, not so much an issue with the a7r, which makes a wider central area more usable with this lens. But the tradeoff is the edges, where on the M9, the wavy plane of focus recovers with quite good sharpness already from wide open. Vignetting is quite strong and remains so on the a7R all the way through f/11. There has been discussion about how much the compact hood design affects vignetting, and at least one test I’ve seen indicated that it was a factor on Leica cameras. Considering the slightly wider angle of view with the a7R, it is possibly an influence, but I suspect the majority of the vignetting is sensor related, as it is with the M9, if left uncoded. This lens has very strong CA… a design trade off, but does clean up well in post. It also tends to purple fringe in high contrast transitions areas at wider apertures. The central sharpness revealed by the a7R makes me consider this lens somewhat differently, as I never really thought it to be a strong performer for distant scenes.

Buy from B&H
Manufacturer’s product page

Zeiss Biogon T* 21mm f2.8 ZM

Some vignetting and color shift. The problem is more so poor edge sharpness performance that never catches up to how the lens performs on the M9. Might be usable at f/8 and higher and likely is better when there is greater subject/background separation that places no emphasis on edge performance. Coded as the 21/2.8 pre-ASPH for the M9.

Buy from B&H
Manufacturer’s product page

Leica Super-Elmar-M 21mm f3.4 ASPH.

Vignetting and possibly some color shift (difficult to tell from this scene). Central performance eclipses the M9 at all settings. Edges are disappointing. a7R doesn’t really come close until f/11, and who wants to buy a $3000 lens to shoot at f/11? The potential of this lens is best exploited by the M9, but the a7R reveals the lens will be able to deliver a lot more resolution with an appropriate sensor. Looking forward to this one on a 50MP+ (rangefinder lens optimized) sensor.

Buy from B&H
Manufacturer’s product page

Zeiss Biogon T* 25mm f2.8 ZM

First time I’ve used this lens. Some vignetting and color shift on the a7R. I can finally appreciate its reputation firsthand. On the M9, from f/4, it’s blazing sharp across the frame. Wow! Unfortunately, it’s the usual story on the a7R… poor edge sharpness that never matches the M9 through the tested aperture range, though from f/8 and higher, may be acceptable, depending on the application. Coded as the Leica 24/2.8 ASPH for the M9.

Buy from B&H
Manufacturer’s product page

Leica Summicron-M 28mm f2 ASPH.

28mm… a rangefinder focal length that doesn’t yet have an equivalent to the Leica Super-Elmar-M 21mm f3.4 ASPH above. But the Cron might be the closest. Vignetting is pretty strong on the a7R, but a bit more than usual with the M9 too. Continuing with the theme, edge performance is poor on the a7R but might be acceptable at f/8 and smaller. In this respect, it never catches up to the M9… where it needs until about f/4 for quite good edge performance. But centrally, like the Leica Super-Elmar-M 21mm f3.4 ASPH, the a7R shows that this lens is also capable of revealing a great deal of fine detail…

Buy from B&H
Manufacturer’s product page

Voigtlander Nokton 35mm f1.2 Aspherical VM II

Vignetting is pretty strong until around f/2.8 and never really improves much beyond f/4, but isn’t so bad. Doesn’t seem to be a color shift problem. It’s the first lens in this comparison that appears to work equally well on both cameras, with the a7R extracting a lot more central information due to its higher resolution, evident even in these downsized files. Also apparent is how critical wide open focus accuracy is, revealed in the M9′s images, looking quite soft until f/2. I don’t think it’s a problem of the camera’s sensor, rather, the inability to set the exactly correct point of focus. Compounding this is a problem with this specific copy: the focusing ring has lost its correct positioning relative to the helicoid (but not the focus cam, meaning rangefinder focus is still reliable) and seems to allow focusing past infinity with the M9…

The lens is surprisingly decent wide open at this subject distance. Central sharpness is certainly usable wide open and stopping down slightly to f/1.4 does show some improvement in contrast. Even greater improvement is achieved at f/2, probably peaking at f/4, but edge detail continues to tighten up through f/8. I think I see a mid-zone drop in sharpness with this lens that requires greater than f/4 to smooth out.

All in all, very respectable and I think I like it better on the a7R…

Coded as the 35 Lux ASPH pre-FLE for the M9.

Buy from B&H
Manufacturer’s product page

Canon 35mm f2.0 LTM

A lens from Canon’s rangefinder era, vintage is probably early 60s. Vignetting improves after f/2.8 and I can’t really see any color shift. Performance is very, very similar on both cameras, with the a7R perhaps extracting slightly more fine detail. Compared to modern glass, this lens has a smaller central sweet spot wide open and is lower in contrast with more veiling flare at wider apertures, but stopped down around f/5.6 and it’s quite good across the frame. It just needs some contrast boost in post. With specular light sources just outside the frame, it can exhibit some interesting/funky flare shapes. Coded for the M9 as one of the 35 Cron pre-ASPH flavors.

Manufacturer’s product page

Zeiss Biogon T* 35mm f2 ZM

This one was shot uncoded on the M9, which reveals similar vignetting characteristics between the two camera, with the a7R being slightly stronger. Color shift doesn’t appear to be a significant problem. Perhaps a slight shift with the Sony.

Slightly glowy/hazy wide open, this clears by f/2.8 where the central area jumps in contrast and sharpness, expanding outwards as the lens is stopped down. On the M9 most of the frame is great by f/4 though does continue to tighten up through f/11. Here the usual problem with the a7R is evident – the edges are very smeared wide open and lag the M9 through the entire range. Maybe acceptable by f/8, though f/11 is certainly better.

A great landscape consideration on a Leica M camera, but for the a7R, the Voigtlander above seems to be the better across-frame performer at wider apertures, though does have more mid-zone softness as a compromise for better edge performance with a more moderate contrast character.

Buy from B&H
Manufacturer’s product page

Sony RX1R with Zeiss Sonnar T* 35mm f2 lens

Obviously only available mated to the RX1/R cameras, this lens is a very strong performer already from wide open and outclasses all of the other 35mm lenses here for smooth sharpness across the frame. Note that it’s somewhat wider than 35mm… Would love to see this available as an M mount with similar performance! A 36MP version likely would extract even more image detail along the lines of what can be seen centrally from the Biogon above on the a7R.

Buy from B&H

Zeiss C Biogon T* 35mm f2.8 ZM

A lens with a strong reputation, it seems to fall a bit short against most of the other 35mms here in one way or another. I am suspicious whether my copy might be somewhat decentered, as edge sharpness on the left side appears to be worse than the right, particularly evident with the M9 because the a7R’s images suffer from the usual terrible edge smearing. Somewhat stronger vignetting on the a7R that never really goes away. Perhaps a consequence of the C-Biogon design? Maybe some color shift. Coded as the 35 Summarit for the M9.

Buy from B&H
Manufacturer’s product page

Sony Carl Zeiss Sonnar T* 35mm f2.8

Another lens about which I’m suspicious decentering might be a problem in the test copy because the left side seems a bit weaker… or is there a trend here and suspicion of the camera would be appropriate? Performance is decent but never really amazing, perhaps because it seems to lack a bit of bite/punch compared to some of the other Zeiss formulas. It’s more consistent than the Voigtlander across the frame, but the RX1R’s lens is still best overall. Considering its price, I’d spend the extra ~$400 for the Voigtlander’s speed, unless AF and compactness is important. If you love 35mm, then consider the RX1. :)

Buy from B&H

Voigtlander Nokton classic 40mm f1.4

Vignetting is comparable on both cameras, though a bit stronger in the a7R images, as this copy was uncoded for the M9 and I didn’t set a correction for it manually (there isn’t a 40mm option anyway).

A quirky lens not really intended for this kind of scene, but I thought I’d include it anyway. First of all, it has very strong field curvature. Secondly, I muffed focus on the M9… making the mistake to set the lens to infinity rather than checking the actual rangefinder to see if I should tweak the focus for the steeple near the center of the frame. As a result, the focus point is drastically different between the two cameras. I was initially going to discard the M9′s images, but after a second inspection, I think they have some relevance.

The a7R images are correctly focused on the steeple whereas focus with the M9 falls on the buildings in the background. At wider apertures, the central area is very much the sweet spot, with the mid-zone and edges looking terrible. There is some evidence of edge smearing in the a7R images, as the blur looks jittery compared to the M9… But the poor edge performance is mostly due to the lens’s massive field curvature. Look at the wide open M9 image, where the far buildings are in focus, the steeple in the center is very soft, then the two houses at the center bottom of the frame are reasonably sharp…

I like this lens a lot for people photos in part due to it’s somewhat softer rendering (though background bokeh can be harsh), and this kind of field curvature tends to better blur background details the image periphery. But I have not tested the lens to determine if field curvature moderates at nearer focusing distances. It’s something to keep in mind because many lenses will change behavior as the focus distance changes.

Focus shift is another problem with this lens at nearer distances, but a live view camera such as the a7R will avoid this if focus is done at the shooting aperture.

Don’t buy this lens for typical/traditional landscape photography. :)

Buy from B&H
Manufacturer’s product page

Leica Summilux-M 50mm f1.4 ASPH.

A lens many want to love on the a7/a7R series… It’s probably fine for nearer distance work where the subject is separated from the background and peripheral details don’t require perfect focus.

For this kind of landscape application, unfortunately it falls flat on the a7R compared to what it can deliver on the M9. The problem is the edges, again. While not as bad as the wider lenses, distinct detail smearing detracts from the lens’s rendering, as correctly seen in the M9 files.

Wide open edge sharpness is surprisingly decent on the M9. But, as indicated by the lens’s MTF values, it suffers from a mid-zone drop in sharpness through the wider aperture range. Whereas with the M9 this mid-zone drop just looks softer than other areas, the a7R’s images look softer and more nervous. Vertical lines have a distracting, jittery, double image effect.

Centrally, the a7R beats the M9 for absolute detail and its extra resolution helps it avoid nasty moire/aliasing such as the unfortunate vertical green line running down the church spire in the middle of some M9 frames. Creep towards the edges though, and the a7R’s images feel nervous, needing until at least f/8, or even better, f/11, to match the M9.

What makes this lens appeal to me is its versatility on the M9. I use it for everything, including landscapes and distant subjects if I keep in mind the mid-zone dip limitation. But even at such distances, it’s still very usable wide open, if needed. At nearer distances it has a fairly unique, smooth rendering that many other 50s can’t match, yet is still very sharp with excellent contrast and color. f/1.4 also adds to the versatility, being fast enough without making the lens too large. Maybe the closest competitors are the new Zeiss Otus 55/1.4 (B&H) and the 55/1.8 below? But neither work on the M9…

If shooting a lot of landscape type images with the a7R where across-frame consistency and sharpness is important to you, then consider other options. This lens likely will be a reasonable paring on the a7 series for nearer distance, three dimensional work where there is subject to background separation without the need for edge detail sharpness. But even here, it’s closely matched or beaten by the new 55/1.8 at 1/4 the price… While I am a fan of the Lux, I’m not going to deny that 2013/2014 has finally presented some worthy competitors. I’m sure Leica will respond, but won’t be eager to know the cost!

Buy from B&H
Manufacturer’s product page

Canon 50mm f1.4 LTM

A lens from the same era as the Canon 35/2 above, it shares many characteristics with it, including a lot of softness/glow/veiling flare wide open and gradually improves as it is stopped down.

This test reveals a problem with the corresponding M9 images, in that from f/2.8 on, the focus point on the spire has shifted to the background. Yet, the a7R images don’t reveal such a shift, and I did not adjust focus between each exposure. This copy has a very stiff aperture ring, which sits next to the focusing ring. It’s possible I accidentally shifted the focusing ring while stopping down during the M9 sequence (it was cold, so I was wearing thin gloves, and as the light was fading, I tried to work through the lenses faster on the M9)…

I think I slightly prefer the look on the M9, but real-world, it’s likely a minimal difference. Lens was uncoded for M9 use.

As a side note, there are a few other older fast 50mm rangefinder lenses I could have included in this test, such as the Nikkor 5cm/1.4, Leitz Summarit 5cm/1.5 and the Zeiss Opton-Sonnar 50/1.5, but experience has been that none are decent infinity landscape lenses and all are bested by the Canon. Therefore I didn’t bother.

Manufacturer’s product page

Voigtlander Nokton Aspherical 50mm f1.5 LTM

This is the LTM version of the M mount lens currently available from Voigtlander. Unfortunately Cosina, the manufacturer of the Voigtlander line, is very stingy with product information and therefore it’s difficult to pin down definitive information about the optical differences, if any, between the two models. From what I have read from various sources, the optical formula is supposed to be the same, but changes/improvements were made to the coatings that result in somewhat higher contrast and color saturation, particularly with backlighting.

Apparently it’s designed to mimic the 50 Lux pre-ASPH. Unfortunately I have very limited experience with that lens and can’t offer a subjective comparison.

Difficult to draw a conclusion here because there seems to be some side to side sharpness inconsistency on each camera, with the Sony being sharper on the right side and the Leica sharper on the left. It also reveals the challenge of accurate wide aperture rangefinder focusing on distant subjects where a minute twist of the focusing ring can cause a dramatic change in focus. Typical Voigtlander purple fringing at wider apertures…

We’ll call it a draw. The M9 should be able to achieve similar central sharpness, though the a7R will certainly have a resolution advantage (as seen again by the M9′s green line of moire/aliasing on the church spire). Uncoded on the M9.

Buy from B&H
Manufacturer’s product page

Zeiss C Sonnar T* 50mm f1.5 ZM

A situation where I messed up focus with the a7R? I can only assume it is… as the M9 looks a lot better on the central church spire. Assuming similar performance is possible with the a7R, this lens does result in some edge smearing that never clears to the same degree as on the M9. Again, probably not a lens most would buy for landscape use because of the relatively weak outer zone performance… Uncoded on the M9.

Buy from B&H
Manufacturer’s product page

Zeiss Planar T* 50mm f2 ZM

Results from both cameras are very similar. Central detail/sharpness very much the same. At the edges the M9 seems to be slightly better until around f/4. This lens has always been a solid performer on the M9 and appears to be a good option for the a7R too… however, it doesn’t quite resolve as highly as the Lux ASPH in the center of the frame and just above its price point is the 55/1.8 with AF and auto aperture… Coded as the latest 50 Cron pre-AA for the M9.

Buy from B&H
Manufacturer’s product page

Sony Carl Zeiss Sonnar T* 55mm f1.8

Mini Otus? This lens seems to be quite the amazing package. Looking at the a7R-only results from the two other 50s that exhibited decent performance: 50 Lux ASPH and ZM 50/2 Planar, it doesn’t quite match the Lux for central resolution but does not suffer from a drop in mid-zone sharpness. The ZM lags slightly in terms of resolution and edge performance, but by around f/4-5.6, it’s difficult to tell the two apart outside the center. The 55 is already performing very nicely across the frame at f/2.8 and also sharpens up a bit centrally at this point compared to wider apertures.

One concern about my results with the 55: I didn’t notice that the shutter speeds dropped into the a7R ‘danger zone’ by f/5.6, where it hit 1/250. Other testers have revealed that below 1/500 is a questionable zone and in the 1/60-1/250 range is almost certainly compromised, which could explain the somewhat mushy look at f/8-11.

But, overall excellent results already from wide open!

Buy from B&H

Voigtlander Heliar classic 75mm f1.8

A lens I don’t use much, especially for infinity scenes, but threw into the mix anyway. The ‘classic’ designation should serve as a warning that this lens is not intended to be a resolution monster. It has a rather gentle, moderate contrast rendering and is never blazingly sharp. But here it does look better on the a7R. Difficult to say if this is the norm, or the typical rangefinder problem of increasing difficulty to nail focus with longer lenses. Evidence of purple fringing common with many ‘mid-range’ Voigtlander designs. Uncoded on the M9.

Buy from B&H
Manufacturer’s product page

Zeiss Tele-Tessar T* 85mm f4 ZM

Comparable results on both cameras but the a7R’s resolution advantage better manages moire/aliasing problems. Again, rangefinder focusing accuracy will be a factor with such a lens if critical results are desired. But it looks like stopping down to around f/8 is best anyway. Uncoded on the M9.

Buy from B&H
Manufacturer’s product page

Leica Summarit-M 90mm f2.5

Very similar results but the a7R benefits from its higher resolution to better control moire/aliasing problems and live view allows precise focus placement on the church steeple. This is the only lens I focus bracketed on the M9, and it was necessary to get these results. A problem with this lens (and the ZM85 above) is the focus throw is very quick, meaning fine, precise focus adjustments are more difficult to make.

Buy from B&H
Manufacturer’s product page

Cameras:

Leica M-E at B&H (the M9′s ‘replacement’)
Leica M Typ 240 silver or black at B&H
Sony a7R at B&H
Sony a7 at B&H
Sony RX1R at B&H

Conclusions:

I’m heavily invested in rangefinder lenses. Not because I was anticipating a Sony a7/a7R type full frame camera, but rather, I use and enjoy the Leica M9. It matches well my way of working and makes for an enjoyable process, if at times more challenging. Therefore I’m not overly disappointed in the generally poor performance of most of these lenses on the a7R, because I already have a camera that is proven to work very well with them. That said, it’s always nice to have options, especially ones that cost less than Leica!

My expectation is that the conversation around these cameras will be much different in one to two years when the native lens options are much greater and have been proven to be competitive with other systems. Most exciting will probably be anything Zeiss develops as an alternative to adapting their ZM line.

For the time being, I will pass on the a7 series. One reason is I’m not looking to add another set of system lenses in addition to my Canon and Leica kits. These two are distinct enough with unique applications, yet also overlap nicely for my needs. As a result, I’m unsure how the Sony would fit in. There are areas where it could replace both systems, but others where it can’t and I would need to retain what I already use anyway (such as for sports photography and other high-paced event type work where some might be surprised to find that the Leica works quite well alongside a fast DSLR). Then there are aspects of a7 series design that don’t work well for me. The cameras are too small for my hands resulting in cramped button and dial placement. When wearing gloves I too frequently found myself accidentally turning the camera off rather than adjusting the front dial. In very cold conditions the battery just didn’t have the oomph to last very long. Yes, carrying extras is a given, as is keeping them warm inside the coat… And the shutter button placement is awkward. It should be at the top of the grip extension, rather than the top camera plate. I guess it was an aspect of the retro design? As I’ve already suggested to Sony, the 9-series camera(s) that will likely follow should be made somewhat larger to allow better grip and handling. Not DSLR huge, but just a bit larger. I would be happy if they were made to Canon 1D series build quality, but of course smaller. This process also put the whole lens adapter thing into perspective for me. I found it a bit tiring in the end, adding unknown variables into the mix that may degrade image quality while also removing some conveniences. It’s my opinion, and Sony will probably be happy to read this, that the a7 series will be better as a system with native lenses. Yes, Sony may be indirectly pushing the cameras as highly adaptable, but my feeling is the compromises, at least from the rangefinder side, are currently too significant. I might change my mind if Sony releases another version with a properly optimized sensor, or if a 3rd party offers reliable conversions. If you’re a vintage SLR lens lover, this system should still work for you, but critical lens performance will be somewhat at the mercy of adapter quality.

I’m grateful to Sony for the opportunity to test the cameras without spending my own money to come to this conclusion. And hopefully others will benefit from the above information. While it may cost Sony a few lost ‘alt’ enthusiast sales, I believe it’s in everyone’s interests to find the most harmonious solution. Sony will ultimately benefit once the system fills out and the right mix of photographers determine it’s the system that best matches their needs.

And finally, the mandatory blogger’s pitch to encourage you to kick some cash my way, in exchange for this information, by making any purchase at B&H Photo Video through the affiliate link or search box below…

Or send a donation via this PayPal button:






B&H Photo - Video - Pro Audio








Equipment mentioned in this article:

Tags: , , , , , , , , , ,

7 Responses to “Sony a7R vs. Leica M9 infinity test with 20+ Leica, Zeiss, Voigtlander rangefinder lenses”

  1. » Sony a7 torture test with Leica, Zeiss, Voigtlander rangefinder lenses - TechTalk with Ron Scheffler Says:

    [...] I was finally able to run the test I originally wanted to back in October. Please click the link to read the Sony a7R vs. Leica M9 shootout with 20+ Leica, Zeiss and Voigtlander rangefinder lenses. [...]

  2. Sony a7R vs. Leica M9 infinity test with 20+ Le... Says:

    [...] “Some lenses tested on the M9 were not 6-bit coded. If correction was not applied (and therefore embedded in the DNGs), it's sometimes indicated in the watermark. – Two Sony-Leica M adapters were tried.”  [...]

  3. riz Says:

    Hi Ron,

    I am planning on getting Sony A7 camera. I am really confused between which lens to go with: 35mm 2.8 Zeiss from Sony or Voigtlander 44mm 1.4.

    I am also planning to do some video with my A7. Do you know if I set the zoom to infinity on Voigtlander, how is the video quality compared to Zeiss infinity (can you even do it on this?).

    I know that Voigtlander is only manual focus but reviews I read so far doesn’t seem like I will have any issue with focusing manually. Since I do want to be lazy sometimes does it require a lot of attention to get the focus right on Voigtlander?

    Lastly, which is better at lower light? I really don’t want to use flash for any of my pictures.

    Thank you in advance for your help and nice review

  4. Gavin Says:

    Hi, this is a wonderful study that you’ve done. Thank you very much for the hard work. I did my own study shots on a limited set of ZM lenses a couple of weeks ago, and just didn’t have the time to finish it. I found that anything below and including 35mm smears at the edge of the frame, all the way around.

    Did you ever test the Voigtlander 21mm f/1.8 Ultron? I heard that it is one of the only ~20mm lenses that will not smear at the edges of the frame, and the color shift is also quite well managed. Unfortunately, I can’t find a lens locally (in NYC, can you believe that??) to test out. Perhaps I have to try lensrentals.com.

    Anyway, thanks again for all this. It’s just so helpful.

  5. Ron Says:

    Hi Gavin,

    Sorry for not replying sooner. I haven’t yet tried the 21/1.8, as I no longer own one. From what I’ve seen elsewhere, such as the review at 3D-Kraft, it still appears to smear the edges and needs to be stopped down a fair amount to be useable, compared to how it would perform on a digital Leica. About the only rangefinder lens I’ve seen in this focal length range to have minimal smearing problems is the Leica 16-18-21 Wide Angle Tri-Elmar (WATE), but of course, it’s over $5000 and a difficult purchase to justify for use on a Sony body when there are cheaper SLR lens options available.

  6. Gavin Says:

    Ron, I didn’t realize you had responded. I didn’t even think to check :)
    I got my hands on a WATE attached to a novoflex (which is much better quality than the metabones, IMO) for my A7r. It’s a real beauty! I’m still putting it through its paces, but here are some initial observations:
    - The WATE is incredibly compact. It could fit inside my Nikon 14-24/2.8 and have room enough to jiggle around. I would certainly carry it around a lot more at this size, but does it justify the eye-popping price?? (It’s my wife’s eyes that are popping, btw).
    - At 21mm it outperforms the 14-24 at every aperture setting. In fact, doing this quick study has made me realize that even the venerable 14-24 also smears at the outer edges. The WATE loses sharpness toward the edges, but the image integrity remains fairly well intact. No muddiness there.
    - At 16mm it’s a little weird. The Nikon is sharper in the center, but the WATE is sharper in the edges with the exception of the right side of the frame which is smeared. The left side of the frame is relatively clean, and is sharper than the Nikon. Does it make sense that the lens is smearing only on one side? I feel like you encountered this on one or more of your tests, but I can’t remember. Anyway, I’m not really sure how it makes sense that the lens would only be lopsided at a certain focal length.

  7. Ron Says:

    Hi Gavin,

    Thanks for the feedback. The WATE does seem to be one of the more promising M lenses on the a7 cameras. It seems there are also promising signs of better compatibility also with the a7S, though I have yet to try one.

    I would suspect the odd performance at 16mm, with one side good and the other side smeared, could be a problem with the adapter. Only way to rule it out would be to try another adapter, or another lens, or try your lens on a different camera, such as an M9 or M240. Unfortunately the need for an adapter adds another variable into the equation and makes it more difficult to determine where the problem lies without access to another camera, lens or adapter.

    It could be a problem with the lens too, as any UWA zoom is fairly complicated and can change performance throughout the zoom range.

Leave a Reply


Copyright © 2014 TechTalk with Ron Scheffler. All Rights Reserved.
No computers were harmed in the 0.262 seconds it took to produce this page.

Designed/Developed by Lloyd Armbrust & hot, fresh, coffee.