hubsand
Offline
• • • •
Upload & Sell: Off
|
 p.10 #1 · Contax N 17-35mm Evaluation | |
Lotusm50 wrote:
But of course a beverage can get some points out of a hundred. Parker doesn't assign points, however on an all or nothing basis. In a recent wine rating, the 1995 Krug Brut Champagne gets a score of 95, and a bottle of Korbel Brut from California gets an 86. If we applied a ranking system similar to used for these lenses, the Krug would get a 100 and the Korbel a 0, and we would be no wiser as to how close in quality these to sparkling wines are.
I would also hae to say that if you have to look at 300% crops to determine if one is sharper than the other, then there is clearly no significant difference between the two images and any rating (or any all-or-nothing points awarded) based on it is a purely meaningless exercise -- and can only distort the results you are trying to obtain.
I've criticized in the past the Photodo system of creating a single number rating of the lenses it tests. However, at least the numbers that go into those single number scores actually mean something and tell you something about about the relative quality between lenses. That's something a simple scoring and tabulation of 0 or 1 can't do -- adding those 0's and 1's together can't possibly give an accurate overall picture of relative quality. There are just too many ways to produce perverse results.
...Show more →
I would submit that the evaluation of wine is considerably more controversial and subjective (and therefore less amenable to a points-based scoring system) than lenses.
Robert Parker wrote:
In terms of awarding points, my scoring system gives every wine a base of 50 points. The wine's general color and appearance merit up to 5 points. The aroma and bouquet merit up to 15 points, depending on the intensity level and dimension of the aroma and bouquet as well as the cleanliness of the wine. The flavor and finish merit up to 20 points, and again, intensity of flavor, balance, cleanliness, and depth and length on the palate are all important considerations when giving out points. Finally, the overall quality level or potential for further evolution and improvement—aging—merits up to 10 points.
Scores are important for the reader to gauge a professional critic's overall qualitative placement of a wine vis-à-vis its peer group. However, it is also vital to consider the description of the wine's style, personality, and potential. No scoring system is perfect, but a system that provides for flexibility in scores, if applied by the same taster without prejudice, can quantify different levels of wine quality and provide the reader with one professional's judgment. However, there can never be any substitute for your own palate nor any better education than tasting the wine yourself.
...Show more →
My 'system' is rather less mature than Parkers, and mainly consists of chalking up a mark for the best image. This might sound (and indeed be) crude, but with each test I'm trying to differentiate lenses: to make them behave differently in the same circumstances and expose weaknesses. What matters with this method is the size of the gap between the final numbers: the aggregate number of 'wins' and 'losses'. If there is a consistent superiority only visible at 200% (it's rare that a 300% view is helpful!), then it's fair game.
The reduction to a single number is intended to summarise baseline optical behaviour across the broadest possible range of competence: it's not necessarily a recommendation. It doesn't even make reference to build quality or convenience or handling, or even subtler characteristics of drawing styles.
In my summary, I don't dismiss or disallow the convenience of automatic control, but how many points does that fact deserve on a scale assessing image quality?
Many difficulties arise in even attempting to quantify the results: if I'd only tested for centres and corners, the Contax would have won the test outright. If I'd included the test results at all apertures at 24mm and 35mm, the Nikon would have been favoured – just as it would if I'd run a test at 28mm where the Nikon has the advantage, again. If I'd decided that flare resistance is worth more than 5 points, the Contax would have benefited. I believe – but not without grounds – that accutance, plasticity and resolution are more important than any other aspect of a lens' behaviour. Flare resistance and control of internal reflections are important, too, but in the digital environment, CA and distortion are far less meaningful: what you can fix doesn't matter as much. Colour falls somewhere in the middle: get a well exposed picture into 16-bit ProPhoto colour space and you can pretty much achieve whatever colour you want, though it helps to have a head start.
If I tested a Leica 28mm Elmarit against a 28mm Distagon and a Nikon 28mm f2.8 AIS using this method, the Leica would win. The Zeiss would probably be very slightly better centre frame f8-f11, but the Leica would be better across the frame at f2.8-f5.6. The Nikon would be better than the Zeiss from f2.8-f4, and have less distortion, but couldn't compete for resolution at f8. The Leica would rack up lots of little victories; the Nikon and Zeiss squabbling for second and third place, and the Leica would emerge with a clear lead once the points are tallied. The points would fairly reflect the reality: the Elmarit is a better lens.
And if you can't afford one, is the Zeiss better than the Nikon? According to the test, they're both second class (fair enough). Their scores would be very close . . . the best lens for you would depend on how you prioritise wide and small aperture performance, distortion, or flare resistance. The points system would give you a fair picture of the lenses' competence across the broadest possible range.
In the summary, I don't claim that the Nikon is better than the Contax at 17mm. What I say is that, for resolution, accutance and plasticity, it's impossible to pick a winner: it depends where in the frame you look. When we take into account CA and flare resistance, the Contax is better at 17mm, but by any measure 21mm is a stalemate, and from 24mm up, even the 16-35L is an improvement.
It's hard to support the idea that the Contax has 'better' colour than the Nikon. It is warmer, which which gives more weight to the many yellow and green areas in the test target (note to self: include more red and blue), but when you look at a section of the image with a different palette, we don't see the Contax producing a richer spectrum of hues than the Nikon in the same way that, for instance, the best Leica lenses do:
http://www.16-9.net/lens_tests/cz17_35/v_nikon1735/contax_nikon_17_35e.html
I'm open to reconsider the verdict of each individual comparison: each of which impacts the final score. If I've called a decision badly according to the majority of informed opinion, I'll change it. If you agree with each pair verdict, though, you inevitably support the final score as a measure of broad-range competence.
The most competent lens, on average, may not be the right one for you. The Helios 85/1.5 would score terribly by this system, but it's still a great tool for certain tasks. That's why your review of the pictures, and even my conclusion, are more important than the final number.
|
)