Quote:
Originally Posted by eloisecox12
I think that an evaluation of noise is both reasonable and appropriate. My concern is that dpreview sets both too high and too narrow of a threshold for what is "acceptable" noise in an image. It's like saying only a Lamborghini has good acceleration or handling; it's neither reasonable nor meaningful to compare all cars in terms of a Lamborghini's performance.
|
That may be true, but the assumption on the part of most reviewers who write for the sort of audience that DPReview do, or work at that level of technical detail, is that the readership are experienced or intelligent enough to evaluate the data, assimilate the results and draw their own conclusions.
Another tricky thing for reviewers is that they have to compare against the best in class. This means that you're forever saying that something is not quite as good as X, where X is your benchmark.
DxO (as H has already linked to) offer more information and don't make a call about quality or output, they simply present empirical data. Is that more meaningful than a review? Maybe. Maybe not.