Anyone care to espouse on DOF for digital

Discussion in '35mm Cameras' started by Alan Browne, Feb 27, 2004.

  1. Alan Browne

    Alan Browne Guest

    considering the smaller capture area of most 35mm DSLR style cameras.

    Which leads to a smaller CofC (say 0.02 instaed of 0.025mm) for a print
    of size x.

    What will be the effect in the viewfinder when using DOF preview? Will
    DOF preview remain a reasonable way to evaluate the DOF using our lenses
    designed for 36x24mm cameras?

    Thanks,
    Alan
     
    Alan Browne, Feb 27, 2004
    #1
    1. Advertisements

  2. Alan Browne

    columbotrek Guest

    Yes it does. Think of what would happen if you snipped an PS sized
    frame from the center of your 35mm film. Would the DOFF be any
    different? Another way of looking at things is that with an SLR, you are
    looking through the lens which is going to be taking the shot. The
    perspective you see will be the one you record.
     
    columbotrek, Feb 27, 2004
    #2
    1. Advertisements

  3. Alan Browne

    columbotrek Guest

    Yes it does. Think of what would happen if you snipped an PS sized
    frame from the center of your 35mm film. Would the DOFF be any
    different? Another way of looking at things is that with an SLR, you are
    looking through the lens which is going to be taking the shot. The
    perspective you see will be the one you record.
     
    columbotrek, Feb 27, 2004
    #3

  4. Alan

    DOF preview with DSLR's works fine - What you see in the viewfinder is what
    you get. DOF remains the same as with 35mm - the one differing factor being
    the distance you will be taking from to 'fill the frame'. Look at the DSLR
    sensors as nothing more than a crop of a 35mm frame and you'll have exactly
    the right idea
     
    Simon Stanmore, Feb 27, 2004
    #4
  5. Alan Browne

    S. Cargo Guest

    Just what everyone wants to see.. A good digital camera discussion
    in a 35mm equipment group.. No wonder this group is on it's knees
    puking blood and dying a miserable death :)

    To answer your question, it all depends if the sensor is a Bayer or
    Foveon design. There are more megapixels in the Foveon sensor which
    allows for a greater DOF.
     
    S. Cargo, Feb 27, 2004
    #5

  6. It'd be a *larger* CofC ;-). By 30 to 60%. Which sounds good, but
    it's the same effect as enlarging (and cropping) a standard film image by
    that much more. In most circumstances, it probably doesn't make too much
    difference.

    Unless I'm mistaken, you can see the same thing by applying Gaussian
    Blur in Photoshop by 0.3 to 0.6 pixels, for a direct comparison.

    As for the viewfinders, they're magnified to match the capture area,
    so you *should* be as accurate as with a film camera - which I've never
    found to be all that accurate, or at least useful in crucial applications.
    Viewfinder screens aren't sharp enough for me...


    - Al.
     
    Al Denelsbeck, Feb 27, 2004
    #6
  7. Alan Browne

    Peter Chant Guest

    Surely sensor design is nothing to do with it. Sensor or negative
    size is, as is the size of the final print as they define the size of
    the circle of confusion which is used to derive the depth of field.
     
    Peter Chant, Feb 27, 2004
    #7
  8. These statements seem to me to be confused and contradictory.

    If you take a crop from a 35mm film and enlarge it to the same print
    size you used for a full-frame print, the greater enlargement would
    result in increase in size of the circles of confusion. IOW, the useful
    DoF would be less. The same is true if you enlarge the "crop" which
    comes from a reduced size digital sensor.

    If you use the same enlargement for the print from digital, you get the
    subject at the same magnification as the full frame 35mm print, and thus
    the same sized circles of confusion, but a smaller angle of view (i.e. a
    smaller print)

    If you move further back to get approximately the same framing of the
    main subject you get an entirely different picture.

    DoF preview in camera (film or digital) gives nothing more than the
    vaguest hint of what the final picture will be like unless interpreted
    with a great deal of experience. Despite what many people think, useful
    DoF depends on the use to which the final image is put. This is
    especially true of print size, as anyone who makes 16x12 or larger
    prints knows - the bit about "people look at them from further away" is
    largely untrue.
     
    David Littlewood, Feb 27, 2004
    #8
  9. Alan Browne

    Alan Browne Guest

    er, nope. CofC for 35mm is typically 0.025mm
    2.25" is typically 0.049 ... 2x the size..

    I'm not sure if that applies...?
    That's really more what I was looking for. It would make sense that
    they do that (in DSLR's), but is it *actually* corrected.

    S. Stanmore's reply seems to say, yes...

    Cheers,
    Alan
     
    Alan Browne, Feb 27, 2004
    #9
  10. Alan Browne

    Steve Kramer Guest

    Not a problem at all. I use DOF preview for most of my shots, and find
    it just as accurate 'for me' as when I used it on my film bodies.

    Steve Kramer
    Chiang Mai, Thailand
    http://www.photoenvisions.com
     
    Steve Kramer, Feb 28, 2004
    #10

  11. Hmmmm. Another new face whose mouth has no connection to brain cells.
    This has to be at least the eighth post I've seen from Cargo with more
    blather than content. Have we got another TP?

    Depth of field is entirely a property of the lens. Period. You can
    make it have a different appearance with how much you enlarge the image
    afterward. But one thing that will make no difference at all is sensor
    density. The circle of confusion covers the same area, regardless of
    whether it has one pixel or a thousand. Film, digisensor, glass plate, or
    UV-sensitive leaves all get the same effect.


    - Al.
     
    Al Denelsbeck, Feb 28, 2004
    #11
  12. Hmmmm, we seem to be at cross-purposes here, and I have no idea where
    some of the figures above came from. So let's approach this a different
    way.

    You are referring to circle of confusion, right? Which is going to be
    the same for a DSLR as it is for a film camera, because the distance to
    focal plane is the same.

    But since the digital sensor is smaller, this is the same as cropping
    the film. To get to the point of a print sized "X" as you said (4x6 as an
    example), the digital image has to be enlarged more from the actual size of
    capture. Thus the circle of confusion is going to appear larger, i.e., more
    indistinct.

    For 35mm film, this is 4.23x magnification to get to 4x6. For the
    Digital Rebel (because I can't find specs on the D-Maxxum) at a sensor size
    of 22.7 x 15.1, this is a 6.73x. So the circle of confusion grows with
    this.

    Or are we talking about something else?


    - Al.
     
    Al Denelsbeck, Feb 28, 2004
    #12
  13. Alan Browne

    Alan Browne Guest

    exactly, and coming at it from the other direction, you are hitting the
    nail on the head of what is bothering me. And from here, we have to
    back into, 'how does it look when I press the DOF?

    BTW: for a larger format camera, a larger CofC is "allowed" as the
    blowup to specific print size ratio is less. The CofC "convention" is
    based on a certain sized print regardless of the origin film. Hence the
    CofC for MF is larger than 35mm CofoC.
    ( http://www.nikonlinks.com/unklbil/dof.htm and the statement that
    begins: "If you accept the proposition...")

    For the "35mm DSLR with a sensor that is smaller than 36x24" the manufs
    can do three things: first, do nothing (see note below "Or, as"); second
    cleverly screw with the viewfinder optics to make it look the way a
    print would look (following the CofC conventions); OR, when you press
    DOF, they just close the aperture to a different aperture than
    programmed to make the viewfinder look the way it should for the CofC
    conventions...... (if you follow that) ...at the limit cases (either
    wide open or crunched down, there would be a minor problem, but enoiugh
    for the user to notice... I just finished dinner and two glasses of
    wine, so I'll leave the numbers alone for now....

    Or, as Simon S. said, (and maybe you did too...?) since the viewfinder
    is 'cropped' to the sensor size anyway, everything is inherently correct
    when you press the DOF...

    Yes Al, we were talking about redesign of DD bra's for ladies in a 0 g
    field.

    Cheers,
    Alan
     
    Alan Browne, Feb 28, 2004
    #13
  14. Alan Browne

    Alan Browne Guest

    DOF is CofC related, CofC is a physical dimension at the film plane.
    How many pixels it covers is irrelvant.
     
    Alan Browne, Feb 28, 2004
    #14
  15. Alan Browne

    Lionel Guest

    It looks pretty healthy to me, Mr Snail.
    Go away, troll.
     
    Lionel, Feb 28, 2004
    #15
  16. Ah! We ARE talking about two different things! I can't say I've ever
    run across DOF being considered/calculated by the final print size as a
    common way of considering it (this could be just me). Since this varies
    according to usage, it's pretty much nonsense to try and develop rules for
    it - same reason why magnification is quantified "on the film".

    But I've also never run across specific figures for allowable CofC
    for any particular format. This is incredibly subjective, and it's unlikely
    I would agree with them for much of what I do. For a scenic it's far less
    of an issue than it is for a macro shot, even if you enlarge them to the
    same size. And usage/viewing distance has to be taken into account, which
    brings us to the old argument over how big an 'acceptable' print can be
    made from 35mm. Since you can do a billboard from a 35mm frame and nobody
    is likely to squawk, it's all relative.

    I see part of the concern now. I had never even begun to consider
    that the manufacturer might change the actual aperture to reflect the crop
    factor (and I doubt they do, but cannot offer any evidence either way).
    Taking THAT into the mix, then yes, the 'real world' (focal plane) CofC
    would have to be reduced in order to produce the same effect as film,
    gotcha.

    But ignoring such an assumption, CofC gets bigger with digital when
    you print the same size. So there, nyaaahhhh!

    And if you suppose that the camera is actually using a different
    aperture than claimed, then adjustments would also have to be made for the
    transmitted light change too, in either shutter speed or ISO. While this
    certainly isn't beyond possibility, I suspect it's something they wouldn't
    tackle for the relatively insignificant difference it can make to DOF.

    Now you're bouncing around and talking about the appearance of DOF in
    the viewfinder, which has exceptionally little to do with CofC calculation
    because you'd never be able to resolve it to a useful extent. But yes, for
    every model I've ever seen/heard about, the viewfinder reflects the image
    size on sensor, so DOF Preview will give an appropriate idea what to expect
    from the final image. Well, as appropriate as any other camera, which as I
    said before isn't too precise...

    Ah, but you realize there is no such thing? Gravity exists in all
    locations ;-). And why would you need a bra in micro-gravity?

    Boy, you're grumpy lately.



    - Al.
     
    Al Denelsbeck, Feb 28, 2004
    #16
  17. Depth of field preview is as subjective on a DSLR as it is on a SLR. If you
    have a preview button to close the aperture of the lens down to the selected
    fstop - which dims the viewfinder and is why we have full aperture viewing
    in the first place... The preview of what you see will be what you get
    because you are after all looking at a faximile of the image on the sensor.

    The 1.6x or whatever magnification smaller capture areas provide has no
    bearing on actual DOF only DOF calculated on the focal length of the lens
    which even that is easily calculated by anyone with a P4 desktop, 21" CRT
    screen and scanner in their gadget bag - and don't we all fall for that
    trick of too much gear!

    The magnification *factor* published by lens manufacturers is the fly in
    the ointment. Thankfully this is not a preview of the DOF but an estimate of
    it and something lens makers are doing away with anyway. This published
    estimate is not accurate when say; a lens of 100mm FL is used on a 10D Canon
    with 1.6 factor of sensor reduction which makes the 100mm lens think it's a
    160mm lens.

    We all know that the DOF for a 160mm lens is different from that of a 100mm
    Lens but this knowledge is hardly needed when you press a button and get a
    preview because all the published data is useless if, while calculating the
    new DOF, the bird flies the coup - so to speak!

    I have recently taken a keen interest in DOF due to the transition from film
    to DSLR cameras and my obsession with long, heavy, expensive lenses that
    don't like getting dumped in seawater. The cost of recovering them from such
    a disaster is close to the replacement price!

    I think now I might carry a 1000mm f2.8 lens with me at all times so I don't
    have to jump in a surf boat driven by a half crazed Australian lifesaver
    hell bent on demonstrating how high the bloody boat can get when jumping
    waves. (That's another story!)

    Douglas
     
    Douglas MacDonald, Feb 28, 2004
    #17
  18. Yep, that's all quite correct. As I wrote, a DSLR capture should be viewed
    with respect to DOF & CoC simply as a crop of 35mm.
    Yep - Given no change in focal length.
    The real usefulness of the DOF preview is in assessing the appearing of OOF
    areas when shooting macro, tele's, wide aperture portraits, etc. For this it
    works wonderfully. It's a pretty pointless feature otherwise IMHO
     
    Simon Stanmore, Feb 28, 2004
    #18
  19. On additional factor. "the same effect as enlarging (and cropping) a
    standard film (taken with a wider angle lens and cropping to the same" image
    by that much more

    We need to remember than most digital cameras use a wider angle lens
    when photographing the same subject at the same distance.

    The effect of this is to increase the aparent DOF.
     
    Joseph Meehan, Feb 28, 2004
    #19
  20. Alan Browne

    Doug Payne Guest

    The way I see it (not that my eyesight is all that great these days :)
    is that DoF preview works just fine to give a rough idea, but there are
    differences in actual depth-of-field, and they make a difference in the
    final product.

    For me it all depends on the size of prints I make. A print from my D100
    comes from a smaller area than one the same size from a 35mm file camera
    and is therefore magnified more. And the greater you magnify, the more
    small details can appear out of focus. The Zeiss-calculated CoC for film
    is .025, and for a D100 is .016 (metric diagonal of image area divided
    by 1,730).

    A good rule-of-thumb seems to be that for same lens, focus distance and
    aperture, the difference between 35mm and the D100 sensor is about one
    stop (useful for hyperfocal calculations, e.g.). Add a stop for digital.
    Other-sized sensors will of course be different.
     
    Doug Payne, Feb 28, 2004
    #20
    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.