Confusion over sharpness-lens aperture

Discussion in 'Digital Photography' started by David Lewis, Sep 5, 2003.

  1. David Lewis

    David Lewis Guest

    Greetings,

    I wonder if someone could help me out by explaining the relationship between
    lens aperture and image sharpness/quality/resolution....and how does wide
    angle vs telephoto figure into it?..see below for reason for my confusion.

    1) In Phil Askey's review of Canon D300 he states....." It's clearly a
    little soft at maximum aperture ..< my understanding=max lens opening/small
    f number>.. (almost all lenses are), but stopped down..<high f number?>...
    it produced good resolution at wide angle and average resolution at full
    telephoto. "

    2) In "Short Courses" it states...."The smallest apertures...<?high f
    numbers>.. may give greater depth of field, but they also have interference
    patterns that soften the image. For the sharpest possible images, you should
    select the largest aperture that will give you the depth of field you need.
    "

    Thanks
     
    David Lewis, Sep 5, 2003
    #1
    1. Advertisements

  2. David Lewis

    David Lewis Guest

    Thanks for reply:

    OK so sharpness suffers (to what ever extent) at both extremes of aperture.

    What is Pop Photo?

    Point taken :>)
    I gather sharpness somehow relates to diffractive errors at narrow lens
    openings?If I am following this logic then my next guess is that wide angle
    lenses additionally narrow the opening..although I don't know how this
    results in a wider field of view ??
    Thanks again
     
    David Lewis, Sep 5, 2003
    #2
    1. Advertisements

  3. David Lewis

    David Lewis Guest

    Ed my intention is to try and take *good* pictures. To do this you must
    understand your equipment. Besides what's wrong with wanting to learn
    something,whether from a book,internet or asking questions..no-one is
    forcing you to answer.I would suggest however if you can't or won't be
    helpful then don't answer at all :>)
     
    David Lewis, Sep 5, 2003
    #3
  4. Sharpness is affected by a lot of factors. The primary factor is
    spherical aberration. This can be corrected by stopping down the lens,
    since the narrower aperture exposes less of the lens, and the spherical
    aberration is smaller. Careful grinding of a lens can give it an
    aspheric shape, but different frequencies of light refract refract at
    different angles at each surface, and the greater the angle the greater
    the difference in refraction is. Stopping down a lens also improves its
    color correction, which results in sharper pictures with truer color.
    Stopping down a lens also reduces the amount of light on the edges of the
    elements, which reduces internal reflections and improves contrast and
    sharpness.

    Diffraction is due to the fact that light has wave-like properties.
    Every time light goes past a sharp edge, like a lens diaphragm, it starts
    to diffuse into the shadow. Red has a longer wavelength, and diffuses
    faster than blue, so diffraction also introduces color aberration. If you
    look at the shiny side of a CD, the rainbow colors you see are the result
    of diffraction. The smaller the aperture, the more likely a light
    wave will be close enough to the edge of the diaphragm to be
    diffracted, and the farther away from the image the diaphragm is, the
    worse the diffraction error.

    That last is why a tiny digital camera lens can perform as well as a much
    larger aperture on a 35mm camera. The diaphragm is much closer to the
    CCD, so diffraction has less of a chance to build up. Lens design also
    makes a big difference. Many compact camera lenses are a retrofocus
    design that has a lens element that brings the image into focus before
    the nominal lens focal length.

    For most lenses, the sharpest *focused* image is somewhere between wide
    open and fully stopped down. Just where that happens varies with the
    lens, but it is usually about one or two stops from fully closed down.
    In other words, if your lens has a minimum aperture of f22, try f16 or
    f11 to get the maximum resolution and contrast of an image you focus on.
    If you want maximum depth of field, stop it all the way down and set your
    focus to the hyperfocal distance.
     
    Larry Caldwell, Sep 6, 2003
    #4
  5. David Lewis

    Tony Spadaro Guest

    Popular Photography. I believe their website is:
    http://www.popphoto.com
    but I could be wrong. They have been testing lenses for many years. I don't
    know how many of their tests they put on line - I rarely buy lenses and
    almost always know a fair amount about them long before I do. In the
    magazine they have a graph showing the performance of the lens at all stops.
    They never have anythign bad to say about a lens, but if you read enough
    reviews you begin to catch the language ("for the budget minded" usually
    means pretty poor - etc.).

    Sharpness suffers at both ends but there are several causes for this and it
    really can only bedetermined on a lens by lens basis. In general it is not
    enough to make a serious difference, but if you are doing landscapes or
    something else where speed and flexibility are not important this is usually
    a good time to set the lens to it's optimum opening and fire away. Most of
    the lenses I've owned over the last 37 years were at their best in the f8 to
    f11 range.

    Besides diffraction there are also aberrations and various other problems
    that crop up in extreme wide angle lenses. Zoom lenses (and wide angle to
    super wide or tele zooms in particular) are very complicated.
    --
    http://www.chapelhillnoir.com
    home of The Camera-ist's Manifesto
    The Improved Links Pages are at
    http://www.chapelhillnoir.com/links/mlinks00.html
    New email - Contact on the Menyou page.
     
    Tony Spadaro, Sep 6, 2003
    #5
  6. David Lewis

    Don Stauffer Guest

    While Ed's reply may have not been the most friendly, I think there is
    an important point in his reply, which is that the subject is way to big
    for the quick replies you'll get here. In addition to looking it replies
    from group, I'd say ALSO go to library and get a good book on
    photography that includes discussion on lenses.
     
    Don Stauffer, Sep 6, 2003
    #6
  7. David Lewis

    NJH Guest

    That's correct.

    Typically, zoom lenses do not have the same performance over their whole
    range of focal lengths. They tend to be sharpest at the wide angle (short
    focal length) end and into the mid-range, and least sharp at the long focal
    length end.

    Correct. F numbers are like shotgun gauges, the lower the number the larger
    the bore. The number is actually a fraction, e.g. f/4 means that the
    diameter of the lens aperture is one-fourth the focal length.

    need.

    Generally, lenses do not give their best performance wide open, though this
    may be less of a problem with the relatively small lenses of digital cameras
    than it is with film cameras. Going to smaller apertures necessarily means
    going to a slower shutter speed (because the smaller aperture passes less
    light in a given time) and that of course increases the possibility of loss
    of sharpness due to camera shake. Since the very small physical apertures of
    digital camera lenses compared to film cameras (i.e., at the same f number)
    provide much more depth of field anyway, it usually is good advice,
    especially when shooting hand held, to lean toward large apertures as
    suggested in the piece you've quoted.

    Neil
     
    NJH, Sep 6, 2003
    #7
  8. David Lewis

    NJH Guest

    No, it isn't the narrow lens opening that produces the wide field of view.
    Changing the aperture doesn't change the field of view at all. Only changing
    the focal length (assuming the same format, of course) does that.

    What Tony is saying is that the relationship between f number and the actual
    physical dimensions of the aperture depends on the focal length of the lens.
    An f/11 aperture on a 24mm lens would be twice the PHYSICAL size of an f/11
    aperture on a 12mm lens. For the apertures of those two lenses to be the
    same physical size, the 24mm lens would have to be set to f/22.

    This is because, as mentioned in my post just previous, the f number
    actually expresses the aperture size as a fraction of the focal length. In
    the case of a 24mm lens, an f/4 aperture would be 6 mm in diameter. In a
    12mm lens, an f/4 aperture would be 3 mm in diameter.

    Neil
     
    NJH, Sep 6, 2003
    #8
  9. David Lewis

    David Lewis Guest

    Exactly but it needs to be delivered in a helpful way like yours is in order
    to be credible.

    Yes and No.I didn't expect an encyclopaedic response but I *did* get an
    answer to my question..thanks to Tony and Larry!
    The irony here is that my question was generated from an apparent
    contradiction from different *reading* sources outside of this forum.I
    needed clarification on that point.

    Thanks for input :>)
     
    David Lewis, Sep 6, 2003
    #9
  10. David Lewis

    David Lewis Guest

    Neil,many thanks for clarification.
     
    David Lewis, Sep 7, 2003
    #10
  11. David Lewis

    NJH Guest

    That's correct, at least with film cameras. Most 35mm camera lenses will
    give their best sharpness closed down two or three stops, but "fast" lenses
    (i.e., large maximum aperture) are often at their best stopped down even
    more. A fixed focal length 50mm f/1.4 lens for example may give its best
    performance at about f/8, which would actually be closed down five stops
    from wide open. My own lens testing over many years has shown that a wide
    variety of lenses, both fixed focal length and zoom, are best at around f/8.
    It should be noted though that a quality lens would give excellent
    performance over a range of many stops; finding the critically best stop is
    not necessarily of great importance. And also that's not a hard and fast
    rule.

    I suspect it's different with digital camera lenses, because of their much
    smaller size and consequently smaller physical aperture size at any f-stop.
    That is, I believe diffraction becomes more of a problem at larger f-numbers
    with digicams; with 35mm cameras it doesn't become much of a problem until
    you get down to around f/22 or so, generally speaking, and even then it's
    not terribly troublesome. The digital cameras I've seen won't even go down
    to that small an f-stop, and my guess is that this is because diffraction
    becomes more worrisome with the physically smaller lenses. So all things
    considered, shooting wide open with a digital camera is under most
    circumstances likely to be a much better idea than it would be with a 35mm
    camera, I think. I've only been in digital cameras a couple of years myself
    and haven't gotten around to do any lens testing.

    If the f number is kept the same, the physical aperture of a zoom lens gets
    smaller as it moves to shorter focal lengths, yes. But that isn't considered
    stopping down. Stopping down means actually changing to a higher f number
    (i.e., a smaller RELATIVE aperture). It's the f number that's an important
    determinant of exposure, the relationship of aperture to focal length, not
    just the physical size of the aperture.

    An example sometimes given is that of a window. If you're reading by window
    light, it's not only the size of the window that matters but how close you
    are to it. A small window might be plenty large enough if you were sitting
    right next to it, but if you were 40 feet away the window would have to be
    very much larger to give you the same amount of light. In this case of
    course window size corresponds to aperture, and distance from the window
    corresponds to focal length.

    One other thing in this connection is that MOST zoom lenses today go to
    slightly smaller (sometimes not so slightly) f numbers as they go to the
    longer focal lengths. For example, the 7X zoom lens on my Minolta DiMAGE 7i
    is f/2.8 wide open at the wide angle end, but f/3.5 at the long end. My
    DiMAGE S404 with a 4X zoom is f/3.0 at the short end, f/3.6 at the long. And
    so on.

    Intermediate settings may or may not vary in the same way. Some digital
    cameras have just two apertures, wide open and stopped down a couple of
    stops or so. In such a case the stopped down aperture usually also changes
    in f number as the lens is zoomed. Some digital cameras report this change
    correctly and others do not.

    Any time, David, and you're more than welcome.

    Neil
     
    NJH, Sep 7, 2003
    #11
  12. David Lewis

    David Lewis Guest

    NJH wrote:

    My new found expert Knowledge (joke) tells me that aperture in mm = focal
    length divided by f number.........meaning the physical size of the lens
    shouldn't come in to it ??

    Except that with overall smaller dimensions the image sensor is closer to
    the lens and negates diffraction problems??
    Well explained. The f number will in fact stay the same then as you zoom in
    or out because the physical size of the aperture changes to compensate.The
    exposure stays the same but there is still altered physical characteristics
    due to change of size of lens opening eg smaller size=more potential for
    diffraction.

    A helpful analogy.Thanks
    I'm glad you said this because I did in fact notice this. I presume these
    lenses are trying to "catch light" with the trade off being less sharpness
    etc at wider apertures.

    Thanks for your time and patience :)
     
    David Lewis, Sep 7, 2003
    #12
  13. David Lewis

    David Lewis Guest

    Oh. In that case they would tend to relatively underexpose the image by
    cutting off more light but improve the quality of the image by not getting
    too close to the critical zone of poor lens performance (by avoiding ultra
    wide lens opening)?

    Hey,I'm a Newbie...confusion - it's part of the job description ;-)
    Thanks again
     
    David Lewis, Sep 7, 2003
    #13
  14. David Lewis

    NJH Guest

    That's correct, at least effectively. Zoom lenses are such complex designs
    that I'm not sure that's always true depending on where the diaphragm (the
    adjustable part which creates the actual physical aperture) is placed in the
    light path. But for practical purposes, a 20mm lens at f/4 will have an
    aperture 5mm in diameter.

    I THINK the reason it does is that at extremely tiny physical apertures, the
    non-image-forming light resulting from diffraction becomes a larger
    percentage of the whole. Diffraction of course results from light passing an
    edge, and as the diameter of a circle is reduced its area will be reduced
    faster than its circumference.
    It certainly would as far as angular considerations are concerned, yes. That
    is, as diffraction produces a blur circle instead of a point at the image
    plane, the farther away that image plane is the larger would be the blur
    circle. That would imply that diffraction losses scale with size and would
    cancel out, so to speak, in the print or other final image, since a smaller
    CCD-plane image would have to be magnified (and its blue circle would be
    magnified) proportionately more than a larger CCD- or film-plane image.

    Again, I THINK the difference here is that these things change as the format
    size changes, and actual PHYSICAL size of the aperture (not just f number)
    has some bearing on diffraction. I'm not sure of this. But I know it is true
    that with very large format cameras, e.g. 8 x 10 view cameras, many
    photographers commonly used extremely small RELATIVE apertures like f/64. I
    have never seen a 35mm camera lens with that small a stop and doubt that
    there is one; the smallest relative apertures on any of my 35mm lenses is
    f/32, and that only with telephoto, macro or long zoom lenses.

    On none of my digital cameras is there that small a relative aperture. For
    example my DiMAGE 7i only stops down to f/9.5 (and at the shortest f.l. it
    stops down only to f/8), and that seems fairly typical.


    Correct, or at least it would be if zoom lenses really did maintain the same
    f number throughout their range. (That is exactly what almost all zoom
    lenses DID, many years ago. Some still do, but most now gradually shift to a
    smaller relative aperture as the lens is moved to a longer focal length.
    This makes it possible to design much smaller, lighter and cheaper zoom
    lenses--albeit at some loss of lens "speed" especially at the long end where
    a large relative aperture would be most appreciated, unfortunately.)

    Using again the example of the 7X zoom on my DiMAGE 7i, the lens has a
    maximum aperture of f/2.8 at the short (in this case 7.2mm) end and f/3.5 at
    the long (50.8mm) end. The effective physical aperture wide open is
    therefore about 2.57 mm at the short end and 14.5 mm at the long end. The
    PHYSICAL aperture becomes larger almost, but not quite, in proportion to the
    increase in focal length. The RELATIVE aperture becomes slightly smaller
    (i.e., by about two-thirds of a stop over that range) at the same time. The
    camera of course compensates by shifting to a slightly slower shutter speed
    as the relative aperture becomes smaller.

    Well, it's just the way most zoom lenses are made these days; on the whole
    such lenses are most attractive to buyers. Smaller, lighter, and less
    expensive than they'd be otherwise. My guess is that generally they're at
    least as sharp at the short, wider aperture end, even wide open, as at the
    long end. I really will have to get around to lens testing one of these
    days.

    Neil
     
    NJH, Sep 7, 2003
    #14
  15. David Lewis

    NJH Guest

    They would, if the camera did not compensate by shifting to a lower shutter
    speed. But that is what the camera does, or should do, at least in any
    automatic mode.

    No, not really. A faster (wider maximum aperture) lens is not necessarily
    less sharp than a slower one. Stopping down does almost invariably improve
    lens performance (at least it does with larger format lenses), but a zoom
    lens at the long end is not stopped down just because its maximum relative
    aperture is smaller--it's still wide open. Whether a zoom lens is sharper,
    less sharp or equally sharp at either end of its zoom range when wide open
    depends on the particular lens. But generally speaking, zoom lenses tend to
    be least sharp at the long end.

    Neil
     
    NJH, Sep 7, 2003
    #15
  16. David Lewis

    NJH Guest

    Sorry, that should be "blur circle." Damn typos anyway.
     
    NJH, Sep 7, 2003
    #16
  17. David Lewis

    David Lewis Guest

    Makes sense.The size of the 'hole' is the same but the resultant diffraction
    has a proportionately bigger impact.
    Ah OK I wondered why they would do this.


    This was actually based on "smaller f numbers" rather than "smaller
    apertures" which you corrected later.

    Neil, sincere thanks for your herculean effort to educate a newbie. It has
    helped enormously!Some of the concepts are difficult to wrap your mind
    around but the picture slowly comes in to view (even after that second
    scotch!) I can only hope you got something out of the exercise also...like
    they say, to teach is to learn twice!

    All the best
     
    David Lewis, Sep 8, 2003
    #17
    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.