DOF / focus question

Discussion in 'Nikon' started by martin_z., Nov 18, 2011.

  1. Here's a shot at 1/640 @ f/4, using a D90, with Nikkor 24-70 lens. The guy seems to be in focus, while the girl is not. They were standing just next to each other, so depth of field shouldn't be an issue. The chosen aperture was wide, but not completely wide open. Any ideas what's going on here?
    Thanks for any suggestions.
    00Zcpn-416935584.jpg
     
  2. Here's a crop of the same shot.
    00Zcpp-416937584.jpg
     
  3. On my 85mm F/1.4 lens it isn't too difficult to get one eye razor and other soft, matter of fact if I'm at the close focus distance its difficult to get both eyes in focus. So yes, two people standing next to each, even at F/4, it is easier than you think to get one in focus and one slightly out of focus which is exactly what happened here. Remember that focus isn't just a function of what F/stop you are at, but also how close you are to the subject, which judging by your shot here you were pretty close. You can actually see that the back part of her collar was out of focus, and it slowly gets softer toward her face. In the future either get back farther and crop in if necessary or when shooting super close like this, stop down to at least F/8 maybe more depending how close and how far apart the faces are.
     
  4. They're standing next to each other, but the woman is obviously closer to the camera than the man, considering the way her shadow falls on him. You don't say what focal length you were using, or how far from the camera they were; 24-70mm is a pretty big range, though I'd guess you were shooting at about 50mm and they were perhaps four to five feet away. With those values, DOF should be about 4 to 6 inches. His ears look to be in reasonably good focus, so the region of sharp-enough focus probably extends about from his nose to his ears. Her face, then, is too close and consequently appears soft. Note that the hair sticking out from behind her ear looks sharp.
     
  5. As you can see the girl is standing up front a bit, casting shadow on the man's face. So, their front of faces are not in exactly the same focal plane. I would guess the difference in planes is about the length of the girl's nose.
    Also the gfirl's face is cleamy, perhaps a ton of powder smothered on her face. while the man is unshaved and with freckels. This magnifies the visible difference effect.
     
  6. Just for the heck of it, you might want to shoot a brick wall or newspaper tacked to a wall using the same f stop and focal length. It's unlikely but possible that one of the elements in your lens could be misaligned.
     
  7. Your exif info shows 58mm focal length at f4, so a dof of maybe 6" at four feet. Just not enough dof to cover the subject range, she is standing sideways and her head is mostly forward of his, look at how short the shadow of his nose is on his cheek.
    A DOF calculator shows 2.52 inches at 4 feet, 6 inches at 6 feet. At 6 feet, f8 would have given you a whole 12 inches. http://www.dofmaster.com/dofjs.html
     
  8. When I look at the exif data, parts of it are messed up. Things that CAN'T be right.
     
  9. Like what?
    This is what I get:
    EXIF IFD0

    Camera Make = NIKON CORPORATION
    Camera Model = NIKON D90
    Picture Orientation = normal (1)
    X-Resolution = 3000000/10000 ===> 300
    Y-Resolution = 3000000/10000 ===> 300
    X/Y-Resolution Unit = inch (2)
    Software / Firmware Version = Adobe Photoshop Elements 6.0 Macintosh
    Last Modified Date/Time = 2011:11:19 00:15:39
    Artist = Marty
    Y/Cb/Cr Positioning (Subsampling) = co-sited / datum point (2)
    Copyright Owner = 2009 MFitzPhoto

    EXIF Sub IFD

    Exposure Time (1 / Shutter Speed) = 10/6400 second ===> 1/640 second ===> 0.00156 second
    Lens F-Number / F-Stop = 40/10 ===> ƒ/4
    Exposure Program = normal program (2)
    ISO Speed Ratings = 200
    EXIF Version = 0221
    Original Date/Time = 2011:11:18 16:11:56
    Digitization Date/Time = 2011:11:18 16:11:56
    Components Configuration = 0x01,0x02,0x03,0x00 / YCbCr
    Compressed Bits per Pixel = 4/1 ===> 4
    Exposure Bias (EV) = 0/6 ===> 0
    Max Aperture Value (APEX) = 30/10 ===> 3
    Max Aperture = ƒ/2.83
    Metering Mode = center weighted average (2)
    Light Source / White Balance = unknown (0)
    Flash = Flash did not fire
    Focal Length = 580/10 mm ===> 58 mm
    Last Modified Subsecond Time = 00
    Original Subsecond Time = 00
    Digitized Subsecond Time = 00
    FlashPix Version = 0100
    Colour Space = sRGB (1)
    Image Width = 700 pixels
    Image Height = 465 pixels
    Image Sensing Method = one-chip color area sensor (2)
    Image Source = digital still camera (DSC)
    Scene Type = directly photographed image
    Colour Filter Array (CFA) Geometric Pattern = 0x00,0x02,0x00,0x02,0x01,0x02,0x00,0x01
    Custom Rendered = normal process (0)
    Exposure Mode = auto exposure (0)
    White Balance = auto (0)
    Digital Zoom Ratio = 1/1 ===> 1
    Focal Length in 35mm Film = 87
    Scene Capture Type = standard (0)
    Gain Control = n/a (0)
    Contrast = normal (0)
    Saturation = normal (0)
    Sharpness = normal (0)
    Subject Distance Range = unknown (0)
     
  10. Thanks, y'all. I appreciate the info. John, what about the exif is messed up?
    Here's another crop, of a different shot. This was the first shot that made me wonder about the DOF / focus issue on this lens. This one was at 1/2000, f / 7.1. At that aperture, I would expect both kids to be in focus.
    00ZcqT-416951584.jpg
     
  11. Yes, the kid on camera right is a little farther away than the one on camera left, but still...
     
  12. 90mm focal length gives you even less dof than 58mm.
     
  13. Bob, you're correct. Thanks. I guess I should just be a little more careful, and use smaller apertures than I've been using in close-up portraits of more than one person.
     
  14. That's why I prefer older manual focus lenses, they have dof scales right on top, and all manual focus film cameras have ground glass style viewing screens that show when things are out of focus.
     
  15. Really, it's all about learning about Depth of Field and, if need be, using a DOF calculator.
    The shot of the two boys has the same issue as your first image. The boy on the right is sitting behind the boy on the left...enough that you don't have adequate depth of field. (not to mention that the boy on the left appears slightly out of focus to begin with)
    I don't know the distance at which this shot was taken, but my guess is somewhere around 6 feet from the subjects. At this distance you only have four inches of DOF. At ten feet you have a whole foot of DOF. And at 12 feet you have 1.5 feet of DOF....and that's probably enough to get the shot you wanted.
    When in doubt, take a step back, especially when using longer focal lengths.
    RS
     
  16. A lot of my AF lenses have DoF scales too, special the better ones.
     
  17. I'm going to throw a radical statement out for consideration: "There's no such thing as Depth-of-Field!"
    To qualify that; lenses (especially good lenses) will show only one plane in sharp focus while everything else gets steadily more blurred on either side of that plane. The greater you magnify the picture, or the closer you look, the more that becomes apparent.
    Depth-of-field is just an invented concept to allow us to roughly judge how much of a scene will be acceptably sharp. It's not like a sudden transition from sharp to blurred, with everything inside the d-o-f "boundaries" being pin sharp and everything else being totally unsharp. If you pixel peep at 100%, almost nothing will appear to be sharp except the very plane that the lens was focused on.
    Depth of field tables are usually based on a smallish (say 10"x12") print being viewed at about arm's length by someone with averagely good eyesight. When you pixel peep the standard numbers used to calculate depth of field no longer apply.
     
  18. It's just a case of insufficient DOF. That's why I almost never shot anything wider than f/5.6.
     
  19. In such situations it is best to focus on the nearer person. Depth of field is greater behind the plane of focus.
     
  20. Mukul, you`re right, but it should be specified that the smaller the magnification, the greater DoF behind the plane of focus (and the higher magnification, the most symmetrical DoF in front and rear that plane).
    If the pic shown has not been cropped so much, I`d say there is a very little difference.
     
  21. Martin -- does your camera have a DoF preview button?
    It can often prevent this kind of thing. Especially when you seem to have shutter speed to burn.
     
  22. Rodeo - Parts of your post are not radical at all...but...
    Depth of field does exist. It's the area in front of and behind the plane of focus that is "sharp enough" to be considered "in focus" Is depth of field 100% perfectly measurable? No. Is the human eye consistent enough to be able to determine mathematically an approximation of what the area around the plane of focus will be sharp enough to the average person? YES!
    Also consider that the concept of depth of field, (along with understanding how it works), is important to the OP since you can see the problems in his photos without pixel peeping.
    Keep in mind that my statements aren't just simple rhetoric about whether or not depth of field exists... They're statements based upon my own photography and use of large aperture telephoto lenses at close distances. I use DOF tables for prints as large as 20"x30" and as small as 4"x6". DOF tables work equally well for large prints as they do for small-ish 8x10 prints. Even if I were printing billboard sized prints, my DOF scale would work. Why? Because they are not intended to be viewed at the same distances. An 8"x10" print should be viewed from inside 5 feet. A billboard, (depending on the size) should be viewed from only as close as probably 25-50 feet. (Again...it depends on the size of the billboard.)
    Additionally, you will notice that what is and what isn't "acceptably sharp" differs between print mediums as well as different size prints.
    Enough ranting...I've said my piece.
    RS
     

Share This Page