Can somebody explain why the T5i rates 5 pts worse than the T2i in DxoMark?

I think the later rebels have a weaker anti aliasing filter. This results in a sharper image sooc, but also more perceptible noise.

996gt2 wrote in post #16349327

http://www.dxomark.com …3)/645%7C0/(bra​nd3)/Canon


Stuff like this really makes me question how valid DxoMark's testing is. These 3 cameras have essentially the same sensor and similar image processors. So why the difference? I mean, yeah, Canon hasn't really improved their APS-C sensors since the 18MP 7D came out, but going backwards...?

As you know DxoMark only evaluates the sensor performance in low light and nothing else. I don't see anything bizarre here. I think you answered your own question here. The 550D has a slight edge (0,2 f-stops) in low noise, higher ISO performance and slightly better color depth, leading to 5 points better overall score. Infact I was looking at some RAW files comparing 550D, 700D and 7D and to my untrained eye 550D had the cleanest RAW-files with high ISO-values.

But you should note that these are such small differences a normal user won't see the difference. The 700D has made some other significant updates to the body which I find more valuable, such as more cross-type focus points and better video perfomance etc.

However, like you said it does raise a question regarding Canons' sensor development as things have been pretty much stagnated for last few years, whereas in Nikon's side you see a clear pattern of increased sensor performance both in low light ISO perfomance and dynamic range...

The later cameras started to have more aggressive in camera nr settings, perhaps that is also part of the issue.

Can someone explain why 5 points in the Dxo score matters to anyone?

996gt2 wrote in post #16349327
Stuff like this really makes me question how valid DxoMark's testing is. These 3 cameras have essentially the same sensor and similar image processors. So why the difference? I mean, yeah, Canon hasn't really improved their APS-C sensors since the 18MP 7D came out, but going backwards...?

You can't take one small factor of a cameras overall ability and assume because that is slightly reduced the overall performance is worse. Maybe Canon made a big improvement elsewhere and that meant a (very slight) drop in performance in the narrow test that you are viewing. Sometimes a manufacturer will trade off a (very) slight loss somewhere for a big gain elsewhere. I highly doubt the overall image quality coming from the newer sensor has gone backwards.

rral22 wrote in post #16349770
Can someone explain why 5 points in the Dxo score matters to anyone?

I am totally aware that a 5 pt difference is unlikely to be perceptible; I just find it strange that 3 cameras with the same sensor would test differently at all, and variances like this makes me question the validity of their results for others cameras. DxO is supposed to be testing the RAW files so there shouldn't be differences in in-camera NR affecting results here

996gt2 wrote in post #16350445
I am totally aware that a 5 pt difference is unlikely to be perceptible; I just find it strange that 3 cameras with the same sensor would test differently at all, and variances like this makes me question the validity of their results for others cameras. DxO is supposed to be testing the RAW files so there shouldn't be differences in in-camera NR affecting results here

But, the electronics are not the same. Filters may not be the same.

Canon needs to upgrade their sensor, finally.