Monthly Archives: Januar 2013
It took Canon more than 20 years before they started revamping their line of non-L primes. As of January 2013 the EF 24mm f/2.8, the EF 28mm f/2.8 and the EF 35mm f/2, all designs from the early 1990s, have been replaced by new models which, apart from new optics , now include Image Stabilization, USM focusing and an improved build quality.
Before I start writing about the 35 mm IS let me quickly explain for what I use prime lenses.
Before I bought the 35mm f/2 IS USM I owned the 28mm f/1.8 USM, a fairly old lens which I bought to shoot under extremely low lit environments, such as the inside of huts only lit by a candle or fire. Unfortunately, I found that lens always too wide for my taste and the optical performance, particularly the border and corner sharpness on full frame, was very bad until stopped down to about f/2.8.
I also tested the old 35mm f/2 but found the optical performance to be not any better than that of the 28mm/1.8 – definitely not good enough to produce images wide open capable of being blown up larger than A3 or 13×19”. You may now ask why I didn’t go for the Canon 35mm 1.4 L USM, a lens much appreciated by many photographers. The short answer to that is size and weight. I travel a lot and need a wide prime only on very few (but important) occasions. I already carry two bodies, the 24-105 L, a 85mm f/1.8, a 70-200 f/4L IS USM and on occasion a 17-40 f/4 L with me. That is already the limit of what is practical. In fact, if I can predict what I will be shooting I prefer to take only one or two lenses with me, as this allows me to interact much better with my subjects and work much longer without getting tired.
When Canon announced the 35mm f/2 IS USM as replacement for the old 35mm f/2 I was immediately intrigued. If optical performance was good, this would be the lens I had always been looking for: A fast, light weight and silent 35mm prime, with Image Stabilizer as a bonus.
The only downside was of course the price. 849 Euros for a 35mm f/2 prime seemed over the top. It can of course be expected that the price will (at least here in Germany) come down over time. I guess that in a year or so the lens will cost about 700 Euros or even below that. I still decided to buy now, because I have an important shoot coming up in February for which I expect to use a 35mm prime a lot (Details about the shoot soon here on this blog).
Now how is the performance? In short, very good with only minor quibbles.
II. Build quality
Build quality is very good, much better than the old 35mmf/2 but it does not quite feel like an L –lens. The lens is made out of plastic except for the metal mount. The lens barrel consist of of one piece of plastic and the only place where dust and water can enter the lens is the focusring, the switches and the front lens. The latter can be sealed by a UV filter, of course. The plastic of the barrel feels very nice and sturdy but also a bit “hollow”, which is the major difference in feel compared to my L-lenses. The lens is not weather sealed, meaning it does not have any rubber seals.
Focusing on my 5D Mark III is not only silent but also very fast and precise, even in low light. The lens is about as fast to focus as my 24-105L or my 17-40 L but a bit more reliable than these two lenses (even though I have not much to complain about these lenses focusing wise, as they hardly ever let me down). I can probably say that the 35mm f/2 IS USM is the most reliable focusing lens that I own. This is really big. From a reportage point of view your biggest concern is to get the image in focus in the split second you have to capture the shot. The lens silent ultrasonic motor gets the job done without any noise. Perfect.
IV. Image Stabilisation (IS)
Well, it works but not as good as advertised. The EF 35mm IS USM allows me to shoot at a shutter speed of 1/6s and get sharp images 80-90% of the time. Thus, the improvement gained from the Image Stabilizer (IS) unit is about two stops – a lot less than the four stops promised by Canon. At shutter speeds any lower than 1/6s the hit rate goes down dramatically. At 1/4s my hit rate is about 10-20%. (Without image stabilization, I can reliably hold the lens at 1/30s) Keep in mind that IS efficiency varies also with focusing distance. At infinity about 30% of my shots at 1/4s are crucially sharp, while at near macro distance (meaning 0.5m or below,) I wouldn’t drop my shutter speed below 1/15s). Still I consider the IS unit of the 35mm IS USM to be quite useful for still subjects.
If you do not require crucially sharp shots, the Image stabilizer can help you to shoot at shutter speeds down to about 1/2s and still get shots, which are adequately sharp for on-screen viewing or small prints.
For the sake of comparism: With my 70-200 f/4 L IS USM I normally get crucially sharp shots at 200mm at a shutter speed of 1/30s of a second, if I hold the camera properly with good breathing technique. That is an improvement of a little more than 3stops (I need about 1/250 to reliably get a high hit rate of sharp shots at 200mm without IS, although 1/125 may work with a bit of luck).
V. Optical performance
In summary, optical performance is very good and I would not hesitate to use the lens wide open, even if I plan to make large prints. The following evaluations were all made on full frame, namely a Canon EOS 5D Mark III.
I consider Sharpness (and bokeh) the most important characteristics of a lens. All other aberrations are today correctable in software by one click of the mouse with very little, if any, negative side effects.
The EF 35mm IS USM is a very sharp lens for a fast 35mm prime. At f/2 the lens is sharp from center to the borders. Even the corners are quite sharp at f/2 although they suffer from a slight veiling from coma aberration. (A little bit like the 50mm 1.4 but not nearly as strong and only in the far corners). I consider f/2 fully usable, even for large prints. Stopped down to f/2.8, the lens is at its optimum and does not improve much from there.
In particular the corner and border sharpness of the 35mm f/2 IS USM is much better than my 24-105L @35mm and my 17-40 f/4L @ 35mm. The 35mm IS USM is wide open already as sharp in the borders and corners as the 24-105L (@ 35mm) stopped down to f/5.6. My copy of the 17-40L never gets sharp in the far corners at 35mm and the border are only sharp from f/8. (Note that the 24-105L and the 17-40 seem be to particularly week at 35mm. Both of these lenses are sharp corner to corner at most of their zoom range but not at 35mm. According to Canon Service, both lenses are within specs. I believe that since I had two copies of the 24-105L which performed practically identical)
2. Chromatic aberration
The lens produces practically no lateral chromatic aberrations (blue/yellow or magenta /green color fringes on hard contrast edges). There is however a slight tendency for purple fringing at f/2 but not enough to cause any problems (the fringing is much lower than for example from my 85mm 1.8 wide open.
Vignetting wide open at f/2 is very strong. Probably around 2- 3 EV. You need to correct this in software if you need equally bright images from center to corner wide open. I usually find vignetting useful as a creative element. Particularly portraits improve much from vignetting as the attention of the viewer is focused on the face.
Vignetting is much better at f/2.8 and practically not relevant at f/4 and smaller apertures.
Distortion is very low with this lens. Only on very demanding shots, where you require absolutely straight lines, will you need to correct the slight barrel distortion.
The word Bokeh describes the aesthetical appearance of the out of focus parts of the image rather than the degree to which the out of focus parts are blurred While the former is an inherent characteristic of the lens, the latter can be influenced by the photographer through controlling the aperture.
The Bokeh of the 35mm IS USM is, dependent on the focus distance. If you focus on subjects closer than about one meter away, the background at infinity will be blurred very smoothly and beautifully. If you focus on subjects farer away than one meter, the out-of-focus parts of the image will become rougher and less pleasing to the eye. bokeh is however never really bad. If you print your images, you will discover anyway, that bokeh looks much better in print than on screen. So far I have not seen any lens below 50mm which really offers truly great bokeh, so I would not count this as a negative simply you won’t get anything truly better. (The bokeh of some 35mm lenses like Canons 35mm f/1.4 L USM look a bit better, but not by much.)
VI. Practical Considerations and comparism to Sigma f/1.4
Whoever intends to buy a fixed 35mm lens for a Canon camera certainly considers also the Canon EF 35mm f/1.4 L USM for about 1200 Euros and the Sigma 35mm 1.4 HSM for only 850 Euro. In particular, the Sigma seems to be a good alternative to the 35mm f/2 IS USM, as it is almost the same price. I have handled the Sigma and can say that it is an admirably well build and sharp lens with very good focusing. I decided against it however because I couldn’t see much benefit from the larger aperture and because the much greater size and weight is unacceptable to me.
f/1.4 certainly gives you a little bit better ability to blur the back- and foreground of your images. The difference to f/2 is however not great and certainly not enough to make aesthetically completely different photos.
The additional light gathering ability of the f/1.4 lens will also mean that you can shoot at half the shutter speed of the EF 35 mm f/2 IS USM. The EF 35 mm f/2 IS USM is however more handholdable on still subjects, since the IS gives you a two stop advantage. For example: If under a certain light level you can achieve only a shutter speed of 1/15s with the Sigma wide open at f/1.4 , the image will most likely be blurred due to hand shake. Under the same light conditions the 35mm IS USM would only give you 1/6 of a second as shutter speed, because it is one stop slower (f/2 vs. f/1.4) than the Sigma. However, as opposed to the Sigma, and because of the Image Stabilizer, the 35mm IS USM would give you a sharp image.
Sometimes people argue that a f/1.4 lens is preferable to a f/2 lens because it allows you to achieve faster shutter speeds and thus freeze action better. Well as far as achieving faster shutter speed goes, this is surely correct. However, this faster shutter speed will in practice most likely not help you stop action simply because you will not be able to get a moving subject at f/1.4 in focus. Most likely your subjects with a 35mm lens will be in a distance range of 1-6m from the camera. At such short distance even the very capable Af system of the 5d Mark III can hardly track even a slowly moving or fidgeting subject reliably at f/1.4 or f/2 (e.g. a person slowly walking towards the camera if the person is close enough to the camera to fill more than a half of the frame). The depth of field of a 35mm lens at f/1.4 or f/2 when focused to a distance of 1 to 6m is simply so thin, that even the slightest focusing error will render a blurred image. Only from f/2.8 and smaller depth of field will be big enough to cover the slight focus errors of the AF-System to allow good tracking of a moving target. (Tracking ability is much improved if you to track a person more than 6 or 7m away because the depth of field becomes bigger, but these are hardly the distances you find yourself shooting a moving target with a 35mm (wide angle) lens.).
Another argument one hears from time to time is that brighter lenses can focus quicker or more reliably under dark lighting conditions. Well, from a practical standpoint I can tell that I don’t really notice it. So far I have not had a single misfocussed image from the 35mm IS USM, so there is not really room for improvement.
Above I have said that the EF 35mm IS USM focusses a bit more reliably than my 24-105 L, but this covers all light conditions taken together. I have not noticed that the 35m IS USM can focus in low light conditions in which my 24-105L cannot focus. (The camera is much more important in this regard: low light focus on my 5D Mark III is worlds better than on my 5D Mark II) With regard to focusing speed, the lens is exactly as fast to focus as my 24-105L or my 17-40 L. Of course, I have not measured this, but if I do not notice it, what does it matter?
Thus, I would only recommend going for the Sigma 35mm 1.4 over the Canon 35mm f/2 IS USM if you prefer the Sigma’s sturdier feel and better build over the lighter weight of the Canon and/or if you absolutely must have the slightly thinner depth of field offered by the f/1.4 aperture.
The Canon EF 35mm IS USM is a very good lens and a joy to use. I prefer it to the 35mm f/1.4 offerings because of its much smaller size and much lower weight (Canon’s and Sigma’s 35mm 1.4 lenses weigh roughly double the weight of this 35mm f/2 IS USM 335g). In my experience, lighter build quality does not necessarily mean that a lens is less durable. For sure, a heavier metal lens like the Canon and Sigma 35mm f/1.4 lenses do feel much better in the hand. However, plastic lens barrels absorb shocks much better than metal, which tends to transmit shocks from the outside to the glass inside. I also have never had a plastic lens barrel brake on me even though I treat my gear like tools when I am out an about. Thus, since none of the 35mm alternatives is weather sealed I don’t see why any of them should be preferable in build quality over the others even though I admit that the Canon and the Sigma f/1.4 lenses surely feels much nicer in the hand.