Originally posted by stevebrot Agreed ^ ^ ^
Yes, he made the mistake of mentioning the DPR test, but his methodology is clear and much higher rigor than the level of testing generally done by users on this site. My thoughts are that this is something very subtle with possible interaction with light source and sensor design and test rigs similar to his may be appropriate for this sort of work.
Steve
It's easy to mention the DPR test, while forgetting how DPR found similar issues for many other cameras.
What we need is a thorough investigation of this issue, using several types of supports, several cameras (more than one sample of each) and lenses, and checking every potential problematic shutter speed.
Unfortunately, the amount of work is staggering, and I don't think anyone (even Lens Rentals or other organizations with access to a large number of cameras and lenses) can do it.
DPR will unavoidably half-botch it because they lack the resources; and then, they might forget to talk about other cameras.