Review of Olympus Top Pro lenses

[brandoneu]

http://www.zs2.hu/csgzsfoto/technical/lenses/eisazuiko.jpg

Thanks to the opportunity presented by Olympus Hungary Ltd., we had a chance to test three of the „Top-Pro” range Zuiko Digital four-thirds lenses. The range as a whole, presently counting a total of 5 members, have won the EISA "European Professional Lens 2006-2007" award. It consists of the following lenses (field-of view equivalent focal lengths in 35mm terms are given in parantheses):

Zuiko Digital ED 7-14mm 1:4.0 (14-28mm)
Zuiko Digital ED 35-100mm 1:2.0 (70-200mm)
Zuiko Digital ED 90-250mm 1:2.8 (180-500mm)
Zuiko Digital ED 150mm 1:2.0 (300mm)
Zuiko Digital ED 300mm 1:2.8 (600mm)

According to the official roadmap of Olympus, a new member will arrive in 2007:

Zuiko Digital ED 14-35mm 1:2.0 (28-70mm)

It is apparent from the focal lengths that the "Top-Pro" lenses cover the 7mm-300mm focal length range (14mm to 600mm in 35mm FOV equivalent terms), and they do that offering remarkable brightness. In the case of zoom-lenses, the f2.0 maximum aperture is virtually unparalelled. Using the EC-14 teleconverter, a reach of 424mm (848mm in 35mm equiv. terms) can be obtained using the ED 300mm 1:2.8 lens. All the lenses share the same professional-grade sturdy, splash and dust proof finish, along with the incorporated ED and super ED lens elements, promising durability and outstanding color accuracy, respectively. Olympus marks the Top-Pro lenses with a noticable silver ring on the front. Our current review looks at the a 35-100mm, the 150mm and the 90-250mm lenses.

Zsombor Cseres-Gergely:
I have been using the 4/3 system as an amateur for one and a half year now, primarily to shoot people (no pun intended). I choose the system for the (to my eye) pleasing colour, the apparently outstanding optical performance of the Pro-series lenses and the professional-grade, yet affordable E-1 body (apart from a "Rebel"-ish choice, I had OM-s before, so there might be something more going on here, if you care for this detail). Introduction of the E-1 successor is scheduled for spring 2007 and before further investing into the system, I just wanted to know how much potential is there in the top of the range 4/3 lenses - being an indication of the overall system quality a lá F1 in the case of cars. I used my E-1 for the test. Although the 5 megapixel can not show the quality of such lenses completely, it does provide a way to obtain a good estimate. On the other hand, having at least one familiar element in the too-chain helps to generate reasonable results.

A direct motivation for proceeding with the test was the fact that there are very few tests and virtually no high quality sample images for the Top-Pro lenses - probably down to the very few items sold. Here we wanted to present the capabilities of the lenses in a fairly comprehensive way, but of course we too had to face constraints. Having had only two hours to familiarise with the lenses and do the test, we could not quite get to the level of precision and thoroughness we planned to. As this bacame increasingly clear, we concentrated only on the (to us) most important features. We only hope to be able to continue or even expand the test in the future.

The large (and constant) maximum aperture is one of the distinguishing feature of the Top-Pro lenses. This makes low-ligh photography less difficult, while due to the sensor size (smallest among digital SLRs), still provides relative large depth of field. Some think that this is a drawback, rather than a blessing. I beg to differ: excluding fairly rare occasions, f2.0 itself is on the limit of usability, considering (most of) the focal lengths and the bulk of the lenses in question here. I also have to mention that I have quite bad experiences with fast (prime!) lenses that claimed a maximum aperture of f1.8 or wider, but started to produce acceptable sharpness only when stopped down to f2.8, f3.5 or even more. Ressing up images and the sensor noise built up in low-level shots makes maximum sharpness a top priority for me. I also hate to think of aperture as a quality improvement tool - it should serve creativity as a means to control DOF!

Some might ask: why do we need all this mighty expensive brightness when we have all those forms of shake reduction/image stabilisation, etc. out on the market? Within the 4/3 system, it is only Panasonic/Leica that offers IS, but this is no reason for paying thousands of dollars for an optical solution to a seemingly similar problem. It is important to note however that brightness and image stabilisation work in different situations and that one does not preclude the other. IS/AS/SR has no benefit if you are shooting (and want to freeze) moving subjects (let us not talk about the efficiency of the different available solutions). Brightness does allow freezing motion by making a slower shutter speed possible, but it brings shallow DOF with all its benefits and problems. Since the two type of solutions are not yet merged in a modern solution, we can look only at the effect of brightness here.

When testing the Top-Pro lenses we used the same criteria as looking at others, only employing the most demanding criteria realistically possible. We expected spotless build quality (not testing but trusting splash- and dust proofness based on our experience with the "Pro" line) and similarly outstanding performance from the mechanical and electro-mechanical parts, such as the zoom, focus and auto-focus mechanisms. Given the outstandingly wide maximum aperture, we expected superb performance also with this set, falling short of the available maximum only marginally. We of course expct the lenses to be sharp, not only in the center, but also in the field and at the edges. Many try to save bright and unsharp lenses by arguing that they will be employed for portraiture where "softness" is desired anyway. We have to note however that this is partly an uninteded play on words. Although softness and unsharpness can mean similar phenomena in everyday English, they are by no means the same. Softness - or soft-focus - is an effect that can be created on purpose but even that needs sharpness! Sharpness is therefore needed everywhere. The lenses are also expected to be free of geometric and chromatic aberrations, vignetting and should exhibit excellent colour fidelity. Although we did look at geometric distortions, they did not recieve a great emphasis as these are the easiest to correct during post processing. Indeed, we even expected them to be present with the zooms as they represent a relatively inexpensive degree of freedom. Colour fidelity on the other hand is important as it is quite difficult to post-process. We also expected the out of focus areas of the images (the "bokeh") to exhibit a pleasing quality, with nice gradations, full bodied colours and free of secondary chromatic aberrations. Considering the large possible apertures, this quality is a surprisingly important one in building up the final impression an image generates.

It is a major question in the case of all the tree tested lenses, whether the technically available brightness can actually be used in practical shooting situations. Considering the narrow fields of view and the bulk of the lenses, stability of the photographer and the scene shot might become a critical limiting factor. Beside these, shallow depth of field adds another level of difficulty, as any movement that happens between setting the focus and pressing the shutter can kill sharpness instantly. Previous experiences with the otherwise superb ZD 50-200mm 1:2.8-3.5 tell me that this possibility prompts extra agility and skill (at 200mm f3.5, you have about 3mm DOF there when doing a headshot).

My main interest lies in reportage- and portrait-type photography. The 35-100mm 1:2.0 is a typical reportage lens, and the 150mm 1:2.0 is a typical tool for a type of fashion photographers. Although (or rather: because?) I have heard high praises of the quality of the 150mm 1:2.0, I was more eager to meet the 35-100mm 1:2.0. Handling the 90-250mm 1:2.8 is a physical joy, but nature photography is something I mostly enjoy, rather than do.

[brandoneu]

Krisztián Nagy:
I have been using the 4/3 system since the beginning of 2005. I shoot mostly nature, macro, but I pop up on events too.
We had two 4/3 cameras available for the test shots. One of them was the Olympus E-1, the other the Olympus E-330. Since the lenses are quite heavy, and I had a portrait grip on the E-1, I chose to use that for the practical tests. For consistency, Zsombor used an E-1 too. It is true that the E-330 has higher resolution, but I felt that lack of quality had to show up already at the resolution of the E-1. We really badly wanted to get an E-400 for the review, but we could not obtain one (and could not justify to buy one solely for this purpose). However, it seems that one of us will get hold of the E-400 (or its successor) and further tests will use that. In any case, we used the E-330 to shoot the test images.

We conducted two trials: one simulating a real-life situation, the second geared toward the perception of pure technical qualities. For the second one, I created a custom test chart with repetitive geometric patterns. The chart is computer-generated and printed on a laser printer. Considering its size (A3) and the resolution (600dpi), it poses a challenge for 20 megapixel camers too.

All shots were taken at ISO100 or ISO125. We shot RAW files and applied the following procedure for development. In the case of "real life" shots, we developed the ORF-s in Olympus Studio using "Advanced High Function" to get maximum colour performance (no low light situations here). The results were saved as 16 bit TIFF-s. Sometimes we had to apply exposure compensation. Then we went to our friend who has Photoshop and created an image we are pleased with as if we were working on our own images. We have done some "levels" adjusment and corrected saturation if needed. Finally we either resized the images or took a 100% crop. In the first case, we applied a "Sharpen edges" before resizing (using bicubic sharper), and an USM with parameters A:0.4 R:40% T:0 after. In the second, we applied USM with parameters A:0.4 R:60% T:0. The images were saved with "Save for web", given its superior JPG engine (hence the lack of EXIF data). In the case of synthetic shots, we used ACE to develop the files. Although we do not particularly like the colours it gives, it is free from the Moiré effect Olympus Studio exhibits in some cases. Sharpening was set to zero here.

As already mentioned, we had a total of two hours to test the three lenses, so we would like to apologise if the test is imperfect here and there. Criticism can be raise especially in the case of the synthetic test shots of the geometric pattern, where lack of time resulted in blurred, lost and hence missing frames as well.

[brandoneu]

Zuiko Digital ED 35-100mm 1:2.0

http://www.zs2.hu/csgzsfoto/technical/lenses/small/35100_100mm_f2p5_1_small.jpg

Zsombor Cseres-Gergely:

If I was to buy a Top-Pro lens, my pick would be the 35-100 as its field of view matches my style of photography very well. This is the second mandatory reportage-lens of modern times - it offers the same field of view as the 70-200 lenses in the 35mm world (probably targeted on the same PJ audience). I know only the smaller of the 70-200 lenses made by Canon (the f4 L), but that was so well built and sharp, that it set a high standard for the 35-100. Based on the public information however, it is very difficult to judge this lens, as there were quite a number of controversial images shared on the web after its release, especially those taken with wide-open aperture. Many, if not all, of these images were clearly taken in a hurry, which does not lend a lot of credibility to the results. The blurred images might be a result of genuine unsharpness of the lens, a badly taken image (= one with avoidable camera shake) or the fact that the lens is impossible to handle and hence produces blurred images when hand-held, even if it is tack sharp on a tripod.

Grabbing the 35-100 for the first time sends a shiver down the spine. Yes, this is the same "military-grade" equipment we know from other manufacturers, if not better. Good! Handling and overal physical experience with the lens is perfect. Out of the three lenses we looked at, this has the most pleasing ergonomics. Although it weights more than one and a half kilos, the very well shaped body makes it easy to handle. Actually, because of the non-changing size, I felt it is easier to hold steady than the 50-200, its "budget" alternative!

Workings of the autofocus system is a results of both electronic and mechanical elements. We all know Olympus AF as a precise but slow one. I am not sure why it is precise, but surely slowness is at least partially caused by the lack of ultrasonic motors in the lenses. The 35-100 is not entirely different from this experience. It focuses precisely and slightly quicker than the 50-200 did. There is the usual whirring noise of the micro-motor, but focus seems to be tack on. This is a mixed bag for users of other brands, something Olympus should really improve upon (keeping the precision, of course), but also something you can learn to live with, especially using the focus limiter. A somewhat strange experience is that the E-1 - 35-100 was not able to lock focus in some backlit scenes. This is a replicable experience, and being so, it would be nice to know if it is a general issue, or just peculiar to this combination, or even to the sample we had. But who knows what the future brings with new bodies and possibly updated lenses?

Uhm, have I already said that the 35-100 sits perfectly in one's hand? Although my bony wrist ached by the end of the shootout, handling the lens was a real joy. An unfair comparison, but working at 35mm I could not help thinking about the tiny 35mm 1:3.5 macro! If you look at the sample images, you might be wondering why there only a few 100% crops available, with a lot more images on the end of this section. Well, this is a result of our decision to represent our hit rate fairly accurately. It is quite hard to keep the lens steady set at f2.0, but is one succeeds, the result is superb image quality. Actually we received some comments from early readers pushing us to admit that the 100% crops are actually downsampled originals! Although it is very hard to make out the difference on the real-life images, the synthetic tests make it clear, that there is a definite sharpness difference between f2.0 and narrower apertures (and the quality of the 150mm at f2.0, see later). Sharpness catches up however quite rapidly as we reach f2.8 and I do not really see much improvement from than on. Considering the performance of other zooms, I regard this as quite impressive.

Nevertheless, sharpness is not everything a lens can and has to offer. Brightness results in a bright viewfinder, which puts the E-1 in direct comparison with top full-frame cameras (although my comparison is merely the memory of my OM2n with the 50mm f1.4). The other benefit of the large aperture is shallow DOF, which in the case of the 35-100mm was perfect: creamy with saturated and true colour rendition. A DOF offered by the 4/3 system at f2.0 is clearly not up to the razor-thin one possible with a full Leica-frame camera plus f1.4 (or faster) lens combination (such as the Canon 5D and the new 50mm f1.2). However, given the aims of the system (mobility) and the difficulty focusing poses even at f2.0, I do not miss this possibility at all. My personal favourite is image number 19, which is a "miss" (the farther eye is in focus), but has a mood which I like very much and exhibits the wonderful bokeh of the lens very well.

Krisztián Nagy:
This lens is quite comparable to product of other companies with 70-200mm focal lengths and 2.8 brightness. Ergonomics is excellent. The lens' size does not change during operation as both zooming and focussing is internal. The zoom ring turns smoothly, more so than it does on the Zuiko Digital 50-200mm 1:2.8-3.5. On the one hand, this is expected as the 50-200mm extends during zooming, but non-trivial on the other as in this case the mechanism has to move a substantial bulk. Such high level of mechnical precision is quite reassuring. Focussing is of course implemented as "focus by wire" here too, meaning that there is no mechanical connection between the focus ring and the lenses. Similarly to the mice used for computers, the signals sent by an optical sensor are processed by electronics and then executed by an actuator - the experience given by the ring is determined solely by lubrication and sealing. I shoot mostly macros, and I have to tell that I find this type of focussing very comfortable. The often heard worry that this solution produces an experience very different from mechanical coupling is completely unfounded. Focussing is a one-finger job even on the biggest lenses! Materials used for building the lens are all excellent. Everything is in its place, without a hint of looseness. The lens shade supplied is quite substantial and offers an interesting solution for using a polarizing filter. There is a small cover near the front lens element which can be pulled to the side, making the rotator of the filter accessible for rotation.

Despite of the relatively large depth of field of the 4/3 system, handling this lens demands a master of the craft. Already a small movement parallel to the optical axis between setting focus and exposing the image renders the image noticably unsharp. This is of course old news for those using 70-200mm 1:2.8 or 1:4.0 lenses on APS-H / full Leica frame cameras. Consequently, in the beginning of our trials, we were able to get only few accaptable images, but this improved over time. The 3-way AF-limiter makes it less noticable that it is a micro-motor that sets the focus. I myself concentrated more on the 150mm 1:2.0 lens, but I have an interest in the 35-100mm 1:2.0 too. Using this lens professionally would clearly benefit from some means of lens stabilisation. We probably do not have to wait for this too long, as there is already a stabilised 4/3 lens available and Olympus is offering a compact camera with CCD-shift stabilisation. Which version we shall see in future E-system cameras, I do not know.

I have read about image quality in a few tests. My experience is similar to other, precisely executed tests' results. The lens offers outstanding image quality already at f2.0. There is a small drop at 100mm, but as image number 03 shows, there is no need to worry at all. Vignetting is invisible at the lower end of the focal-length range, even at the largest aperture. It is getting visible as we approach the long end, peaking at 100mm. Stopping down to f2.2, vignetting becomes unnoticable and practically disappears at f2.8! Possibly thanks to the incorporation of 4 ED and one Super ED lens elements, chromatic aberration seems to be absent. Contrast drops in heavily backlit scenes when using the lens wide-open, but this is noticable only if back-lighting is so hars that the background burns out completely. There is no visible distortion at any focal length.

Contemporary high resolution lenses are often criticised for not rendering out-of-focus elements of the scene particularly beautifully - many call this an indecent bokeh. Highlights in these cases appear as sharp polygons instead of following the shape of the actual light-patch. There is nothing to critisice on 35-100mm 1:2.0 in this respect. As it is apparent on the images, the blurred background is quite "creamy". It looks like that a good bokeh was a consideration already when designing the lens. Anoter interesting detail is that while focal-lengths are divided to sections in the case of the Zuiko Digital 14-54mm and the ED 50-200mm lenses, producing discrete jumps in the EXIF data, this lens gives a feedback with millimetre precision (an example: we show images with 50mm and 51mm focal lenghts!)

[brandoneu]

Zuiko Digital ED 90-250mm 1:2.8

http://www.zs2.hu/csgzsfoto/technical/lenses/small/90250_203mm_f2p8_1_small.jpg

Zsombor Cseres-Gergely:
THE big gun! As already mentioned, I am not one of those planning to use such a lens on a regular basis. It is too big, too heavy, and too long for me. Nevertheless, being a technophile person, my heart beats faster when I took it in my hands or merely look at its specs - especially since the synthetic tests seem to confirm what the figures and the MTF data suggest. Its build quality and usability was not distinguishable from that experienced with its siblings. Based solely on its weight however, I am sure that those buying into this piece of rig should definitely count with the cost of a high quality and heavy-duty tripod, if they do not own one yet!

Krisztián Nagy:
Lenses covering the field of view implied by the 90-250mm focal length on the 4/3 system are not frequently offered with a maximum aperture of f2.8, the only exception being probably the Sigma 120-300 F2.8 EX when used on a camera equipped with an APS-C sensor. This Zuiko Digital lens is designed however for 4/3 sensors from the ground up. The 90-250mm 1:2.8 was not designed to shoot handheld: the lens weights 3270g, which is impossible to hold without serious support. The minimum required accessory is a monopod. Shooting nature asks for a really sturdy tripod and a matching head. If this is available, the lens attaches greatfully by means of a massive tripod-collar. The zoom and focus ring have a smooth action just as the one on the 35-100mm 1:2.0. Setting focus on such a monster with a touch of one's little finger is quite a feeling! The distance limiter helped a lot in setting the correct focus also in this case.

Looking at the factory published MTF data, we can conclude that this lens is more than fit for being used with a teleconverter. Attaching the converter and setting a focal length of 250mm, we get a 354mm (708mm equivalent) 1:4.0 lens. The full sized image and also image number 62 shows what this lens is capable of wide open with a converter on. Although this image is heavily flash-lit, increasing sharpness to an unrealistic degree, I do not think that a possible minimal loss of sharpness would be a something to worry about. I did not observe any chromatic aberrations. Distortion is barely detectable at any focal length. The syncthetic test images show a minimal barrel distortion when used with the teleconverter. There is a minimal vignetting at f2.8, but that goes away when stopping down. Using the teleconverter however makes vignetting visible.

Considering all of it capabilities, I could recommend this lens to those who want to cover the whole tele-range with only one lens. Having said this, it is quite reasonabe to contemplate on buying an ED 300mm 1:2.8 and an ED 150mm 1:2.0 and using them on two bodies. Although this requires roughly 50 percent more investment, we get more brightness with the 150mm and more reach using the 300mm. For those preferring tight composition on the spot, the 90-250mm 1:2.8 seems to provide an ideal solution.