Fixing Focus in post processing

Discussion in 'Australia Photography' started by Ken Chandler, Nov 22, 2005.

  1. Ken Chandler

    Ken Chandler Guest

    Ken Chandler, Nov 22, 2005
    #1
    1. Advertisements

  2. Ken Chandler

    k Guest

    k, Nov 22, 2005
    #2
    1. Advertisements

  3. Ken Chandler

    Noons Guest

    Noons, Nov 23, 2005
    #3
  4. Ken Chandler

    kosh Guest

    I get the feeling the method they use to do this involves taking
    advantage of the Circle of Confusion.

    I vaguely understand this concept, but does anyone happen to have a
    concise description of what the Circle of Confusion and how it relates
    to depth of field?

    I have found the following definition which explains it somewhat,
    however I think this also comes into play with issues relating to sensor
    size and depth of field control.

    -Circle of Confusion
    -The largest on-film circle that you can see as a well defined point on
    -an 8x10 print when viewed at from a "normal" viewing distance of 2 to 3
    -feet. Anything larger is seen as a small circle, not a point and is
    -therefore perceived as out of focus. For 35mm format the diameter of
    -such point or circle is 0.025mm, commonly rounded to 0.03mm

    actually... whiie writing this e-mail... found sme interesting links and
    definitions.....

    I understan the maximium enlargement stuff.... but still wonder IF this
    also relates to whay a photo taken at equivilant focal lengths both at
    the same aperture will appear to have a shallower depth of field on the
    larger sensor?

    http://en.wikipedia.org/wiki/Circle_of_confusion
    http://www.nikonlinks.com/unklbil/dof.htm#method

    something a bit different than which camera is best eh!?!

    kosh
     
    kosh, Nov 23, 2005
    #4
  5. Ken Chandler

    k Guest

    | I vaguely understand this concept, but does anyone happen to have a
    | concise description of what the Circle of Confusion and how it relates
    | to depth of field?
    |
    | I have found the following definition which explains it somewhat,
    | however I think this also comes into play with issues relating to sensor
    | size and depth of field control.


    here's my effort re an explanation:

    First on DOF and coc, my understanding suggests that the first component to
    this discussion that must be examined is what a lens actually does when
    focussed.


    If you grab a magnifying glass,go outside and start to focus the rays of
    the sun on the ground you will notice that if the magnifying glass is too
    far from the focal length of the lens you will produce a large circle. As
    you move closer to the ground the circle will get smaller until you produce
    a spot, then if you move closer again you will again get a circle.

    Thinking about this little experiment three dimensionally you will realise
    that the lens is producing a 'cone' of light either side of the focal
    length. Imagine now the same application in a camera where the lens is in
    a fixed position where you have focussed on an object nearby - the light
    enters the lens in the same way as above and you have (by focussing the
    lens) brought the film into the position where the cone of light focuses
    _on_ the film plane for your subject.. BUT the other stuff in the view of
    the lens is also bouncing light in, with these cones of light focussing
    either in front of or behind the film plane.

    Obviously you will not have a spot forming on the film, but a 'slice' of
    that cone.. actually a circle (like the one the magnifying glass
    produced) - there will be many, many circles being produced from the almost
    infinite points in the scene in front of the lens, and these will all
    record on the film AS circles (but they blend together to form an out of
    focus fuzziness) and these circles are termed the 'circles of confusion'.

    This comprehension of the 'cones of light' leads to another interesting
    thing about those circles - Think for a moment and presume we were talking
    about the lens being wide open in the above example.. obviously those
    circles are going to be large - hence we get that wonderful shallow depth
    of field look when we photograph a nearby subject - the subject is sharp
    with the background being a blur of colour or tone.

    SO what happens when we stop the lens down? - the aperture becomes smaller
    and the angle of the cones of light become thinner. Lets look at the
    example of the magnifying glass again, presuming this time you had a small
    aperture in the magnifying glass. You quickly come to realise that it's
    hard to see where the lens is actually in focus because the spot looks
    almost the same size when positioned close, near OR far. Translating this
    to the camera/film we see that the circles of light that strike the film
    now are all of a comparable size, not the big cone/small spot as it was
    before - this means ALL things in the image look like they are in focus!
    This is why you get a shallow DOF with a lens wide open and a deep DOF when
    the lens is stopped down - those circles that make up the image are varying
    in diameter as you open or close the aperture :)


    As you can see - the image forming on your film/sensor array/whatever is a
    product of the lens and the lens alone. The varying size of those circles
    is a product of the aperture. Neither contribute to what looks sharp as
    there is only a single plane at which the lens focusses... (and that's
    really only for a given wavelength for many lenses!)

    As to whether it makes a sharp spot that looks 'in focus' - an object that
    when defocussed through a lens produces a circle of 1/2mm in diameter
    whether on film or on a ccd is still producing a 1/2mm circle! - lets say
    this 1/2 mm spot was on an 16x20 neg that we reduce to a 8x10 image - it
    will look sharp - if it were in a half-frame neg and we enlarged it to an
    8x10 it would be a blurry spot...

    ...that is assuming you were using the same focal length lens.

    so DOF relies on the point appearing sharp as viewed on a print at an
    accepted viewing distance, by convention this print is given as being an
    8x10 held at about 40cm. (if it were viewed from 3m away it would appear
    smaller and the image would appear sharper)




    lets look at circles of confusion, and accept the usual definition that it
    relates to the percieved depth of field seen in an 8x10 print.

    the coc values are generally accepted as:
    diameter film format
    0.250mm 8x10
    0.125 5x4
    0.60 6x6
    0.029 35mm

    now lets see what the DOF for some of these are and how the DOF relates

    H=hyperfocal distance
    f= focal length
    c= diameter of coc
    u= object distance
    N= f number of lens

    a couple of formulae:

    Hyperfocal distance= f^2 / (NxC)

    near limit of sharp focus= HxU / (H+U)-f

    far limit of sharp focus = HxU / (H-U) -f

    run a few calculations through this and you will find that a larger format
    will have a greater DOF than a smaller format using the _same_ focal length
    lens. therefore If I were to use a 50mm lens on a ccd camera the DOF would
    be substantially less than for a 35mm camera.

    BUT if I were to use 'equivalent' focal length lenses (lenses on each
    format that has close to the same field of view) the smaller format will
    produce a greater depth of field. a 50mm lens on a 35mm camera has a
    greater DOF than an 80mm on a 6x6
     
    k, Nov 23, 2005
    #5
  6. Ken Chandler

    Noons Guest

    Looks to me like a combo of a bucketload of small lenses
    between the real lense and the sensor plus software
    to make sense of the resulting pixel blur.
    If they can establish what it should look like from
    standard targets, then if the lenses are really small
    and heaps of them, it should be possible to extrapolate?
    Interesting stuff.
     
    Noons, Nov 23, 2005
    #6
    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.