>>> http://forums.dpreview.com/forums/read.asp?forum=1019&message=13523054
>> Yeah, we had a thread about that here like a billion years ago. As I
>> recall, from the description of the article, which I don't have and
>> wouldn't be able to read anyway, they were measuring something, but
>> ISO sensitivity wasn't it.
>It's interesting idea. I thought the point was to show which cameras has the
>most noise. Did it fail to do so?
I think so. Noise should be tested with detailed targets, with external
metering; not the camera's. Should be done with raw RAW data, too, IMO,
as conversion has so many variables.
JPEGs are not "the camera", unless the camera only does JPEGs.
There really isn't a monolithic noise figure, anyway. The issues are
too complex to be handled the way they usually are.

Signature
<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS@no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><