The problem of SLR lenses on DSLRs

Discussion in 'Digital SLR' started by Rich, Sep 11, 2005.

  1. Rich

    Rich Guest

    Rich, Sep 11, 2005
    1. Advertisements

  2. Rich

    Jeff R Guest

    Jeff R, Sep 11, 2005
    1. Advertisements

  3. Rich

    Colin D Guest

    Sod me, Rich, you're beyond any help.

    One: The bloody lens was wide open. ANY lens wide open will vignette
    to some degree. Even non-Canon ones (horror!)

    Two: despite the write-up, the image was shot indoors. Any schoolboy
    will tell you that the upper walls and ceiling have less illumination
    than lower down. High areas in a room are lit only by reflection from
    the floor and furniture. This is shown by the lower corners of the
    image not being 'vignetted' to the same degree as the upper corners.

    Three: Post processing may well have included darkening the top corners
    - a well respected practice to focus attention on the subject and
    prevent the eye from wandering out of the picture. Practically a
    *requirement* for good portraiture.

    In short, you are clearly ignorant about photography in general, and you
    do yourself a major disservice every time you open your very large mouth
    about things you have no grasp of.

    Conclusion: You're just a troll who has managed to jerk two reactions
    from me today. Congratulations. Laugh yourself sick.

    Colin D.

    PS: Part of the reason behind my reactive answers is to try to undo the
    harm you are doing to posters or lurkers who may be tempted to believe
    your irrational outbursts.
    Colin D, Sep 11, 2005
  4. Rich

    Rich Guest

    Vignetting is a consequence of using 35mm lenses (because the light
    The most obvious conclusion is that the lens vignetted, not the
    convoluted "path" you suggested was the cause of what you see in that
    image. Vignetting is one of the failings you see when 35mm FILM
    lenses are used on digital cameras. This was all explained when the
    4/3 system was devised, too bad you didn't read about it.
    Rich, Sep 11, 2005
  5. Rich

    Darrell Guest

    4:3 sensors suck so their promise is pretty much lost. 4:3 is digital APS, a
    promise that never made it.
    Darrell, Sep 11, 2005
  6. Rich

    MarkH Guest

    Ummm, I can't find the comparison images with the same lens and same
    settings but with a film SLR. So the evidence that suggests that the
    problem is because the camera is a D-SLR using a lens designed for a film
    SLR is what?

    Mark Heyes (New Zealand)
    See my pics at (last updated 5-September-05)
    "The person on the other side was a young woman. Very obviously a
    young woman. There was no possible way she could have been mistaken
    for a young man in any language, especially Braille."
    MarkH, Sep 11, 2005
  7. The photograph you posted as evidence was likely caused by the lighting
    conditions in the room where the image was made, not by the hardware
    involved. But then, you've yet to demonstrate that you know shit about
    Randall Ainsworth, Sep 11, 2005
  8. What is more interesting, is that if you want the same DoF and field of view
    on an 1.5x APS-C dSLR, you need a 33/0.93.

    I'm not aware of any 35/1.0 lenses. So you simply can't take this shot with
    an APS-C sensor.
    Philip Homburg, Sep 11, 2005
  9. Any idea how that lens would vignet on film???

    Good sensor arrays use microlenses. Those microlenses are probably
    more efficient in concentrating the oblique incident light on the
    photosensitive surface than film without microlenses is. The fact that
    light has to travel further to reach the corners remains.

    Bart van der Wolf, Sep 11, 2005
  10. SNIP
    Is it? I've yet to see quantitative proof.

    In fact, in competent designs (with microlenses) a FF-sensor array
    like this 1Ds Mark II with the EF 50mm f/1.4 shot at f/8.0 is quite
    well behaved:
    , and I'd be surprised to see much different results from film.

    Bart van der Wolf, Sep 11, 2005
  11. Exactly, unsubstantiated speculation (at best)!

    In fact, even with just a DSLR (EOS 1Ds Mark II in this case) it is
    trivial (with the right evaluation tools) to demonstrate that lens
    design and aperture have a dominant influence on light fall-off (just
    like with film):

    Although a nice experiment for the future, the light fall-off at f/8.0
    is so small that I won't even bother to waste a film on it to
    demonstrate what the effect of film surface reflection has, compared
    to the micro-lenses on a good FF-sensor array:
    The microlenses almost guarantee a perfect angle of incidence.

    Bart van der Wolf, Sep 11, 2005
  12. Rich

    Rich Guest

    Don't the microlenses actually increase the angle of incidence of the
    light beam as it goes through them? Unless of course physics has been
    suspended and the microlenses display negative refraction?
    Rich, Sep 11, 2005
  13. Rich

    Rich Guest

    It only stands to reason. If what the other poster says is correct,
    that wide open that lens will vignette with film, then because the
    light rays coming from the edge of the lens are highly oblique, the
    vignetting would exist unless the lens was physically much larger than
    it actually is (so you'd only be using the centre of the lens to form
    the image instead of the whole aperture). If vignetting is made worse
    when using a sensor instead of film and if the lens already displays
    vignetting with film when wide open, then it's reasonable to assume
    that the edge darkening shown in the picture (gross as it was) was
    vignetting and was so severe only cropping (and not adjusting the
    illumination levels in software) would be needed with the image.
    Many of the "add-on" tele and WA attachments for P&S cameras exhibit
    this problem. Talk of using "f8" really makes no sense in the context
    of that image since the goal was a blurred background, as DSLR owners
    constantly point out is one of the benefits of DSLRs against P&S
    cameras, for portrait work.
    Rich, Sep 11, 2005
  14. Yes, most lenses will vignet more when wide open than when stopped
    down, no different for the f/1.4 although the geometics of the design
    will cause an f/1.4 vignet more likely than e.g. an f/2.8 or smaller.
    That's not really what happens. I think the reasoning flaw can be
    explained from a too simple mental diagram of ray propagation through
    a lens. The light entering the lens should be envisioned as a cone of
    light emanating from each point of the original scene as seen by the
    entrance pupil of the lens. That cone will travel through the lens,
    its diameter will then be restricted by the aperture and will leave
    the lens at an apparent exit pupil, still as a 'cone' of light.

    The cone of light, which may look like an elliptical (due to looking
    through a tube from an angle) cone of light will also have to travel
    farther to the corners of the flat plane of projection than to the
    center of the projection plane (rectilinear design).

    The result for the EF 50mm f/1.4 lens for the 4 largest full stops
    looks like this:
    .. Mind you that's all on the same sensor, with only the corner ellipse
    of light turning into more of a cone of light, but still traveling a
    longer distance to the corners (so some loss is inevitable).
    No, it would mostly have to be caused by surface reflection
    characteristics by part of the light ellipse, but it's
    reduced/eliminated by microlenses as present on the sensor array
    tested above (but not on film).
    Yes, as explained above.
    Mostly lens vignetting, nothing to do with the sensor yet! Try looking
    through a detached lens as you move your eye off-axis. If the exit
    pupil turns into an ellipse, the lens will vignet.
    Vignetting adjustment in software is very useful if needed, and the
    relative under-exposure of the corners re-enforces the "expose to the
    right" axiom for RAW capture. Cropping is not necessary under most
    circumstances. The excellent higher ISO performance often allows to
    stop down a little.
    f/8 is relevant to demonstrate that it is mostly the lens that
    vignettes at larger apertures, not the sensor.

    Bart van der Wolf, Sep 11, 2005
  15. SNIP
    No, the polymers slow down the light (refractive index > 1.0) and thus
    behave as condensors/lenses.

    Bart van der Wolf, Sep 11, 2005
  16. Rich

    Sheldon Guest

    The only vignetting I see is not vignetting but the way the image was lit.
    Except for the kit lens, I use all 35mm lenses on my D70, and I've never had
    any problems with vignetting, at least that I or anybody else can see.
    Sheldon, Sep 12, 2005
  17. Rich

    Rich Guest

    That's a 1.5 sensor, the lenses more than cover it so it's not an
    issue I guess. A full frame however might be different.
    If a lens open at f1.4 vignetting with 35mm film, then for sure it
    will with a full-frame sensor.
    Rich, Sep 12, 2005

  18. Please stop posting bollox - Olympus claim that their lenses are 'designed
    for digital', yet many vignette so badly Olympus had to include 'shading
    compensation' on their DSLR's

    I know for a fact that the 14-45, 40-150 and the 14-54 Zuiko lenses have a
    vignetting problem - so please stop pissing about with your puerile
    trolling - you're clearly in need of an obliging woman to take your mind off
    these forums.
    Pinky & Perky sing for royalty, Sep 12, 2005
  19. Rich

    Skip M Guest

    Yes, but, Rich, your original implication was that 35mm film lenses would
    vignette on digital cameras (FF) BECAUSE they were used on digital cameras.
    To wit: "Vignetting is one of the failings you see when 35mm FILM lenses are
    used on digital cameras." Since the 50mm 1.4 doesn't vignette on film, I
    certainly wouldn't expect it to vignette on digital, and the image you
    posted doesn't show any vignetting, only light variations on the background.
    Skip M, Sep 12, 2005
  20. Rich

    Ed Ruf Guest

    That is tautologically correct, so then your point is? Changing the sensor
    the image is projected upon doesn't change the optical properties of the
    Ed Ruf, Sep 12, 2005
    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.