All non-product images in this review were shot with a Leica M9-P and the 50/2 APO-Summicron-M ASPH, with the exception of the one B&W image lower down the review. Clicking on an image will bring you the Flickr page where you can access larger versions.
Part one of this review took a look at Leica’s new M-Monochrom I’ve had a chance to review already earlier; however, it’s now time to take a look at the accompanying lens. Personally, I find this a much more interesting announcement than the other two items – but then again, I’m also the sort of person who has eight ways to get to 28mm (and can find a use for all of them) – and I do appreciate the gentle field curvature of a Tessar design.
Note: I’ve been informed by Leica that both camera and lens are prototypes, and there may be changes between now and the final release product.
The full name is a bit of a mouthful, so I’m going to refer to it as the 50 AA from here on. The 50 AA is a complicated design: 8 elements in 5 groups, in what appears to be a heavily modified double-Gauss design with a complex rear floating aspherical group, along the lines of that used in the recent 35/1.4 ASPH FLE. Leica claims it’s their most perfect lens ever – the MTF charts certainly support this, but we will of course have to use it in the field to see if a) these claims bear out, and more importantly b) whether we can see the difference under typical shooting conditions. Perfection does come at a price – the lens’ RRP is around 7,000 USD, which is double that of the already excellent (and actually apochromatic) 50/1.4 Summilux-M ASPH.
Leica claims it’s the most perfect 50mm they’ve ever made – given the family lineage also includes the legendary 50/1.4 ASPH and 50/0.95 ASPH, that’s a tall order. MTF charts of this lens look flat – I don’t know if these are measured or computed, but f5.6 promises some pretty incredible resolving power and fine microcontrast structures. Let’s just say expectations were very, very high as I opened the files from this beast.
Aside from the obvious optical formula changes from the previous design (which was a simple double-Gauss with no floating or aspherical elements) – the lens has a number of physical changes despite retaining approximately the same size. Firstly, I’m pleased to see the return of the focusing tab – it makes life a lot faster, especially when you’re trying to track moving objects. Better still, it incorporates the best of both worlds and has both a wide knurled ring and the tab, which is handy when shooting in portrait orientation. Secondly, the previous telescopic hood design has been modified a bit; it telescopes but also twists to prevent it from accidentally collapsing. Finally, the cap is no longer the plastic clip-on type, but a deep metal slip-cap that makes the lens look like the 50/1.4 ASPH when it’s attached. Unfortunately, the aperture ring seems to follow recent Leica designs and is far too loose to prevent accidental changes when taken out of a bag – you have to frequently check that you’re actually shooting at the desired aperture. It’s a complaint which I have of my 35/1.4 ASPH FLE, too – how hard can it be to make slightly firmer detents?
I want to make one thing clear upfront: we’re already at the bleeding edge of diminishing returns with most lenses; your shooting discipline and rangefinder calibration are going to make far more difference to the pixel-level results than switching from the 50/1.4 ASPH to the 50 AA.
Performance on the M9-P
Very nearly flawless. It’s easier to talk about what’s wrong with the lens than what’s right; let’s just call it perfection minus a little bit. What are the little bits? Well, it does flare a little with strong side lighting; curiously, I don’t see this behavior shooting into the sun or point light sources. There’s also vignetting that doesn’t go away fully until f5.6, though admittedly it’s fairly minor and mostly not noticeable in everyday shooting. I really don’t have anything else to say – it’s that good.
Sharpness is outstanding across the frame even from wide open. There’s a very, very slight visible improvement until f4, but it’s so borderline you might as well just use the aperture for depth of field control only. Once again: if your rangefinder is properly calibrated, and you’ll want it to be for use with this lens because the focus plane transition is so sharp, you will have nothing to complain about.
I didn’t see any lateral chromatic aberration in any of the 500 or so test images I shots with this lens, many of then deliberately framed and composed to provoke it. There is a tiny bit of spherochromatism (bokeh fringing), but it too is fairly minor. The apochromatic nature of the lens is actually given away more by its micro contrast structure than edge CA/ fringing; let me explain why. Non apochromatic lenses focus different wavelengths of light at slightly different distances; if you cut the sensor plane through this, you’ll land up with some wavelengths being focused and others not; this is why edges aren’t perfectly defined, and why the vast majority of lenses have trouble reproducing extremely fine detail structures. The 50 AA has no such issues, by the way. Bottom line: yes, it’s apochromatic, or as near as we can tell with current sensor resolution. What would be interesting is to see how it performs on say a Sony NEX-7 body, which is somewhat notorious for not playing nicely with Leica M glass – especially in the corners.
Bokeh is pleasing – probably best described as neutral – it doesn’t get in the way, it isn’t distracting, but it also doesn’t dominate the image in the same way as say, a Noctilux. I did see some odd internal structure in extremely bright out of focus highlights, suggesting a formed aspherical element, but only in a couple of images. Color is also neutral and matches the palette of the modern Leica ASPH gestalt. Perhaps the best way of describing this lens is transparent. You see your subject, and not the signature of the lens – which is what I personally prefer.
Interestingly, the aperture blades – much like every other Leica lens – don’t make a perfect circle at any aperture other than wide open, yet somehow the lenses always manage to produce great bokeh. The same is true of the Noctilux 0.95, too.
Performance on the M-Monochrom
It is clear that this lens is capable of resolving beyond the definition of the M9 and MM sensors; even on the MM, the lens continues to be an outstanding performer wide open, though the improvement on stopping down is slightly more obvious. It delivers extremely fine micro contrast, which makes it a great B&W tool. Macro contrast – whilst great for delivering saturated color – is perhaps a little bit too high for B&W, especially if you’re shooting scenes with high native dynamic range.
The bonus bit: A comparison between the 50 AA and Zeiss ZM 2/50 Planar
Unfortunately, I don’t have a 50/1.4 ASPH or previous 50/2 Summicron-M handy for a comparison, and the Noctilux 0.95 really has a different design objective (though admittedly their prices are uncomfortably close) – so I thought the most interesting head to head comparison would be with another highly regarded 50/2 – the Zeiss ZM Planar. I tested this lens against the older 50/2 Summicron extensively before making my purchase decision, the ZM 2/50 won both optically and financially. It’s a hair sharper and doesn’t exhibit any focus shift. The coatings allow it to be about 1/2 to 2/3 stop faster than the old Summicron for a given aperture and exposure histogram, too. I also preferred the way it rendered – there was just more ‘pop’ to the micro contrast, and (personally) preferable color.
Reference polar bears. This image shot with the Zeiss ZM 2/50 Planar. Clicking here will get you a larger version.
And this one with the Leica 50/2 APO. Clicking here will get you a larger version.
A note on testing methodology: the lens was shot on a tripod-mounted M9-P at base ISO with manual (i.e. fixed) white balance, with several shots taken for the purpose of focus bracketing. However, you can still see very slight differences in calibration have resulted in a slightly different focal plane between the two lenses. However, if you look at the fine detail structure in both focus zones, you should get a fairly accurate impression of lens performance. You’re probably wondering why I didn’t do the testing on the higher-resolution M-Monochrom; simple: how else would you know if the lens was truly apochromatic or not without using a color body? Please also go by my comments rather than the images (which are for demonstration purposes) – I’ve reviewed and examined far more full-size, uncompressed images on my PC than I can post here.
100% comparison screenshot at f2. Original (100%) size
And again at f8. Original size
Focus plane differences aside, (and this is a prime example of how rangefinder calibration can seriously affect results) there are a lot of differences – and all of them in favor of the 50 AA. Neither one quite gets the color right – the 50 AA is too cool, and the Zeiss is too warm and saturated. What I previously thought was excellent and three-dimensional micro contrast (look at Bob’s scarf) on the Zeiss frankly now looks a little, well, soft (look at the label that wraps around the edge of the cigar box). The 50 AA has better bokeh, too; there’s a hint of double-edging and more spherochromatism on the Zeiss.
Microcontrast, 100% crop at f8. Original size
100% crop of bokeh at f2. Original size
And again at f8. Original size
There are two pieces of good news, however – firstly, the differences by f8 are negligible; secondly, in actual shooting, the two lenses are a lot closer than this appears. Your focusing accuracy is going to make far more difference than the MTF charts. Secondly, both lenses are already at an excellent starting point – though the relative differences may be very noticeable here, I think perhaps comparing these to a Nikon 50/1.8 D might put things into perspective somewhat.
Remember my earlier article on T stops and f stops? The 50 AA meters the same as the older 50 Summicron, which is to say the Zeiss is 1/2 to 2/3 stop faster still. What this means in reality is that you can use the same aperture, get the same exposure histogram, but use a shutter speed that’s 50% to 75% faster on the Zeiss. It matters because you’re effectively getting more light into the camera, which can be critical especially in marginal situations.
The 50 AA is undoubtedly the best 50mm lens I’ve ever tested by a small, but very clear margin. I’d love to have something this good to make the most of the D800E’s resolution, but the only thing that might come close is the Zeiss 2/50 Makro-Planar or Coastal Optics 60/4 APO UV-VIS-IR. The former has a very, very short focus throw from about 1m to infinity, which makes it tricky to use as an everyday lens; the latter is limited to f4. The only problem I have with the 50 AA is the price – I don’t know how this can be priced at the same level as much more complicated optical designs like the 21/1.4 Summilux ASPH; one can only hope it’s the degree of quality control involved (and I’ve definitely seen clear sample variation amongst even the 50/1.4 ASPH and 35/1.4 ASPH FLE lenses.). It’s also clear that the resolving power of this lens goes far beyond the current M body sensors, and perhaps reading a little deeper, says something about the resolution and pixel density of future M-mount cameras. Having said all of that – I’ve been hugely impressed with the results I’ve seen out of this lens. If you have the money, and must have the best, then this is the 50mm for you.
Come back in two days for the final part of the Leica May 10 series: how much resolution does the M-Monochrom really have?
Visit our Teaching Store to up your photographic game – including Photoshop Workflow DVDs and customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!
Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved