M is for Magenta

Discussion in 'Digital Photography' started by Scott W, Nov 10, 2006.

  1. Scott W

    Scott W Guest

    Over the last few days it has become clear that the M8 has far too weak
    of an IR blocking filter. The problem is textiles that are dyed black
    often reflect a lot of IR, which the camera sees as both red and blue.
    The result is that many clothes that should be black photograph as

    In watching the Leica forums I find it interesting that a number of
    people are saying that this is not a problem since it is easy to change
    the colors in Photoshop. Now call me crazy but I would be more then a
    little upset if I paid $5,000 for a camera only to find out that I had
    to hand adjust colors in my photos. The problem is that you can't
    just do a global replace of black for magenta since the real world does
    have magenta in it.

    It will be interesting to see how this plays out over the next few

    Scott W, Nov 10, 2006
    1. Advertisements

  2. And it also shows nasty streaking in some conditions with bright lights
    - perhaps related?...

    But Leica's damage control has swung into action:

    Leica statement:

    Dear Friends of Leica,

    Customer feedback to us following the start of shipment of the LEICA M8
    points to a performance under certain conditions that does not meet the
    expectations in the Leica brand. Our technical and engineering teams
    have isolated the source of these concerns and have identified
    definitive solutions. Over the next two weeks we will communicate a
    plan to further assure you of your investment in the LEICA M8.

    We hope for your understanding and support. Please accept that I will
    not be able to answer all messages in the forum personally.

    Kind regards

    Leica Camera AG


    Sounds like a recall is looming, but I'll bet it is only "for those
    customers who are experiencing problems...."

    Hmm. Where's Dallas to defend his beloved?

    mark.thomas.7, Nov 10, 2006
    1. Advertisements

  3. Scott W

    nathantw Guest

    Sounds like the beta testers, ahem I mean early adopters, have done a
    great service for Leica while instead of charging the company a salary
    they're paying the company $5000 each to test their camera. Wow, isn't
    the computer age wonderful? I remember the old days when stuff was
    tested before it went out the door, now companies just let their
    customers do the testing for them.
    nathantw, Nov 10, 2006
  4. Scott W

    Mike Fields Guest

    And, after you find the bugs for them, they fix
    the software and charge you for the "upgrade"
    (which fixes many of the things that did not work
    as advertised but introduced new ones ... Acronis
    "True Image" anyone ?)

    Mike Fields, Nov 10, 2006
  5. I doesn't work that way: the problems with the M8 can't be fixed in

    When you figure out for Leica that Bayer CFA filters are transparent to IR
    (which everyoner knew all along), and you figure out for Leica that Bayer
    cameras desperately need a low-pass (AA) filter (which everyone knew all
    along; heck the Foveon fans have been banging that into our heads for years
    now (they're right about Bayer arrays needing AA filtering, they're wrong
    about Foveon or monochrome sensors not needing it)), you'll realize that you
    spent US$5000 for a camera that is physically defective and simply not

    One completely and totally buck-nekked US$5000 emperor.

    David J. Littleboy
    Tokyo, Japan
    David J. Littleboy, Nov 10, 2006
  6. Scott W

    Scott W Guest

    And now the big problem, this statement has made a lot of people who
    have ordered the M8 happy, but they have their limits. This is the
    response from one such person "That's all the M8 purchasers wanted
    to hear. Well done (as long as the answer isn't filters!)."

    And this is the rub, Leica has already said the solution was filters in
    front on the lens. Will Leica change their mind on this and re-chip
    the cameras with a sensor that has a working IR blocking filter?

    Digital cameras that are overly sensitive to IR are not new, but the
    other camera companies fixed the problems years ago. I would imagine
    that are some pretty tense phone calls between Kodak and Leica right
    about now.

    Scott W, Nov 10, 2006
  7. Scott W

    acl Guest

    Well, maybe they'll retrofit IR filters in the existing cameras
    (unlikely, I suppose) and fix upcoming models (more likely, I'd say).
    Being such a small company, and not having much beyond their reputation
    to sell their cameras (for the asking price), I'd have thought they will
    make a serious effort to please the people who already bought the
    cameras; otherwise they're in serious trouble.
    Yes, it's pretty bizarre. You'd have thought that they (Kodak and Leica)
    would have taken into account the years of experience with the IR
    problems of sensors. Not to mention the streaking business, which is
    also strange (but that may be fixable by firmware, who knows? It depends
    on what it actually is).

    But anyway, I think this is pretty funny (if you're not one of the
    sufferers, of course). Especially given some of the statements about
    this camera made here recently.
    acl, Nov 10, 2006
  8. Scott W

    SimonLW Guest

    Can't be too hard to recall the 10 cameras sold ; )
    SimonLW, Nov 10, 2006
  9. Scott W

    acl Guest

    OK, here's a thought (advance warning: I just thought of this, so take
    it with a grain of salt). According to
    the passband of these IR blocking filter depends on the angle of
    incidence of the light. If the angle of incidence is increased (0 angle
    is normal incidence), the peak wavelength is shifted to lower
    wavelengths (towards the visible). Since the distance of the lens' rear
    element from the filter/sensor is much less for the M8 than for SLRs,
    the largest possible angle of incidence will be larger than for SLRs.

    This means that they may not have been able to use a filter that cuts
    off low enough wavelengths, because then you would end up blocking some
    visible wavelengths at the edges of the frame (for some lenses).

    If that's the case, it's not very good news I'm afraid. But maybe
    someone who knows more about IR filters can chime in here and explain
    why I'm wrong.
    acl, Nov 10, 2006
  10. Scott W

    Scott W Guest

    An interesting idea. I really don't know if Leica is using a
    interference filter or a absorption filter, I kind of think maybe an
    absorption filter. The reason I am thinking this is Leica talked about
    wanting to use a thinner then normal filter, most likely because they
    have very little working distance on their lenses. They said they are
    using a 0.5 mm filter, the standard thickness is 1.0mm and whereas this
    would not matter for a interference filter it is not good for an
    absorption filter.

    Scott W, Nov 10, 2006
  11. Scott W

    acl Guest

    Ah. So, there are two types of IR blocking filters: Absorption and
    interference. It's not hard to see why interference filters would have
    such a strong angle dependence, now that I think about it. And I take it
    that absorption filters do not display such a strong effect. If so, then
    maybe that is not the problem.

    I was trying to find some kind of rational justification for them not
    having a strong enough IR blocking filter. Hmm....
    acl, Nov 10, 2006
  12. Scott W

    Scott W Guest

    Here is what Lieca has said regarding the IR filter

    "The glass cover of the image sensor of the LEICA M8 is a combination
    of the IR barrier filter and a specially coated protective glass. The
    transmission in the red and infrared region of the spectrum can be
    controlled by the layer thickness of this filter. In the case of the
    Leica M8, which is a very compact system, the thickness of the filter,
    0.5 mm has proved to be ideal. The short back focal length is the base
    for the compactness and the high quality of the standard and wide-angle
    lenses. However, the resulting oblique angle of the incident light on
    the sensor requires special adaptations of the filter.

    Absence of color fringing / Image Resolution
    The extremely thin layer of the filter, 0.5 mm prevents color fringing
    at the corners of an image. This phenomenon, which is also known as
    astigmatism and is frequently encountered with digital SLR cameras, is
    not a problem for the LEICA M8 because of the thin glass cover on the
    image sensor. This feature, plus the particularly high imaging quality
    of Leica M lenses, is the reason for the high corner-to-corner image

    Now when they talk about special adaptations it would seem they are
    saying for their lenses a thicker filter would mess up the image
    resolution, which could well be true. As best as I can figure they did
    the retracing with a 1mm think cover glass and did not like what they
    saw and so went for a thinner filter. This thinner filter does not seem
    to be up to the job.


    Scott W, Nov 10, 2006
  13. Scott W

    acl Guest

    I see, you mean that it's an absorption filter but they had problems
    with the thickness if it was thick enough to do its job, again because
    of the angles of incidence probably. But couldn't they just use a
    different kind of absorption filter that would have the desirable effect
    with the .5mm (or whatever) thickness? I assume the absorption
    characteristics come from adding impurities to the glass or something
    like that (so do not affect the refracting index).

    I mean, it's not as if they are constrained by cost here. Is it possible
    that they simply could not get the right absorption strength with that
    acl, Nov 10, 2006
  14. Scott W

    Scott W Guest

    They may well be constrained by cost, a special filter melt may cost
    more then they are willing to pay. And filters are a standard thickness
    of 1mm. It is easier to cut a filter thinner then it is to get a
    special glass made up but then you had a much weaker filter.

    They may also have gone with a weaker absorbing filter glass to get
    more light through in which case they could go to a more aggressive
    filter glass and perhaps be ok. But even an aggressive filter glass
    like BG18 gets pretty weak at .5 mm. At 1 mm it only passes 1% at
    700nm, but at 0.5 mm that will go up to 10%.

    The other thing I am not sure of is how far can Hoya or Schott push
    their absorption, I have been through this before where we had a lens
    what was not designed to work with a cover glass and wanted to get the
    filter as thin as possible. In the end we redesigned the lens to work
    with the 1mm thick filter.

    Scott W, Nov 10, 2006
  15. Scott W

    Scott W Guest

    So people are just now getting the IR blocking filters that they order
    and these seem to be doing wonders for the images. This is sort of a
    good news bad news thing. For those who really want to use the M8 they
    now have a way to do so and have it take photos with the colors right.
    The bad news is the comparisons with and without the filter show just
    how bad the colors are without the added filter. Green trees turn
    pretty yellow without the filter for instance. So it is clear that
    this is not just a problem with some dark clothing but a problem that
    every M8 owner is gong to need to address if they want their colors to
    be close to right.

    It is was me and I has an M8 and really wanted to use it I would buy
    the needed filters in a heart beat, but also be pretty mad that this
    was not disclosed up front. People will understand that a camera by
    itself if not a whole system and that you need to buy other items to
    make it work. If I buy the body only of a camera I know I will need at
    least one lens and this does not bother me. If I know the camera does
    not come with a memory card again no big deal. But what is a big deal
    is if there is some rather expensive item I need and you don't tell
    me up front that I need it.

    So it will be interesting to see what Leica comes up with and how well
    they take care of their customers. I would imagine that Leica and
    Kodak are going back and forth right now on who is going to pay for the

    It will also be interesting to see if Leica continues to make the M8
    with the weak filter. If they had any sense they would offer in with
    both the weak and normal filter.

    The immediate problem that people are now running into is that the
    supply of these filters seems very limited right now.

    As for the AA filter I don't think Leica is going to do anything about
    that. And after reading a number of posts from Leica fans it is clear
    most are willing to put up with some pretty big flaws in the M8. Even
    with the colors looking like total crap where were/are some who felt it
    was no big problem since you can fix the colors in Photoshop.

    Scott W, Nov 10, 2006
  16. Scott W

    John Turco Guest

    Hello, Scott:

    Totally unacceptable! That's the type of performance one would expect
    of a lousy "keychain camera" (of which I own a few), but not from a
    ridiculously-overpriced, status symbol of a digicam.

    Why, my own, lowly Kodak models produce stunning and accurate colors.
    Oh, hold on...the M8, itself, contains a Kodak sensor! What it probably
    needs (even more than a better "IR blocking filter") is a Kodak "Color
    Science" chip.

    This tiny DSP (digital signal processor) would undoubtedly improve the
    M8's image quality significantly, if not enough to justify the Leica's
    exorbitant cost. <g>

    John Turco <>
    John Turco, Nov 11, 2006
  17. But your Kodak camera uses a Sony sensor...

    David J. Littleboy
    Tokyo, Japan
    David J. Littleboy, Nov 11, 2006
  18. Scott W

    John Turco Guest

    Hello, David:

    It does? Not according to Kodak <http://www.kodak.com>, itself:

    DX3900 Zoom Digital Camera - Frequently Asked Questions http://www.kodak.com/eknec/PageQuer...estionId=27269&pq-locale=en_US&_requestid=265

    Are the Kodak EasyShare digital cameras affected by the Sony
    charge-coupled device (CCD) issue?

    Answer: Answer last updated:08-21-06
    No, Kodak EasyShare digital cameras are not affected by the Sony
    charge-coupled device (CCD) issue because they do not use the
    sensors in question."

    Perhaps (although, unlikely), you had the following, obscure device
    in mind?

    Kodak Dental Systems - KODAK 1000 Intraoral Video Camera

    "Intelligent Precision
    The new KODAK 1000 Intraoral Video Camera, with its improved
    ergonomics and image quality, is more than just a communications
    tool. Equipped with the latest generation SONY high resolution
    CCD, your images are clear and precise. The macro mode (x60) can
    display details invisible to the naked eye."

    John Turco <>
    John Turco, Nov 11, 2006
  19. That was a _particular_ Sony sensor that had a problem. Read the fine print

    "We cannot disclose which manufacturer's sensors we use in our digital
    cameras. The Sony charge-coupled device (CCD) issue does not affect Kodak
    digital cameras."

    (Kodak _might_ not be using Sony sensors, but it's exceedingly unlikely. The
    two other main CCD mfrs (Panasonic and Fuji) don't appear to be selling this
    class of sensor. I think.)

    David J. Littleboy
    Tokyo, Japan
    David J. Littleboy, Nov 11, 2006
  20. John Turco wrote:
    How is a DSP going to distinguish between IR and other energy illuminating
    the sensor?

    David J Taylor, Nov 11, 2006
    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.