Matrix metering - Camera to Camera Variation

Discussion in 'Nikon DX DSLR' started by Dave Rosser, Jul 27, 2007.

  1. I have been playing with my 2 cameras, a D1H and a D40, comparing the exposure metering on them. I shot the same scene (the view I get out of my front door every time I go out :smile:) using the same lens (12-24 Nikkor) at the same focal length (24mm) and same aperture (f/8) and same speed (ISO200).
    The initial idea was to see if there was any disernable difference between prints from the 2 cameras made at my usual 8.4 x 5.6inch on A4 print size. However when I downloaded the images into Lightroom V1.1 I was surprised to see that the 2 cameras idea of correct exposure were quite different, the D1H gave an exposure of 1/750th sec at f/8 which is just what I would expect from the sunny f/16 rule, here is the image:

    82915018.

    Now I did bracket this and the shot with a third of a stop more exposure is probably technically ideal with the white walls of the far buildings just blowing but it is marginal.
    When I shot with the D40 the exposure was 1/320sec at f/8 with this result

    View attachment 108231

    It needed -2/3 to -1 stop bias to match the D1H result. One other interesting point here is that as well as quite distinct metering differences the sensors, though both nominally ISO 200, seem to have different sensitivities; the D40 seems to be 1/3 to 1/2 stops slower than the D1H :confused: I obviously have to do some more controlled tests but would anyone care to comment?
     
  2. yamo

    yamo

    Jun 28, 2007
    Santa Cruz, CA
    Dave,

    Greetings. I've experienced similar differences between the D80 and the D200. Constant adjustment on the D80 from -1/3 to -1 stop, occasionally more. What drove me nuts was the variability. One of the reasons I upgraded to the D200, where I will occasionally need to adjust exposure but most of the time no adjustment is necessary.

    I think that this is a deliberate choice by Nikon with at least the D80 to "push the exposure to the right" (as in the histogram) or "expose for shadows" to acquire greater detail in the shadows. From what I gather light sensitivity/resolution is better in the highlight end while the shadow end suffers. Perhaps, this choice is what you are seeing in the D40.

    As an exercise you might try to equalize (in terms of brightness, maybe contrast) the shadow areas in your two posted images (the doorway on the far right of the images) for instance, and see how they compare in terms of noise, detail & chromatic aberations.

    In any event, for me, the stability of the metering is more important and I found that in the D200.

    Cheers,

    -Yamo-
     
  3. fks

    fks

    Apr 30, 2005
    sf bay area
    hi dave-

    the matrix metering algorithms have changed from model to model, so you are going to see some variation.

    from a hardware POV, the D1-series has 1,005 sensors vs the 420 in the D40-series. again, you will get some exposure variation, depending on the scene that you're photographing.

    keep in mind that a ±1/3EV variation in metering is supposed to be acceptable from body to body, even within the same model type.

    FWIW, the D100 i used to have tended to meter on the underexposure side, compared to my D1.


    ricky
     
  4. cwilt

    cwilt

    Apr 24, 2005
    Denver, CO
    As a reference you could try spot metering the grass and maybe include a gray card in the shot for later review. Looks like one exposed for highlights and the other for shadows.
     
  5. Thanks for the comments and suggestions I will follow them up. By the way when I compared the prints (D1H as metered, D40 with -1stop bias so both pegged same for highlights) the D1H print was the one I prefered but this may have been something to do with the way Lightroom processed them.
     
  6. Dave, I recently bought a D40 for my wife's use, and we are finding that it consistently overexposes images just like the one you posted. We're finding that it often requires as much as -1 eV compensation, except for scenes that are totally shaded, where it seems to get it about right.

    But on the other hand it is a dandy little camera which produces excellent images. You just have to watch the histogram and compensate accordingly.

    Ron Reznick's Rule of Thumb is that green grass in bright sunlight should register at about 128 on the Nikon histogram scale (which runs from 0 to 255) for a properly exposed image. That seems to work pretty well for me.
     
  7. To expand on what I posted above, I've reproduced below the histograms from your two images as produced by Nikon Capture NX:

    [​IMG]

    The large green spike near the middle almost certainly represents the grass. If you imagine a linear scale along the horizontal axis running from 0 to 255, you can see that your D1H photo is slightly underexposed (below 128), but the D40 shot is definitely overexposed, and needs about -2/3 eV compensation.

    The blue spike on the far right represents the sky, and is definitely blown out by the D40.
     
  8. I have a theory abt the d40 / d80 (not having used either of course!)

    something like this,... the overall / matrix exposure on the d40/d80 seems to be more weighted for the subject i.e. what the focus point is over,. possible reason being that a lot of ppl are going to be pointing the camera at a person and it's that which should be properly exposed

    Sil
     
  9. Sounds like a reasonable hypothesis.
     
  10. fks

    fks

    Apr 30, 2005
    sf bay area
    hi sil-

    good observation. to quote thom hogan's D40 review:

    "As you may know from my D80 review, that camera had perhaps an over reliance upon the area underneath the active AF sensor in its matrix metering. The D40 doesn't seem to have that problem near as much, though I do find it often meters a bit "hotter" than the D200 or D2 series, and this tendency still tends to follow what's under the current AF sensor."

    http://www.bythom.com/d40review.htm

    ricky

     
  11. Nchesher

    Nchesher

    579
    Jul 7, 2006
    Lansing,MI
    I've noticed this too after getting my D200. Except my problem was underexposure. I've had EV +1/3 on my D70s for over a year now. I did it on my D200 when I first got it and was overexposing alot. I was purposefully overexposing another 1/3 stop with the meter and kept doing it. I've since been metering properly with the D200 and it's spot nearly every time.
     
  12. ah, humm you know what I might easily have read it there first ages ago :redface: :smile:

    Sil
     
  13. I have a naive, simple question....which shot looks most like the scene you saw when you stepped out the door that morning?
     
Loading...
Similar Threads Forum Date
D7000, about sunny 16 and matrix metering Nikon DX DSLR May 13, 2011
Matrix metering idiosyncracies Nikon DX DSLR Aug 6, 2009
Center-Weighted vs Matrix Metering Nikon DX DSLR Feb 5, 2009
D90 matrix metering tests . Nikon DX DSLR Oct 26, 2008
Problems w/Matrix metering Nikon DX DSLR Sep 27, 2008