dilbert said:
Well then why don't you post your own equation for determining a sensor score?
I would not make one, because it is impossible to tell with one single score how good a complicated thing like a sensor is.
A star rating system is better because there you a less "accurate" rating, something like, this is among the best available, this is very good, this is okay and so on.
The other problem with DXOMark is, that it is a highly synthetic way of analysis, and rating. They do a lot of measurments, and give the result in an abstract number, and a standardizied picture you can look at, that say, this is about how it could look. Sure some measurments are important, because it takes away the tiresome task to compare dozens of pictures; but on the other hand, the human eye works different than a measuring instrument, and something that should be very good according to a measurment, can still look about the same as something that should be worse, according to the measurment. Noise is a typical example for that, an instrument or analysing tool will calculate a number for noise in a synthetic picture. But to the human eye, the noise pattern is as important as the the amount of noise, but it is very difficult to analyse and quantify the pattern. So even if DXO scores a sensor better in niose, it does not mean that the pictures of this camera will better to the human eye.
And this brings me back to the beginning. It is actually impossible to score a sensore with a single number.