Thorrulz wrote in post #17208003
Unless Canon starts using Sony sensors they'll never catch up. Look at the 7DII, they are already admitting that their focus as a company isn't on upgrading the sensor tech but on other features that are cheaper for them to implement in their products.
Now it's up to the consumer to decide what they would rather shoot with for their needs.
Hrmmm, I completely disagree with this.
Sony makes excellent sensors. They have really perfected the column adc tech. They did not, however, invent it. Their sensors do rock, though. 
Canon makes excellent all around sensors that are lacking in one area only. However, their design gives Canon some advantages (other than price).
However, Sony doesn't have a monopoly on getting good low ISO DR. The D7100 chip is not a Sony chip, for example. The Nikon D4 has better base ISO DR than the 1Dx (though imho the 1dx is the better cam) but uses a Nikon designed sensor with a more traditional readout, although the ADCs are on chip they are not per column.
Canon could get amazing base ISO DR without moving the ADCs on chip, as well. Proof of this is magic lantern dual ISO. However, Canon clocks the CMOS imager chip pretty fast. This has ups and downs. The design on the Sony chips lets them do the readout slower (thousands of ADCs).
One can also use digital CDS (supposedly, I could be wrong about this, but supposedly it has it) and on chip ADCs and produce crap (The Leica M--I don't care what anyone says or what DXO says, but that sensor, while having the high tech features and being done on a more modern fab, is nowhere near Sony or Canon sensors lol).
So, we'll see. Sony makes amazing sensors.
The DXO results saying the 5d3 has less DR than the 5d2 is flat out wrong though. It may technically be true, but due to pattern noise in the real world the results are way, way, WAY in the other direction. The 5d3 sensor is a huge improvement, and the 6d one is a decent step above that imho.
And, I like DXO's measurements. They are useful.
But, don't panic~.