<div id="linkdxomark">This a comment for <a href="http://www.dxomark.com/index.php/Publications/DxOMark-Reviews/Sony-E-35mm-f1.8-and-Sony-E-10-18mm-f4-A-Very-Good-Standard-and-a-Wide-Zoom">this page on the website</a></div>I wonder if the sharpness comparison made here is fair since the NEX7 carries a 24MP sensor yet the Lumix carries a 16MP sensor, i.e. if they have the same score of 9MP does that mean the Lumix G lens is actually sharper? I know system-wise the score is objective, but I believe more credits should be given to the Lumix G lens engineers for making sharper optics.
Fair? How about the fact that the Panny 25mm costs more and lacks OSS, while Panasonic bodies don't have IBIS? Shall we test lenses while we mount them on a handheld motion simulator so more credit can be given to the Sony engineers, who were able to engineer effective optical stabilization while maintaining a high level of optical integrity?
Did it just drift into your mind to stipulate (against DXOMark precedent) that the Panny 25mm should be tested cross-brand on an Olympus E-M5 body? Well gee, the NEX has no 5-axis IBIS... I wonder if such a performance comparison would be fair in the same vein of your original demand to handicap the competition "to be fair"?
Framing a comparison in a specific way to favor one side or the other is the oldest trick in the book.
Read carefully, sir. I'm just talking about the SHARPNESS (TECHNICALLY), not about the price/performance ratio whatsoever, which is definitely another matter.
A Ferrari may cost a lot more but it can do 200MPH, while your average Toyota may do no more than 120; when driven by a certain so-so driver both may do just 60MPH in a quarter mile, but does that mean it's fair to say "the Ferrari is just about as good as a Toyota, but actually the Toyota is better since it costs just a fraction"? Hell no, sir. Faster is faster, sharper is sharper, it's just a matter of fact.
There's naught fanboyism here, I'm merely questioning the methodology DXO does the sharpness score. If DXO is going to do another test over effectiveness of image stabilization, that's yet another thing to question.