关于哪颗星上次ZT的DXO测试结果

转一下色坛器材专家pieq314同志的发言。

http://web.wenxuecity.com/BBSView.php?SubID=photography&MsgID=605171


A fundamental error (fatal error) with DXO's methodology is that they are doing pixel peeping (from glancing at what they did. Granted that I did not read their paper in great detail, so there is a 5% chance that they are not doing pixel peeping).

Assuming I did not misunderstand them (that is, they are indeed doing pixel peeping - comparing pixels among different cameras), they are just making the same mistake that claimed "the higher pixel count of 7D will produce blurry photos due to diffraction effect" which we discussed a few weeks ago.

What they should do is comparing the whole photo, not each pixel. If a camera has a low pixel count, its high ISO noise and dynamic range will show up higher for each pixel than that of a high pixel count photo. But if you downsize a high pixel count photo to lower resolution, its noise will decrease (and this decrease is not taken into account in DXO's test, thus making their conclusion irrelevant).