Here's something interesting: http://www.dxomark.com

DxO Labs, as you may or may not know, is a company who makes some fairly advanced RAW image editing software. In order to make a RAW converter, they need to reverse-engineer the output from every camera that they will be supporting. In doing so, they have built up quite the database of sensor performance in a number of categories and now published it online for your reference.

These tests don't compare JPG performance (which is affected by camera settings) or any other aspect of the camera's abilities such as autofocus or other features, but it's still a very useful tool when trying to decide which sensor best suits your needs.

In related news, the guys on the Canon forums are all having a fit right now.

gawd, that damn forum.

One little surprise the D90 rates better than the D300, but, I was talking to someone about that at the camera store on Saturday. I'll need a back up shortly, guess I'll wait for the D400.

The d90 noise performance is a bit better than the d300 i suppose.

I don't regret buying a 5d at all. Seems like it competes favorably against almost all more modern cameras. I was comparing it mostly to the 50D and D300, and somewhat to the D90. the d700 looks like a beast though

the 40D ranked higher than the 50D...

and the damn D60 ranked better than my D80...psh...cool site though...

You shouldn't have any, it's a good camera.

Yeah, it's interesting, yet kinda....... oversimplified.

G10's low light ISO number is 157.
The D700's is 2303. So uhhh, it owns the G10, but what does that mean?
Are the cameras all compared against the highest ISO supported by that camera?
(which should be low the non-slrs)

And minefield sure does hate that site.

- Frank

briefly glancing at the comparison between the 5D's low-light ISO vs. the D300's makes me seriously doubt the validity, application, and relevance of the entire page.. although I have no trouble believing the #1 and #2 spots.