Attention *ALL* owners of Ixus 500

Discussion in 'Digital Photography' started by Guest, Jul 1, 2004.

  1. Guest

    Guest Guest

    I have reported previously a fault in the metering of this camera. Please
    check your camera so that we can assess whether all cameras are affected or
    only some. The evaluative and centre-weighted metering are transposed. Could
    all owners please check their Ixus 500 and report whether or not it has this

    This is how to check:
    Correct exposures would be: spot> (more than) centre-weighted>evaluative.
    However, the faulty Ixus 500 gives spot>"evaluative">"centre-weighted." If
    not sure you can read exposures from
    downloaded photos into ZoomBrowser EX. (ie spot > [lighter/more
    exposure]than "evaluative etc

    Or: light coloured oblong surrounded by darker area should be
    spot<centre-weighted<evaluative [spot setting darkest - evaluative
    brightest] but "centre-weighted" is brightest - thus wrongly labelled. >
    means more than; < means less than.

    This is important as the automatic mode will give wrongly exposed photos.
    Canon will correct the fault if there is enough evidence that it is present
    in all cameras. PLEASE CHECK

    Guest, Jul 1, 2004
    1. Advertisements

  2. Guest

    Jack Guest

    Can't canon check on the cameras it has?

    Jack, Jul 1, 2004
    1. Advertisements

  3. Guest

    Basiltoo Guest

    It is quite impossible to do a useful test unless one has much more data
    about the number and size of metering zones and their weighting in the
    evaluative mode relative to the centre weighted average mode. I would say
    that it is quite possible under some conditions for the evaluative metering
    to appear to be more centre weighted that the c/w average. Just accept
    that Canon have probably got it right and get enough experience of the
    camera to know how to meter to produce the sort of pictures you like.
    Basiltoo, Jul 1, 2004
  4. This is nonsense. Using spot meter mode will give you *more* exposure
    than center-weighted mode if the area you measure in spot mode is darker
    than average for the scene, but *less* exposure if the area is lighter
    than average. You can't say anything about the relative exposure
    without considering the image content.

    If you want to test spot metering mode, put a grey card on a white
    background, meter the grey card only, and see if the card ends up
    approximately mid-grey in the captured image. Change to a black
    background and see if the exposure remains the same (but the spot
    metering area must be in the grey card only, with no background

    To test center-weighted mode, use the same setup, but now the overall
    exposure should be affected by the black/white background. Grey on
    black should get more exposure than grey on white, and the spot-metered
    grey card should be between these two values.

    Evaluative metering is even more complex. The camera meters a number of
    points in the image, decides "what kind" of image it's looking at, and
    guesses an exposure from that. Since the method isn't documented,
    there's no objective test for whether it's working right other than "do
    the pictures look ok?".
    Again, this depends on image content, and whether the spot meter was
    measuring the "light oblong" only.
    I suggest that you are just leaping to conclusions without doing any
    sort of controlled tests, based on assumptions about the behaviour of
    the metering modes that just aren't true. What does your camera do when
    shooting the grey/white/black test subjects described above?

    I have a S410, which is almost the same camera. I just shot a series of
    3 photos of a greyish bag sitting on a darker grey carpet in the
    hallway. The camera exposed all three at f/2.8 (wide open) with
    different shutter speeds:

    center-weighted 1/6
    spot 1/4
    evaluative 1/5

    Looking at the image, the differences in exposure are pretty much
    exactly as I'd expect. The spot meter mode rendered the portion of the
    bag that I metered just about mid-grey in the digital image; the average
    green-channel pixel value in the metered area is 127.4. That's as close
    to "mid grey" as I'd expect any camera to get. The mean green pixel
    value over the whole image is 133.8.

    The center-weighted mode "saw" the white wall nearby and decreased the
    exposure 2/3 stop. This gave an average green-channel pixel value of
    114.8 measured over the whole image.

    Evaluative metering gave an exposure halfway between these two
    extremes, giving a mean green pixel value of 124.0. Subjectively, it
    looks the best of the 3 exposures.

    In other words, all of the exposures seem reasonable given what the
    meter is supposed to be doing in each mode. For this subject,
    my camera gives spot > evaluative > center-weighted, which is correct
    for this subject.

    How do you conclude that your 500 is faulty, given the same ordering of
    meter results, without taking into account the subject matter? Why
    don't you post some examples of your badly-exposed images where others
    can look at them?

    Dave Martindale, Jul 1, 2004
  5. Guest

    Peter Guest

    Apologies: my post omitted vital information and was unclear. Thanks to
    those who replied: sorry for wasting time. I will repost with clear and
    complete information.

    Peter, Jul 2, 2004
    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.