Technical tests of Nikon D3 camera

The Nikon D3 camera, announced in 2007, is a landmark for Nikon. It seemed they had not only caught up with Canon but perhaps overtaken them in terms of low noise and high sensitivity of the sensor. With a shooting rate up to 11 fps and a 12 megapixel sensor, the D3 is suitable both for action photography and for the more sedate landscape or portrait work.

For the first time since I bought my Canon EOS 5D in 2005, Nikon has a product that is really exciting and tempting. I therefore hired one for a week (thanks to Hire a camera based in Crowborough, East Sussex) and did some tests.

I tested the D3 side by side with my 5D. These pages report my results. All my testing was done with 14-bit raw files, so I make no comment on jpegs produced by the D3.

My emotional attitude to this Nikon development is ambivalent. On the one hand, I am delighted that Nikon is at last fully competitive with Canon; that can only be good for photographers. On the other, having made a big investment in Canon lenses, and having owned Canons for forty years, I was hoping that Nikon had not jumped so far ahead as to put pressure on me to switch brands. Also, being very fond of my 5D, a part of me was hoping that the D3 was no better. However, I am not a brand fanatic and I report here honestly what my measurements show.

Much of the methodology for these tests is based on work by astronomer Christian Buil and space scientist Roger Clark. I am grateful to Roger Clark, Mike Unsold (creator of ImagesPlus software) and Dave Coffin (creator of dcraw software) for useful email correspondence. They are not however responsible for any mistakes I may have made.

  1. D3 summary of findings
  2. Technical comparison between Nikon D3 and Canon EOS 5D
  3. D3 signal to noise ratio at mid-grey, as a function of ISO speed
  4. D3 ISO speed linearity
  5. D3 response to light at ISO 200
  6. D3 signal-to-noise ratio at ISO 200
  7. D3 gain and full well capacity at ISO 200
  8. D3 sensitivity relative to Canon EOS 5D
  9. D3 behaviour at saturation
  10. D3 versus 5D tested by rabbits
  11. D3 gretagmacbeth ColorChecker™ analysed by Imatest (contains PNG files up to 700KB each))
  12. D3 firmware multipliers on red and blue
  13. D3 response to light at ISO 1600
  14. D3 signal-to-noise ratio at ISO 1600
  15. D3 read noise
  16. D3 versus 5D servo autofocus test

If you have trouble printing any of these pages, and find that tables are clipped or the text within them gets jumbled up, try the following:

You can contact me using the email address below (which you won't see unless you have Javascript enabled)

Peter Facey, Winchester, England
20080207 updated
20080119 originated