Color balance downgrades on Photozone

Looking at photozone user lens review criteria, the grading system or optical

verdict is as weighted as follows: performance (wide-open) - 20%, Performance

(stopped-down) - 40%, Distortions - 10%, Vignetting - 10%, Color Balance - 10%

(neutral is maximum), and Flare - 10%.

 

 

 

 

Color balance other than neutral is considered a negative. Why?

 

 

 

 

Is it because one would have to develop one's film or digital images in a

special manner to bring it back to neutral? This would assume neutral is the

gold standard. Many of the great lenses such as the 200mm f/1.8L or 85mm f/1.2

or 600mm f/4 are considered to produce warm colors. So having to correct color

balance (if desired) would require extra work by considering film, use of

filters, or in photoshop. Thus, non-neutral is considered a knock against the lens.

 

Just trying to understand photozone rating system. Naturally we take user

reviews with a grain of salt.

It's all totally abitrary. The scores assigned to various lens parameters is entirely up to who is designing the tests.

 

I personally don't care about color balance since it's easily corrected. I suppose in the days of shooting slides it mattered quite a bit more since in that case, what you see is what you get.

I know squat about Photozone's grading system, but at least they publish objective standards, and it appears they try to cover the territory. The result is that most criteria other than the two performance categories have only a slight impact on the overall score assigned a lens.

 

Consider the alternatives to rewarding neutral color balance: Either rewarding some sort of color bias or else not considering it at all. Some may want to take one of these alternatives, but rewarding neutral color balance has considerable face cred.

 

Another thing to consider: All the categories probably have positive correlations, which means they are not really independent even though they are measured separately. That is, a lens with good performance probably has fewer distortions, less vignetting, etc. Not always, but probably on average. In other words, it probably doesn't matter greatly whether or not color balance is included, EXCEPT when it picks out a particularly good or bad example.

 

Still, its worthwhile to consider the criteria used for lens evaluation as you have and to reflect whether you think they make sense to you in light of how you use lenses.

One strength of Photozone's lens database is that it does try to address rather more parameters - all of which you can see reported, so you are free to weight the factors for yourself - than many so-called reviews, which often tend to concentrate almost exclusively on sharpness. Colour rendition is to some extent a matter of taste unless you are doing scientific work with controlled light sources - portrait photographers often prefer a slightly warm balance, for example. However, there are definitely times when e.g. flare performance might play a rather bigger part in deciding between different lenses than slight differences in sharpness.

One thing I didn't know about Photozone's grading system was the fact that the OP was talking about the user-contributed ratings -- I thought the reference was to the reviews by the site's owner/operator.

 

The value of user-rating weighting system depends critically on the degree of correlation between the scores on the independent categories, that is, the two Performance categories, color balance, etc. There are two extreme states that may be instructive to consider.

 

If scores on the categories are perfectly correlated, that is, if you can predict perfectly how a person will grade a given lens on one category based on the score s/he gives in another category, then the category weights are irrelevant: A set of lenses will have the same relative order no matter how you weight the individual categories. It's a straightforward albeit tedious exercise in algebra to work this out from the basic formulas. Or you could cheat and use matrix algebra :-)

 

OTOH, if the categories are perfectly uncorrelated, that is, if you can do no better than random guessing in trying to predict the score in one category based on the score in another category for a given person/lens combination, then the category weighting scheme will determine the relative ranking of the lenses: Give 100% weight to one category, and you'll get one ranking of the lenses; give 100% weight to another category, and you're likely to get a different ranking.

 

In sum, the more correlated the scores in the individual categories, the less important the category weightings are.

 

If you argue that the lens raters are essentially doofusses who don't know what they're doing, for example, they buy lenses based on evaluations they read at Photozone and then rate them according those evaluations, then perfect correlation is more or less guaranteed, and the weighting scheme is irrelevant. And so, incidentally, is the marginal value of the user-ratings -- might as well read the owner/operator review and skip the dross.

 

OTOH, if you argue that the lens raters are seasoned, discriminating professionals, then the category scores are probably less correlated, and the user ratings are more valuable in providing marginally more relevant information.

 

Nevertheless, as previously argued, there probably is a fair degree of positive correlation among the category scores based on the inherent nature of the lens being rated, which would suggest that the category weightings are not too important.

 

There are many other problems with user rating schemes. For one thing, we don't know the qualifications of the raters. For another thing, cross-lens ratings aren't comparable unless exactly the identical raters are doing the ratings in all cases. There are more problems, but enough already!

 

User-contributed ratings can still be a useful adjunct to broad range of information sources. I prefer and give more weight to the ratings of the Photozone owner/operator, because he does dozens of them a year, uses well-established criteria, seems competent, and doesn't rely upon rote tests in making his final assessment of a lens' value. Similarly, reviews contributed by Bob Atkins and other professionals are well rounded and informationally intensive.

Thanks for the responses. Seems like photozone user ratings are a good start then one can go to the seasoned veterans for more detailed dissections.

 

One thing I was wondering is if there intentional shifting of color balance by lens manufacturers? Because optics is big business it would be hard to believe that color balance is just a by-product of the lens formula. However, it doesn't seem color balance can be used as a rough determinant of optical quality as the best to worst lenses seem to overlap in color balance.