Now that Nikon is outselling Canon....

Discussion in 'Nikon' started by RichA, Oct 8, 2007.

  1. RichA

    Patrick L Guest



    Perhaps I'm wrong, but it seems to me that back or front focusing on a
    manual lens is quite possible in theory.

    Isn't it true that all these flaws are is that when the shot seems to be in
    focus in the viewfinder, the shot is actually focused in front or in back
    of the subject? You are confusing the idea of not being able to AF focus
    at all with this phenomenon (for which manually focusing would be the cure),
    it is not the same thing.


    Patrick
     
    Patrick L, Oct 11, 2007
    #21
    1. Advertisements

  2. Nope. The matte/ground glass/split screen or whatever you use
    is not, and can not, be used for AF. It's used for AE (auto
    exposure) in at least some cameras. Google for submirror.

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 11, 2007
    #22
    1. Advertisements

  3. You're correct. It's misalignment of the mirror which would throw off
    autofocus. But either way, it wouldn't be the fault of the lens.

    What led me down this path was the original poster's report that if he
    adjusted the focus screen for the Noct, none of his other lenses would
    focus correctly.
     
    Michael Benveniste, Oct 11, 2007
    #23
  4. Either in focus in the viewfinder or via use of the AF confirmation
    light, yes. I can see how a misadjusted mirror, submirror, or focus
    screen could cause front- or back-focus, but those are all in the
    camera.

    Similarly, a misadjusted open-loop AF system could in theory cause
    lens-based frontfocus or backfocus, but a manual focus lens doesn't know
    what's behind it and can't misinterpret any instructions from the
    camera.

    So while backfocus or frontfocus could occur while using a manual focus
    lens, how could it be the fault of the lens? The only thing I can think
    of is if the stop-down action somehow knocks the lens out of focus, but
    then it would be visible after the shot.
     
    Michael Benveniste, Oct 11, 2007
    #24
  5. RichA

    frederick Guest

    Mentioned also up the thread was the 80-200 AF-D Nikkor
    where this occurs at 200mm, close focus distances.
    AFAICT, that's a consistent problem with Nikon consumer
    level DSLR bodies through to the D200, but not a problem
    with the D2 series (and hopefully D300/D3).
    There's some other AF-D lenses where this happens (35-70 2.8
    etc)
    And it's not misalignment of the mirror, as if you adjust to
    correct for the backfocus at close range, the lens will
    front focus at infinity, and close at shorter focal lengths.
    The severity of this in practice with the 80-200 is easy to
    overestimate. DOF at f2.8 / 200mm / close distance is so
    shallow that only part of the area enclosed by the AF
    brackets will be in sharp focus unless the subject is very
    close to being on the same plane as the sensor. It's better
    to use MF, as with Macro, and ignore the electronic
    rangefinder light.
     
    frederick, Oct 12, 2007
    #25
  6. RichA

    jean Guest

    On B&H's site, you can't even get on a waiting list for a Canon 40D, not so
    with out of stock Nikon's... wonder why? The smart asses will just say
    Canon is too small a company to supply as many cameras as people want, but I
    have my doubts.

    Jean (I want a 40D)
     
    jean, Oct 12, 2007
    #26
  7. Nikon notes it as an issue in footnote 9 on pages 244-245 of the D2Xs
    user manual as well. My own test with a D200 showed a slight error, but
    not the inch or two that others have reported.

    When I asked Nikon USA about it, they claimed the problem was "an
    anomaly of the autofocus software in the camera," but I have no
    idea why it happens with dSLR's and not with film SLR's.

    While I'd like to upgrade my 80-200mm to a 70-200mm VR, this minor
    glitch will have no impact on my decision.
     
    Michael Benveniste, Oct 12, 2007
    #27
  8. RichA

    Robert Brace Guest

    Read pages 244-245 note 9 again carefully and compare it with the table on
    page 244 and you will find the condition described in note 9 specifically
    DOES NOT refer to the AF-D lenses. In fact all Nikkors with letters
    FOLLOWING the AF designation i.e. AF-D, AF-S, AF-I, AF-G etc, etc are NOT
    AFFECTED. Neither is the AI-P nor the PC 85mm f2.8D.
    The only AF lenses that ARE AFFECTED are "Other AF" lenses except those for
    the F3AF.
    Now, in my interpretation, that means all older pre-letter lenses.
    Thus endeth the lesson!
    Bob
     
    Robert Brace, Oct 12, 2007
    #28
  9. RichA

    frederick Guest

    The D70 manual has the description:
    "AF 80-200mm F/2.8S"
    My 80-200 AF-D behaves exactly as described WRT back-focus
    at 200mm close range, and electronic rangefinder error.

    AFAIK with the possible exception of coatings, the optical
    formula for all AF 80-200 lenses is the same - except for
    the AF-S.

    I don't think the above refers to AF-S, because they also
    refer to the 35-70 as "AF 35-70 F/2.8S".

    So if you know of AF f2.8 nikkors that aren't "SWM S", but
    are some other "S", but aren't "D", please let me know.

    I suspect that it's a typo by whoever Nikon contracts out
    it's english manual writing to.
     
    frederick, Oct 12, 2007
    #29
  10. RichA

    Paul Furman Guest

    It's not easy to imagine how misfocus could be the fault of anything but
    the body. If the in-lens focus motors twist wrong, the camera AF sensor
    should see that and correct for it.

    I suppose it's concievable there is some feedback loop that fools the
    body into overreacting and not re-focusing the final correction and that
    could be skewed toward close or far focusing though it seems possible
    that would vary depending which direction you were focusing from.
    Doesn't one of the newer Canon's have an AF correction on a lens by lens
    basis? If they bothered with that, I guess it's possible.
     
    Paul Furman, Oct 12, 2007
    #30
  11. Think on these assumptions:
    - The Noct is focussed on the ground glass.
    - All the other lenses are focussed by the AF.
    - The ground glass is misaligned, but not enough to be very visible


    In that case you'd misfocus if you really use the ground glass
    for focussing, but not notice that the AF focusses slightly
    differently. Then you get good focus with AF, but not with the
    Noct, but if you correct the focus screen for the Noct, _all_ of
    his other lenses would focus correctly (even on the ground glass).

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 12, 2007
    #31
  12. Thank you for your interpretation. Nikon USA's interpretation is
    that "the problem would be limited to the 80-200mm f/2.8D AF-Zoom-Nikkor,
    Nikon product #1986." Here's a link to product 1986:
    http://www.nikonusa.com/template.php?cat=1&grp=5&productNr=1986

    The wording on the footnote for the D2Xs is identical to that for
    the D200 and D70.
     
    Michael Benveniste, Oct 12, 2007
    #32
  13. Agreed, but the person explicitly stated that he was using the
    focus screen for all his lenses, and the Noct was the only one
    which focused incorrectly.
     
    Michael Benveniste, Oct 12, 2007
    #33
  14. An autofocus system can be "open loop" or "closed loop." In a
    closed-loop system, the camera checks focus after the final correction
    is made. In an open-loop system it does not, so your "not-focusing"
    statement applies.

    The notes I've seen indicate that Canon's system is open-loop. For
    example:
    http://www.fredmiranda.com/forum/topic2/241524
    So at least in theory, front- and backfocus is possible.

    Nikon's "screwdriver" system has to be closed-loop, since the camera
    doesn't know how much focus changes with change of angle of the
    screw. I don't know if Nikon's AF-S system is closed-loop or open-
    loop; it may depend on operating mode. Nikon does use open loop systems
    elsewhere, such as aperture adjustment.
     
    Michael Benveniste, Oct 12, 2007
    #34
  15. Hmmm.
    Used the focus screen explicitely for manual focussing?

    I do use the focus screen on all my lenses. Framing and so on,
    you know?

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 12, 2007
    #35
  16. The AF itself is open loop, but setting the lens is closed loop.
    (you may notice some lenses overshooting, then correcting.)

    -Wolfgang
     
    Wolfgang Weisselberg, Oct 12, 2007
    #36
  17. RichA

    Robert Brace Guest

    The wording is there for all bodies from the D40 on up.
    D40 & D50 wording differs slightly and leave the impression the phenomenon
    only is evident if you minimum focus THEN zoom to maximum zoom setting. The
    wording in the others leaves a slightly different impression of the order in
    which these actions must take place.
    I agree that it must be down to the translation interpretation. Nikon USA's
    version notwithstanding.
    I have never encountered it in the D2 series of bodies with neither AF-D nor
    non-D AF lenses nor the AF-S series.
    I also have a 50mm f1.4 Nikkor-S (AI converted) which focuses accurately on
    the stock groundglass.
    Bob
     
    Robert Brace, Oct 12, 2007
    #37
  18. RichA

    Paul Furman Guest

    The only lens I really have hunting with is the 105 VR macro. That thing
    struggles so much.
     
    Paul Furman, Oct 13, 2007
    #38
  19. In theory the explanation can be like this:
    - when you are looking through the view finder, you effectively see an
    aperture of about f/2.0 or f/2.8 (depending on the camera) or smaller
    (for slow lenses, when stopping down, etc).
    - So when you are focusing an f/1.2 lens for a picture to be taken with the
    lens wide open you are not exactly getting what you see through the view
    finder
    - certain optical effects may change the middle of DoF when you stop down
    - so, if the Noct exhibits this effect then you may be focusing the Noct on
    the wrong stop, and adjusting the focusing screen can be used to compensate
    for that.

    The obvious thing to try is to see of the problem goes away if you take
    pictures with the Noct at an aperture of f/2.8.
     
    Philip Homburg, Oct 14, 2007
    #39
  20. It's a better explanation than what I've heard so far, but would
    you care to expand on what these "certain optical effects" are?
     
    Michael Benveniste, Oct 14, 2007
    #40
    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.