Can DXO Mark be relied upon as being truthful and unbiased?

I suspect there's a bias toward Nikon in the test results?

Well, largely their tests *are* "truthful and unbiased."
It's just that WHAT they test isn't always an actual measure of the performance they're trying to report, or representative of the "quality" of image at the end of the process.

Like any other standardized test, manufacturers can "game" their tests by making a few tweaks that don't actually result in better images, just better scores on their tests. And it's clear some manufacturers have done just that (and not just Nikon).

They provide some data points to consider. Certainly not the only ones, and purchase decisions made based solely on their results are not very intelligent purchase decisions.

I think they are incredibly reliable.

The only bias is how some of their data is interpreted.

DxO's main function is producing software for processing of digital images.

http://aviewfinderdarkly.com.au/2011/12/01/dxo-filmpack-3-1-upgraded-digital-photo-software/

One of the factors they take into account is the performance of the lens and digital senors so they can provide the best interpretation of the raw data and process that into tiff and jpeg images for printing and viewing.

They seek to optimise their products to work well with the widest range of camera manufacturers not just one.

Why should there be a bias towards Nikon when Pentax DSLRs consistently outperform them in sensor tests for similar price and specification cameras.

http://aviewfinderdarkly.com.au/2013/01/02/pentax-k5-ii-features-new-autofocus-system/

It can't be relied upon for lots of test results.

We know that lenses may be able to resolve more detail on a different sensor and DXO mark has attempted to create a clever system that should in theory allow a person to know whether or not using a particular lens with a given digital sensor will be able to resolve all of the megapixels for the maximum sharpness up to a certain resolution so that way you can know how effective a lens would be on that camera body.

That's great but right on the "Perceptual Megapixel" page it says:

"The number of P-Mpix of a camera/lens combination is equal to the pixel count of a sensor that would give the same sharpness if tested with a perfect theoretical optics."

http://www.dxomark.com/...n-help-you

Meaning a lens that is perfect in terms of sharpness and free from defects such as softness in the corners. A lens that is perfect in terms of optical quality is said to be diffraction limited because it is perfect.

I could be wrong but am pretty sure there's not one lens available for usage with DSLR's that is diffraction limited so that is not realistic we aren't using perfect lenses therefore those measurements are not accurate.

However the system presented might be more reliable for the center of a frame perhaps.