Article Index

Resolution

The Contax version of 85/1.4 Planar, of which I owed two different copies throughout the years, has never really managed to 'feed the worm' in me - the lens was always producing mediocre resolution at wider apertures. My sweet spot for fast medium telephotos is up close and personal - meaning, as close to the subject as possible, to fill up the frame and achieve the shallowest DOF. And while I don't necessarily care much about border quality in such situations, Contax did not deliver good center resolution either, which pretty much killed the joy for me. So when I finally got my hands on the new ZE version of the lens, the first thing I did was to test the new Planar at f/1.4 and f/2 around the MFD. Results were surprisingly good - the new designed seems to have improved on the optical characteristics and now delivers a fairly good performance (caveat emptor on 'good', perception of which would vary from user to user). The MTF50 chart below summarizes results captured by Imatest - here we see that the center image resolution is outstanding throughout the aperture range, while border resolution has a typical 'fast-lens' characteristic: a falloff at wider apertures, and strong performance at moderate and small apertures.

A big surprise for me was that in pure resolution terms, Planar seems to best Leica's Summilux at its widest apertures - I can't really see much difference in the center, bot Planar produces slightly sharper borders at both f/1.4 and f/2, reversing the weakness exhibited by its Contax predecessor (at least in MTF50 terms). From f/2.8, both lenses are kind of indistinguishable in resolution and we need to look at other characteristics to decide which one might be more suitable for personal use.


Canon 5DMk2 (21Mp)

 

The imaging chart crops below confirm what Imatest seems to claim - borders at f/1.4 do look noticeably soft. While this is typical for most fast f/1.4 lenses, one would hope that such an expensive lens would perform better than average. Still, results are not throw-away - while the crop is not very crisp, there is enough detail to at least see the small print on the imaging target. Better than nothing, but let's first take a look at some real life test examples.

 

Canon 5DMk2 (21Mp)

Image borders - f/1.4 vs f/8

 

If you look at the series of crops below, all done at the infinity, you will notice a pretty typical to fast lenses pattern repeating itself here too - visually, center image quality is excellent wide open or stopped down, but there is a noticeable difference in quality around corners. Dismissing vignetting, which lowers the visual quality of the crops at f/1.4 somewhat, you can notice that the resolution is indeed worse around corners. There is less definition at f/1.4, with color palette also being more subdued. The results are not what I would call 'throw-away' though - for many users this is probably going to be acceptable, it is certainly acceptable for me. The lens also shows minor field curvature, which also adds

 

Center

L.L.Corner

U.R.Corner

 

Finally. the chart below plots MTF50 results recorded by Imatest based on the focusing distance to the focusing target. Each point is an average of all measured apertures taken both in the center as well as around image corners. As you could see, the lens performs noticeably better at longer focusing distances, primarily because of better corner resolution.

 

 

Color & Rendering

Like with most fast lenses out there, Carl Zeiss Planar 85/1.4 shows some variation in behavior, which predominantly depends on the aperture used and focusing distance. At its widest apertures, f/1.4 through f/2,and particularly at closer distances, Planar renders images in a slightly subdued colors, with overall lower global contrast. This by itself should not be a major issue for the users, since global contrast is easily fixed in post-processing. However, one should be careful not to destroy detail in shadows and mid-tones - gamut compression, which would boost overall image contrast typically has nasty side-effect of also compressing tonal detail. Speaking of the tonal detail - Planar does carry a fair bit of detail in both shadows as well as midtones. This is probably helped by the fact that the lens under-exposes images by ~0.3-0.5EV pretty consistently. Not too bad though - under-exposure is

ISO 100, 1/4000, f/1.4, 85mm

quite common in many manual focus third-party lenses and can be adjusted for relatively easily (Leica's Summilux for example, also underxposes images by ~0.5EV).

On the other hand, when the lens is shot at smaller apertures and moderate to long focusing distances, its characteristics change noticeably - higher global contrast, richer colors.

Color balance out of camera (with images taken in Adobe RGB mode) is very good - it is not as warm as Canon's 85/1.2LMk2, but certainly is warmer than Distagon 21/2.8. The color gamut is tilted a little bit towards reds and greens, thus also capturing a little bit more yellow than your typical Zeiss lens (except probably Sony;s Planar. which is even warmer), which is probably a good thing for a portraiture type lens - with warmer colors your skin tones would look more natural and eye pleasing and would require less meddling during post-processing. Of course with the flexibility and power of many modern post-processing applications like Adobe Photoshop, out of camera color balance becomes less of an issue for a skilled user.

One area where Planar 85/1.4 seems to continue its long-lasting traditoon is in handling lateral CA. I am being sarcastic here obviously - the Contax version of the lens was quite bad in this department, Sony's Planar did not seem to improve much, and neither does ZE Planar 85/1.4. Imatest's lateral CA measurements are captured in a graph below and seem to indicate that the lens is prone to some noticeably fringing at wider apertures. You can see that CZ Planar 85/1.4 produces CA slightly exceeding ~1.5px at f/1.4 and ~1.3px at f/2, which would indicate that color aberration should be visible to a naked eye. Now, how much of a problem is this going to be in real life? It depends on whether you decide to pixel-peep the image at high magnification or not - for the most part CA would not be visible unless you blow up the image to 100% and start examining it inch by inch. Even then, the fringing probably would not bee too too bad and I don't believe it would cause any distracting effect or reduce overall image quality significantly.

And, as one might expect from an f/1.4 lens with no aspherical glass employed, Planar 85/1.4 exhibits a little bit of longitudinal CA in OOF areas (most notably at f/1.4 and f/2). The artifact is subtle and under most circumstances is not even visible to the naked eye, so, yea, I would not stress about it either...

 

Canon 5DMk2 (21Mp)

 

DOF & Bokeh

As I mentioned in my recent review of Sony branded Carl Zeiss Planar T* 85/1.4 ZA lens, manufacturers these days go apes about marketing the 'smooth, out of focus rendition' that their lenses offer. While such claims are probably justified for fast, f/1.4-f/2, lenses, the bandwagon reached some absurd stage these days with one manufacturer touting OOF rendition of a variable aperture f/4-f/5.6 zoom lens! You heard that right...

Regardless of any manufacturer's claims, a fast 85mm lens is an ideal choice for a portraiture type photography. Product and even landscape photography are two other potential applications for a medium telephoto lens like Planar, but because the lens lacks AF capabilities, shooting any moving target would probably be a frustrating exercise for all but the most patient. Another potential use with f/1.4 lenses was night time photography, where users had to use the largest possible aperture to capture even the faintest shred of light. But this application seems to be becoming less relevant in the days of ISO levels exceeding 125,000. And so this leads us to the ever elusive and often misunderstood bokeh - considering that any 85mm lens shot at the same focusing distance and same aperture setting would produce an identical DOF, the main differentiation between the lenses would be the character of the out of focus areas, rather than the shallowness of the DOF itself.

But before jumping into a more detailed discussion on bokeh, let's take a look at how shallow the DOF is with Planar 85/1.4. For that, take a look at the sample image below. Shot at 1m, with the lens pointed at 45 degrees to the surface, you can observe how thin DOF really is in this picture. Per online DOF calculators, DOF is only 1cm, which should give you an idea of how hard it would be achieve a precise focus with this lens af close focusing distances, even if the lens did not exhibit focus shift. On the bright side though, such a shallow DOF opens doors for some really interesting creative applications  - as you would see a bit later in the section, Planar 85/1.4 produces a superb (to my taste at least) bokeh when used under right circumstances.

 

 

ISO 400, 1/1000, f/1.4, 85mm (Canon 5DMk2)

 

The series of shots below demonstrate the DOF and quality of bokeh at the MFD of 1m and stepped away to 3m. As expected, the widest aperture gives smoother rendering, whether up close or stepped away, but the quality obviously changes quite noticeably. At close to MFD we get backgrounds that basically lack any definition to objects, with shapes rendered into big, uniform blobs of color and practically non-existent contours. At longer focusing distances we can differentiate objects, at least big ones. Notice how the lens mutes small color gradations, practically shifting the gamut towards uniform, dominating colors. If you are looking to make the foreground to stand out, f/1.4 at MFD will do the trick very nicely. At longer focusing distances the background starts to loose its smoothness as more and more detail pops into the frame and so the bokeh starts to feel jittery .

 

DOF @ 1m

ISO 100, 1/8000, f/1.4, 85mm

ISO 100, 1/1250, f/4, 85mm

ISO 100, 1/400, f/8, 85mm

DOF @ 3m

ISO 100, 1/5000, f/1.4, 85mm

ISO 100, 1/640, f/4, 85mm

ISO 100, 1/160, f/8, 85mm

 

Since I purchased Sony;s 85/1.4 Planar some 12 months ago, the lens has become one of my all time favorite portrait lenses - the lens performs flawlessly when shot at close focusing distances (a typical focusing range for me is 1m-2m). The lens renders OOF backgrounds in smooth, wide strokes, leaving only the slightest hint of detail and often rendering colors in Monet-style, pastel saturated palette. However, carrying a Sony body just to do portraits, while also dragging another body for everything else is often a hassle for me and so up until recently, I have been using Summilux-R 80/1.4 whenever I did not want carry two different bodies. In many regards, Summilux 80/1.4 has similar to Sony Planar properties - similar MFD helps render the backgrounds into a smooth mask, but unlike Sony Planar, Summilux has a little bit more neutral palette. The ZE Planar's rendering in OOF does not leave much room for improvement either though - the lens mutes contrast in OOF at its widest apertures and keeps OOF highlights smooth and evenly illuminated.

 

 

ISO 400, 1/320, f/1.4, 85mm (Canon 5DMk2)