Nikon and the Dpreview jagged edge conversions of RAW

Discussion in 'Digital SLR' started by RichA, May 13, 2011.

  1. RichA

    RichA Guest

    Check out this comparator. Put the cursor on the shield area of the
    Martini bottle. Now, compare the low ISO RAWs of the D7000, 300s and
    5000 to the Olympus E-PL2 and the Canon EOS600D. Look at the edges of
    the figures in the crops. All the Nikon images have serrated edges.
    The other two cameras represent the edges as smooth. I wonder what is
    causing this?

    http://dpreview.com/previews/panasonicdmcg3/page5.asp
     
    RichA, May 13, 2011
    #1
    1. Advertisements

  2. RichA

    Bruce Guest


    It is a well known effect. It is because the Nikon DSLRs have more
    subtle (i.e.weaker) anti-aliasing filters than the other brands. The
    result is - hey presto! - they can record more detail, in this case
    the knurling around the coin.

    One of the main reasons that I like using my ancient Kodak DCS Pro 14n
    is that it has no AA filter and records *even more* detail than Nikon
    DSLRs. The other reason is its outstanding colour rendition. But
    when it comes to sharpness and detail, the Kodak from 2004 *excels*.
     
    Bruce, May 13, 2011
    #2
    1. Advertisements

  3. RichA

    SumWan Guest

    Also, what's going on with the watchface hour markers? some have slits while
    others are solid!
     
    SumWan, May 13, 2011
    #3
  4. RichA

    RichA Guest

    But, the current Olympus E-5 either has no AA filter, or a very mild
    one and yet I don't see the effect with it to the extent I do with the
    Nikons.
     
    RichA, May 13, 2011
    #4
  5. RichA

    Guest Guest

    it records *false* detail (alias artifacts), not real detail that was
    in the original subject. the 14n does not record more detail than a 24
    megapixel nikon d3x.
     
    Guest, May 13, 2011
    #5
  6. RichA

    Me Guest

    Yeah - They claim that these shots are taken from exactly the same
    tripod position with fixed studio lighting - which can't be 100% true -
    as they're going to have to adjust tripod position to allow for
    variations in actual focal length and crop factor if they're using fixed
    focal length lenses. It looks like the lighting position changes
    between tests as well.
     
    Me, May 13, 2011
    #6
  7. RichA

    Bruce Guest


    That's because E-5 has a sensor with a much higher pixel density.
    While the Olympus Zuiko Digital pro-grade lenses are extremely good,
    they aren't capable of resolving the same level of detail on a Four
    Thirds sensor that Nikkors can on APS-C or full frame.

    These Zuiko Digital lenses gave excellent results on the E-1. At that
    time, Four Thirds could compete on level terms with Nikon's 2.7 and
    5.4 MP versions of the D1 and Canon's EOS 1D. But the next generation
    of sensors from Nikon and Canon blew away Four Thirds, and that system
    has never recovered the lost ground.
     
    Bruce, May 14, 2011
    #7
  8. RichA

    RichA Guest

    Isn't the 12 megapixel E-5's pixel density the same as an APS with 18
    megapixels?
     
    RichA, May 14, 2011
    #8
  9. Including moire. False detail. Which you cannot compute away.
    If that's what you want, feel free.

    I prefer being reasonably sure that what I photograph was
    really there.

    -Wolfgang
     
    Wolfgang Weisselberg, May 16, 2011
    #9
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.