|
Further Camera & Lens Comparison
I have recently made the big step and purchased a digital SLR. While still in the process of getting used to it, I got curious about the optical performance, and how the camera and the lenses stack up against my analogue gear. To find out I did a number of test shots, both with my new lenses designed for the digital sensor and my old lenses for film photography. In the second part I compare the results of the digital SLR with the results of my film SLR.
I will not discuss the respective merits of digital and film photography since I have done that in a previous article. You can think of this article as a follow-up.
|
Note: I have updated the article with a section about the 9-18mm super wideangle zoom. You can skip forward to the new section by clicking here.
The Camera
First a few words about my new camera. I bought an almost new Olympus E-420 with two kit lenses from an internet auction for EUR 400. The kit lenses cover the focal length range from 14 to 150 mm. The equivalent range for a 35 mm film camera is 28 to 300 mm. The purchase certainly looks like good value.
Of course I had studied digital SLRs beforehand, particularly the so called entry-level segment. I had been to various photo stores and handled the cameras on display to familiarise myself with them. I do not want to discuss the merits and shortcomings of other camera models in this article since it is a very touchy subject to some people, and it would distract from the actual purpose of the article. Suffice to say that I found some popular models unpleasant to hold, whilst others seemed to be more compatible with my body. Maybe this is also the time to out myself as (a) wearing glasses and (b) being left-eyed. I mean, I still have my right eye, but I prefer to look into the view finder with my left. The reason is that my eyesight has always been better on the left, and I got into the habit from early on.
Anyway, the E-420 struck me as the model most pleasant to handle, even more so than the E-520 which is technically superior by providing image stabilisation in the body. For some reason I am not a big fan of those handle grips on the right, so the more classical shape of the E-420 fitted much better in my hand.
As it turned out, the E-420 was also the smallest and lightest digital SLR on the market. Which doesn't mean a lot, since it is still bigger than some of the film SLRs from the past. For instance, look at the family picture of my cameras. You'll see that the OM-2 and the OM-10 are smaller. The E-420 has the lowest weight, however.
| The Family
Canon Powershot G2
Olympus OM-2
Olympus E-420
Olympus OM-10
|
I also studied various review sites on the internet. Some of them raised the issue of sensor softness due to a strong anti-aliasing filter, but since I am a competent software user this is no issue for me. With regards to colour representation and lens quality the Olympus models usually score well. Note: it is one of my principles never to look just at the verdict of an internet review. I always look at the evidence provided, and then draw my own conclusion. Sometimes the conclusion is about the review itself rather than the object under review. ;-)
The Lenses
In this section I am going to present pictures I have taken under more or less identical conditions with various lenses. I have the luxury of holding the complete originals in raw format on my hard disk. Due to web space limitations I am not able to make them available to the public. I will, however, present 100% crops of the JPEGs created in-camera to illustrate my point. These crops are taken from the centre of the picture. If there is anything to add in order to convey the full picture, I will do so whenever necessary. One of the astounding results was, by the way, that in almost every case the image quality in the corners was not visibly worse than in the centre.
14 mm
This focal length is covered by the kit lens Zuiko Digital 3.5-5.6/14-42mm. The result is nice and sharp wide open, with a very slight improvement at f 5.6. At f 11 the image is getting softer, a result seen with every lens I tested. The reason is the diffraction limit. The phenomenon of diffraction is a consequence of the wave nature of light. The absolute magnitude of diffraction is determined by the wave length of light and the geometry of focal length and aperture. Whether it becomes visible depends on the sensor resolution and the magnification level, which means it kicks in earlier with smaller sensor formats.
17 mm
I have two lenses in my possession that are capable of 17 mm focal length. First I am going to show you my trusty old RMC Tokina 3.5/17mm lens for the Olympus OM system.
This lens has been in fairly high regard at the time, although it has always been known to be slightly soft wide open. However, the shocking amount of blooming at f 3.5 in the sample above is something I have never experienced when shooting film. I assume, the blooming is result of using a lens designed for film photography in a digital environment. You will see the same effect to different extent with the other OM lenses further down in this review. This finding gives some credibility to Olympus' strategy of designing a system from scratch when they entered the world of digital photography a few years back.
For comparison check out the kit lens Zuiko Digital 3.5-5.6/14-42mm at 17 mm focal length.
I think it is fair to say that the kit lens is superior right up to f 11 when the difference becomes negligible.
28 mm
For the 28 mm focal length I have two lenses at my disposal. First the OM Zuiko 3.5/28mm.
There is no sign of blooming wide open. The lens shows very good image quality overall. Compare this with the kit lens Zuiko Digital 3.5-5.6/14-42mm at 28 mm focal length.
There is not a lot of difference to spot. Depending on the f stop and depending on the image detail you study you might see a marginal advantage for one lens or the other, but I call this too close to declare a winner. One advantage of the OM Zuiko is the faster speed: f 3.5 is one stop faster than f 5.0. Actually I would not be surprised if the difference in T stop was even higher, but the lighting conditions were not controlled enough to make a statement on that.
35 mm
Again, two lenses. First the OM Zuiko 2.8/35mm.
Here you can spot some blooming wide open. There is also some light falloff in the corners (not shown here). When stopping down, the lens shows very good image quality. Now look at the kit lens Zuiko Digital 3.5-5.6/14-42mm at 35 mm focal length.
At around f 5.6 the image quality is quite good. However, the softness at f 11 is substantially higher than what we are used to, and also substantially higher than the OM Zuiko. This indicates that diffraction is not the sole cause of the observed softness at f 11.
The OM Zuiko is more than 1.5 stops faster than the kit lens, although the blooming effect will require you to stop down in most situations. Once stopped down to the same level, the OM Zuiko beats the kit lens hands down, both in the centre and in the corners (not shown here). Of course, in the old days you would have expected a prime lens to outperform a zoom lens. However, at 28 mm focal length we have already seen that this need not always be the case.
40 mm
Both kit lenses provide a focal length of 40 mm. In fact, the smaller one goes up to 42 mm which for all intents and purposes is the same as 40 mm. So let's first look at the 42 mm setting of the kit lens Zuiko Digital 3.5-5.6/14-42mm.
As with 35 mm we find a noticeable softness kicking in somewhere between f 5.6 and f 11. Compare that to the 40 mm setting of the kit lens Zuiko Digital 4-5.6/40-150mm.
Not only is the longer kit lens one stop faster at this focal length, it also shows less degradation of image quality at f 11. Apart from a certain light falloff wide open, the longer kit lens also shows better image quality in the corners (not shown here). So in the 40 mm competition, the longer kit lens is the winner.
50 mm
I have three lenses that can provide a focal length of 50 mm. First an OM Zuiko 1.8/50mm.
As it happens, I own a second OM Zuiko 1.8/50mm.
In both cases you can see a strong occurrence of blooming wide open, probably for the same reasons already discussed in the 17 mm section. Apart from that the image quality is quite good across the whole frame, and consistent between the two samples. Let's now turn towards the kit lens Zuiko Digital 4-5.6/40-150mm.
There is a certain amount of light falloff in the corners wide open (not shown here). Other than that the image quality is similar to the primes, albeit a bit less consistent towards the borders of the frame. So it has to be concluded that the 50 mm primes show better image quality once they are sufficiently stopped down to avoid the blooming.
100 mm
There are two lenses reaching 100 mm focal length. First the OM Zuiko 2.8/100mm prime.
And then there is the kit lens Zuiko Digital 4-5.6/40-150mm.
The results look remarkably similar, although at f 5.6 the prime lens has a certain advantage close to the borders (not shown here). And, of course, it is almost two stops faster.
150 mm
150 mm is the highest the longer kit lens can go.
As you can see, the image quality is not too brilliant despite the slow f stop.
500 mm
Just out of interest I finally mounted the mirror lens RMC Tokina 8/500mm and took a shot. Handling and focusing at that type of magnification is quite difficult and certainly not feasible in a dynamic situation. The image quality across the whole frame is acceptable, but as you can see, the sensor outresolves the lens. Remember, the pictures shown here are 100% crops of a 10 megapixel original.
What also needs mentioning is the very low contrast of the images, a fact already known to me from the 35 mm film shots with this lens. The sample above has been corrected in post processing using the curves tool.
Film versus Digital, Part 2
In a previous article I compared scans of a film negative with digital images created by my Canon Powershot G2. The result was a clear advantage for film as far as image quality is concerned. Of course there are other aspects which I discussed at the time. So with the new camera at my disposal I was interested in re-opening the comparison, this time adding the Olympus E-420 as a new contestant. For reasons of time and effort I didn't re-shoot and re-scan the images with all three cameras under identical conditions. I simply waited for a sunny day and took a shot with the E-420 of the same scene that I had used one-and-a-half years previously for the other two cameras. This approach may not be scientifically correct. Nonetheless, I think the results are still useful.
I would first like to show you the resulting pictures, cropped to the same frame and scaled down to very low resolution.
| Olympus OM-2
Agfa negative film
ISO 200
|
| Canon Powershot G2
ISO 50
|
| Olympus E-420
ISO 100
|
Most people will probably react positively to the colour rendering of the E-420 picture. However, I'll wager, with sufficient post processing you can transform the other pictures to have similar colours and contrast. The fact remains that at low resolution the images are pretty much equivalent in image quality.
So in order to compare the detail that can be conveyed we need to compare them at the highest resolution of the three, which happens to be the resolution of the negative scan. They were scanned at 4000 dpi which results in 25 megapixels.
So without further ado let us compare the wall section on the right.
The G2 cannot resolve the brick structure properly. 4 megapixels are simply not sufficient. Both the E-420 and the OM-2 do resolve the brick structure, but in a very different way. The digital camera seems to have enhanced the horizontal and vertical lines, whereas they seem to be overlaid by the film grain in the OM-2 shot. On the other hand the OM-2 seems to resolve the roof tiles on the right much better than the E-420.
This picture of the gate more or less confirms the observation: the G2 is out, but E-420 and OM-2 show similar detail on images that are very different in nature. The digital image looks cleaner and seems to enhance straight lines somehow. The film image shows some dithering-like effect caused by the grain that makes straight lines less easy to follow. On the other hand small details appear sharper.
In the sample above you can see the OM-2 still resolving the roof tiles in the background, whereas the digitals cannot.
The picture of the shed confirms the observation from above once more. Overall I was quite amazed to see that a camera following the Four-Thirds standard is starting to rival the image quality of 35 mm film. Given the undeniable advantages of handling digital images, even I as a late adopter have to admit the victory of digital over film. The E-420 is certainly sufficient for my uses. And as far as total image quality is concerned, I am sure you can resolve a little more with a fine-grain film. But on the other hand there are 35 mm full frame digitals available, and the prices are slowly coming down.
I still miss the look and feel of my OM-1 and OM-2, but the thought of spending long hours in a dark room with all the smelly and potentially unhealthy chemicals does nothing for me. ;-)
I hope you enjoyed this article. As always, feel free to drop me a line, or to donate camera equipment if you wish. You can contact me here.
Martin
The Zuiko Digital 4-5.6/9-18mm is a new member in Olympus' lens line-up, available since end of 2008. Due to initial user reports that were very enthusiastic about the image quality, and due to a cashback of EUR 50 offered by Olympus Europe I decided to treat myself for Christmas.
Having used the lens for a little while, and having finally been able to do a few comparison shots (the weather was not very kind recently), I feel now in a position to present my own user report. Before I begin, let us recap that Olympus offers three ranges of lenses: standard, high grade and super high grade. The 9-18mm falls in the standard range. Nonetheless, the price was EUR 470 after the cashback (unit sold for EUR 520). To put that price into perspective, it is more than I paid for the E-420 and two kit lenses. Comparing with the old days, it is roughly DM 1000. I cannot remember ever spending so much money on a lens in the film days. Obviously my expectations were high.
In order to take the suspense out of this, let me tell you up-front that my expectations have not been met. But before we go into the bad and the ugly, let us dwell on the good for a moment: the lens does extend my wideangle reach significantly, and thereby my ability of artistic expression. And it is the most cost-effective solution despite what I said about the price. Some of the results you can see in my galleries, for instance images 9 and 10 in here.
Let's start with the longest focal length, 18mm, and compare with the two lenses above at 17mm. Please note that all my samples are 100% crops as before, ie. they have been cut out of the 10 megapixel original without any further scaling or processing.
I would call the result just about acceptable. In my opinion the lens does neither reach the sharpness of the Tokina film lens (at corresponding f-stop) nor of the 14-42mm kit lens.
The performance at the border of the frame is on the same level:
Now looking at 14mm focal length:
The result is consistent across the f-stops, but in terms of contrast and sharpness it does not quite reach the 14-42mm kit lens. Also, as expected, the image is softer at f1:11 due to diffraction. Quite the contrary we see at the border of the frame:
Here the performance is quite poor wide open, better at f1:5.6 and peaking at f1:11. This behaviour is even more pronounced in the corners of the frame (not shown here).
Moving right along to 11mm focal length. Here I have no lens for comparison, but the results speak for themselves.
Again we see a mediocre level of contrast and sharpness, akin to what we've seen at 14mm. Looking at the border, a bit closer to the corners this time:
As with the 14mm focal length, we notice poor performance that improves when stopping down. The corners are particularly affected (not shown here).
At last, the 9mm setting.
The centre, if anything, appears sharper than what we've seen at the other focal lengths. But what about the borders and corners? Let's have a look.
Having whitnessed the gradual decline in image quality from 18mm over 14mm to 11mm, this is a big surprise! The optical performance is dramatically better at 9mm. The borders are on par with the centre. Only the extreme corners show some deterioration (the outer 5% of the image circle - not shown here), but nothing like we have seen at 11mm.
So what is the overall verdict? The lens is quite consistent at the extreme positions of the focal range, with a noticable weakness in between. Outside the weak areas, the performance is decent, but it will never be the sharpest tool in the box. All in all, given the extreme range, it is a good performer in the bracket of standard lenses. Only, I feel, the price is too high given the level of performance.
Another Addendum
Since I am in a criticising mood at the moment, and pixel peeping at that, why not show you this. When processing a number of shots to create a panorama, I discovered a slight difference in optical performance between the right hand side and the left hand side. These are two shots with the 14-42mm kit lens, taken at 42mm and f1:8. You can take my word for it that there is no camera shake in the rest of the frame. Judge for yourself.
Finally, I would like to make a comment after all my lens testing. I have not seen a single shot with any of my lenses that is sharp down to the pixel level. Maybe the sensor with its strong anti-aliasing filter has something to do with it, maybe my expectation is wrong, but somehow I have the feeling that I cannot reproduce the results that some internet review sites are showing.
Anyway, enough for today, cu next time.
Martin
|
Self Portrait
Olympus OM-1n
Ilford HP5
around 1981
|
