1. Reminder: Please use our affiliate links for holiday shopping!

Nikon J1 Gets DxOmarked and Compared to Micro 4/3 Cameras

Discussion in 'Micro 4/3 News and Rumors' started by Amin Sabet, Sep 29, 2011.

  1. Amin Sabet

    Amin Sabet Administrator

    Apr 10, 2009
    Boston, MA (USA)
    The 2.7X-crop sensor Nikon N1/J1 cameras have been tested by DxOmark, and I see the usual misinformation in a number of discussion forums. Let's take a look at the data:

    Here are the SNR data for the J1, E-P3, and G3:

    6195739540_e8ac26e4b5_z.

    The closed orange circles show the J1 performance up to manufacturer-specified ISO 400. As you can see, the J1 has a less-than-1 stop (3db) deficiency in SNR compared to the 4/3 sensors. That is a good performance. Based on sensor size, one would predict a greater-than-1 stop deficiency.

    Note: There are no meaningful data for the J1 sensor at ISO 800 and above, because above that ISO the manufacturer irreversibly applies NR smoothing to the RAW files as represented by the open circles. Any comparison between the J1 open circles and the 4/3 sensors is invalid.

    Here are the dynamic range (DR) data for the J1, E-P3, and G3:

    6195223611_f274dd7e14_z.

    Up until ISO 400, the J1 matches or exceeds the DR of the other cameras. This is not surprising, as sensor size is not the major determinant of DR in current cameras. Instead, sensor technology (specifically read noise) determines this, which is why some APS-C sensors (especially from Canon) don't do very well in DR testing while others (especially from Sony) best even the 35mm format sensors. The GH2 uses a sensor with better DR than the J1 as shown below:

    6195223651_da32de3915_z.

    Lastly, here are the tonal range (TR) data for the J1, E-P3, and G3:

    6195223631_06d49b1744_z.

    Again, the data for J1 ISO 800 and above files are misleading (since the files have been tampered with) and should be discarded. As you can see, there is slightly less than 1-stop disparity in TR when comparing the J1 sensor to Micro 4/3 sensors at ISO values for which meaningful data are provided. Eg, the J1 TR at ISO 200 is close to the Micro 4/3 cameras' TR at ISO 400.

    The bottom line here, which is not reflected in the DxOmark "overall" scores, is that Nikon is using a very good small sensor, particularly good at low ISO. With the exception of low ISO DR, the Micro 4/3 sensors are generally better, as one would expect, but not quite to the degree which one would predict based on sensor size.

    For those who aren't familiar, here's how DxOmark defines SNR, DR, and TR (source):

     
    • Like Like x 2
  2. Djarum

    Djarum Super Moderator

    Dec 15, 2009
    Huntsville, AL, USA
    Jason
    I was wondering when I saw this what it meant by smoothed. Why doesn't Nikon give the customer the true RAW file and let them do what they want with it?
     
  3. drizek

    drizek Mu-43 Veteran

    492
    Aug 5, 2011
    Conclusion: Oly needs a new sensor.

    The Nikon is not only giving around 1 stop better DR across the ISO range, but it is also capable of providing maximal DR at around ISO 100. Olympus doesn't offer anything other than ISO 200 and even there they are inferior.
     
  4. Ned

    Ned Mu-43 Legend

    Jul 18, 2010
    Alberta, Canada
    Because they think it'll be embarrassing. Panasonic does the same thing for their lens corrections. But like the Panasonic files, you could probably get past the Nikon tampering with RawTherapee or some other kind of DNG converter which can ignore these things.
     
  5. Hikari

    Hikari Mu-43 All-Pro

    Nov 26, 2010
    That is really quite neat. Obviously there are no miracles as the smoothing shows, but it is impressive.
     
  6. Amin Sabet

    Amin Sabet Administrator

    Apr 10, 2009
    Boston, MA (USA)
    It's only approaching 1 stop difference at base ISO. At ISO 400, it's closer to half a stop, and above ISO 400 there are no valid data for the Nikon sensor.

    As does Olympus. Olympus gives optional, additional perspective distortion correction in Viewer, but the optical distortion correction is baked in.

    This is a common trick. The Pentax K-5 does the same at high ISO. IMO, DxOmark shouldn't even report the data, because it is misleading. It's like when DPReview reports their silly DR figures for in-camera JPEGs. A little tone push here, a little NR there: voila, high DR :rolleyes:.
     
  7. Ruhayat

    Ruhayat New to Mu-43

    6
    Sep 29, 2011
    Kuala Lumpur Malaysia
    I hope Fuji's new X system camera based on the X100 will use the m4/3 format. Fuji has a very good track record of getting high dynamic range out of their sensors.

    Failing that, Olympus needs to design their own sensors and have it manufactured by a third party... it's a shame - I much prefer Olympus JPEGs than Panasonic's (which, from the LC40 and Leica D-Lux 3 I owned previously to the GF1 I have now, have plasticky looking yellows and reds).

    *Edit: Erm, I'm assuming that Olympus have been getting their m4/3 sensors from Panasonic, which is why it lags a generation behind the top-end GH sensor.
     
  8. RT_Panther

    RT_Panther Mu-43 Legend

    May 4, 2011
    Texas
    Most interesting!
    Makes you wonder what Nikon could have done if they had made the sensor larger...:wink:
     
  9. Art

    Art Mu-43 All-Pro

    Mar 13, 2011
    San Francisco, CA
    Is perspective distortion correction necessary? I didn't see such option in LR
     
  10. krugorg

    krugorg Mu-43 Hall of Famer

    Jul 18, 2011
    Minnesota USA
    I hope this holds true! :2thumbs: I have read all sorts of rumors on APS-C, Nikon mounts, etc. It would be great to have Fuji on board making a X100-type body and some cool glass.
     
  11. Amin Sabet

    Amin Sabet Administrator

    Apr 10, 2009
    Boston, MA (USA)
    I was talking about the Olympus software. Perspective distortion has nothing to do with the lens. It's a matter of preference if you want to address it. I almost never do.

    Also makes you wonder what Panasonic could do with some on-chip "smoothing". Seems like the companies which do this get a free pass from reviewers and most customers. I prefer my RAW raw.
     
  12. Promit

    Promit Mu-43 All-Pro

    Jun 6, 2011
    Baltimore, MD
    Promit Roy
    Watching the DPR forum guys argue over the results is painful. And yet I'm still reading it...

    Fact is, Nikon's choice of a smaller sensor was supposed to yield smaller lenses. It didn't. The big expectation/worry was that we were looking at a system that fulfilled the pocketability promise while still achieving near-m4/3 quality. That reality has not come to pass, and in fact Nikon 1's smallest offering is only vaguely smaller than a PM1. I've heard anecdotally that this is because Nikon refused to use software correction for lenses on their system. So now you can pick a Nikon with nearly m4/3 quality, or a PM1 at the same size with m4/3 quality. That leaves Nikon to sell the system on the other merits, or their name. The other merits don't seem to excite photographers. The name might make a dent.
     
  13. RT_Panther

    RT_Panther Mu-43 Legend

    May 4, 2011
    Texas
    Please show me where Nikon stated and/or advertised this..And exactly smaller than what.:confused:
     
  14. Promit

    Promit Mu-43 All-Pro

    Jun 6, 2011
    Baltimore, MD
    Promit Roy
    Psh, it's like you're new to the internet. No one cares what Nikon thinks! INTERNET DEMANDS SMALL.
     
  15. RT_Panther

    RT_Panther Mu-43 Legend

    May 4, 2011
    Texas
    So in other words, you can't back up your previous statement with facts :smile:
     
  16. krugorg

    krugorg Mu-43 Hall of Famer

    Jul 18, 2011
    Minnesota USA
    Not sure about public statements/facts, but why else would Nikon build with a smaller sensor if not to have a smaller system?
     
  17. Djarum

    Djarum Super Moderator

    Dec 15, 2009
    Huntsville, AL, USA
    Jason
    Whether Nikon stated it or marketed it, the only way I'd be interested in the system is if the quality was as good and size smaller. For me, I don't see any point in changing systems.
     
  18. Promit

    Promit Mu-43 All-Pro

    Jun 6, 2011
    Baltimore, MD
    Promit Roy
    I'm being flippant, of course. But I don't think anyone had their fingers crossed for the Nikon high-speed photography system.
     
  19. krugorg

    krugorg Mu-43 Hall of Famer

    Jul 18, 2011
    Minnesota USA
    Thank you for the analysis, Amin. Nice to have a balanced view on the DxO data, versus all the crazy talk on other forums. :biggrin:

    As others have said, no reason to sell our m4/3 gear just yet. I *am* looking forward to updated m4/3 sensors and hope the competition will push them to be even better on DR and high ISO.
     
    • Like Like x 1
  20. RT_Panther

    RT_Panther Mu-43 Legend

    May 4, 2011
    Texas
    This is fine since you're not making a statement you claim to be a fact without any supporting evidence :smile: