DXOMark vs Reality - let's test it

[Andrew Reid]

Here's an example of what I'm talking about...

http://www.dxomark.com/Cameras/Compare/Side-by-side/Nikon-D3300-versus-Canon-EOS-1Dx___928_753

Nikon D3300 ($399) vs 1D X ($6999)

Same score for both - 82.

Now find other odd results and do a test... take a real world shot with both cameras and look at colour, feel, noise, general performance, etc. Post the results on here.

I will be doing.

[Christina Ava]

nikon is the photographers camera, i wouldn't find it peculiar to match the 1dx.

 

[Guest Ebrahim Saadawi]

Your A6000 should be of higher IQ than your 5D3 and 1Dc, and the Leica S medium format and various medium format digital backs. That's why I don't look at DxO. Having a source to produce real world images to test DxO scores would be very helpful and extremely popular on the web.

[Nikkor]

If you look at the measurements they made the 1Dx wins in all of them, except dynamic range below iso 400. Nothing strange when comparing between a high end Full Frame camera and a low level entry crop camera.

[Julian]

Here's an example of what I'm talking about...

http://www.dxomark.com/Cameras/Compare/Side-by-side/Nikon-D3300-versus-Canon-EOS-1Dx___928_753

Nikon D3300 ($399) vs 1D X ($6999)

Same score for both - 82.

Now find other odd results and do a test... take a real world shot with both cameras and look at colour, feel, noise, general performance, etc. Post the results on here.

I will be doing.

​DxO doesn't test 'feel' or 'general performance'.... Also, I don't think anyone can just look at color and say: camera A is better than camera B. That just comes down to personal preference and/or image processing, not something like actual bit depth.

You can disagree with the 'total' score of DxO, but just click trough to the tab 'measurements' and compare all the details.

An article on 'how to read DxO scores' would be helpful for a lot of people though It's important to look further than the total score and decide for yourself what is important for your photography (and photography only, DxO Mark is completely irrelevant for video).

[Inazuma]

I have posted this before, but DXOMark's results for the GX7 and a6000 are not right. The a6000 is supposedly a bit better than the GX7 in all areas but when I had the two cameras to test I found that the GX7 had more latitude (same highlight clipping but much better shadows). Eg.

ISO 500, f2.8, 1/60th, pushed 3.5 stops on LR:

Sony:

GX7:

 

Note the chroma noise and discolouration of the plant and plant pot in the a6000 image.

And then ISO 3200, f3.5, 1/50th. Pushed 2.5 stops but the original raw image already shows some purple colouring in the shadows.

Sony:

GX7:

[fuzzynormal]

DXO might be a fun diversion for people that aspire to be technical photographers and/or like to pixel peep, but it's hard to see why any of their analysis should be a great value for those of us that actually go out and make videos/films.

[QuickHitRecord]

 

Here's what Sekonic's DTS software is telling me about my 5D3 (shooting a Sekonic Exposure Target II with my L-478D meter):

H264:BMDFilm (4K):BMDFilm (4K) + CineLog LUT:All three stacked:

There's more in this thread that I started on DVXuser: http://www.dvxuser.com/V6/showthread.php?331363-5D3-Dynamic-Range-Raw-and-H264

[leeys]

You can disagree with the 'total' score of DxO, but just click trough to the tab 'measurements' and compare all the details.

This. Always this. The assigned scores are absolute rubbish. You must click to the measurements tab and go through the data yourself. That way you can find out how a camera's sensor actually performs. Eg the 7DII might be poor in many areas, but in ISO 800-3200 its DR is much improved over the 7D and it actually is pretty much on par with the Sony 24MP sensor, making it perfect for its target market. Colour sensitivity is still a bit rubbish so have fun with the Canon splotchy noise though.