Different depth of field for old Oly lenses on 4/3 sensor?

Discussion in 'Olympus' started by erik_dorset|1, Jul 2, 2007.

  1. '...on cropped sensors, both the focal length and depth of field increase. So on a 2x cropped sensor, a
    50mm f/1.4 effectively becomes equivalent to a 100mm f/2.8 (on a 35mm body'

    I think 'tis time for me to switch over to digital, and, as a compliment to my OM1 and small hoard of Zuiko
    glass, I have been studying the reviews of the e410. I like what I see, but was a little bit surprised when I
    came across the above comment. I knew about the focal lenth doubling, but was unaware that the depth
    of field would also double.

    Is this true? Does this mean in effect that my 24mm f2.8 would turn into a 48mm f5.6? If so, it sounds
    frightfully frustrating! I was actually hoping for a 48mm f2.8...

    Any light that might be shed on the subject would greatly interest me!
     
  2. The only thing that changes is the angle of view - But wha tthey mean is for the same angle of view If you compared the depth of field to a 100mm lens mounted on a 24x36mm format camera to a 50mm lens mounted on the e410 (which yields the same angle of view as the 100mm lens mounted on an OM-1) you'd see that e410 photo still has the depth of field as a 50mm lens mounted on an OM1.

    Why?

    Because it is still a 50mm lens. The format size change doesn't effect this.
     
  3. The depth of field of a particular lens does not change according to the format. What they are saying is the 100mm lens on a full frame camera has to be stopped down further to have equivalent depth of field to a 50mm lens on a 4/3 camera, even though they both have the same field of view. In other words, the use of shorter focal length lenses on 4/3 is providing more depth of field.
     
  4. Erik,

    Comparing the same lens on a FF vs 4/3 sensor: the focal length doesn't change. Nor does the aperture. The angle of view will narrow as Ellis says, so you will be tempted to move the camera to compensate. If you keep the camera in the same place the only change to the depth of field will be because of the increased magnification of the 4/3 sensor, so the depth of field will decrease slightly.

    If you move the camera position further away so that the subject occupies the same portion of the frame then the dof will increase, purely as a result of moving further away. However you will have changed the perspective on the subject and the amount of background it covers so you're no longer comparing the same picture. That means to my mind this particular comparison is invalid.

    The third comparison you could make is to keep both cameras in the same place but choose different focal length lenses so you got the same field of view on each. To keep them the same f/stop you would have to make the f/f lens much bigger physically, and it would then give you a shallower dof at max aperture. If these two hypothetical lenses used the same size glass elements then the f/f one would have a larger f-number but both would have an indentical dof at their respective max apertures when the (now identical) images were enlarged to the same size.

    So the answer to all these questions depends on what factors you keep the same, and which you vary, when doing your comparisons.
     
  5. "The depth of field of a particular lens does not change according to the format. What they are saying is the 100mm lens on a full frame camera has to be stopped down further to have equivalent depth of field to a 50mm lens on a 4/3 camera, even though they both have the same field of view."

    This is quite correct. It has to be stopped down exactly as far so the physical apertures are of an identical size. At maximum aperture, the dof depends on the size of the front element and the final image size; if you keep those two factors the same it makes no difference what the format is.
     
  6. But what DOES change, is the degree that the image has to be enlarged to get the same size image, and that, effectively DOES alter the depth of field.

    If the portion of the image that is used is enlarged to the same size in each format, then the depth of field will be the same in both images. But if you are enlarging a smaller part of the lens' field of view to fill an 8x10, because of the 2X crop factor, then the depth of field will change to be similar to the more telephoto lens under the 1X crop factor of a film image.
     
  7. DOF depends exclusively on focal length. That is, DOF is given by the optics, nothing more. The thing that varies is the size of the detector (either 35mm film or a CCD chip). A very good example comes from medium format Rolleiflex cameras: Usually they used 6x6cm frame format, but they also sold and adapter in order to use 35mm film. As you can imagine, the lens was always the same (it's even fixed), so DOF is exactly the same. However, when using 35mm film, it seems to behave like a short tele lens (75 or 80mm), while using the full medium format, it covers an area similar to a normal lens.
     
  8. Actually Luis, it depends upon both the focal length and aperture, AND to what extent that an image from that lens will be enlarged when viewed. The circle of confusion in depth of field calculations includes the degree of enlargement in the calculation.

    So if a 300mm lens were used on a 6x7 cm negative, and enlarged only 4x to create an 8x10 inch image, it would appear to have much greater depth of field than if the EXACT SAME 300mm lens at the EXACT SAME aperture were used on a 4/3 camera, and the image were enlarged something like 20x to produce an 8x10 inch image. Of course, a much narrower portion of the 300mm lens' field of view would be used in making that second photo. If, instead, one used a 60mm lens at that same aperture on the 4/3 camera, which had the same field of view on 4/3 as the 300mm lens had on 6x7 film, and blew that image up 20x, to create an image that covered the same field of subject matter on 4/3 as the 300mm lens did 0n 6x7, THOSE two images would have the same apparent depth of field.

    So, a 300mm lens image blown up 4x and a 60mm lens with a smaller image circle used blown up 20x and cropped such that the two lenses projected the same field of view onto the same size viewed image, at the same apertures, would show the SAME depth of field.
     
  9. Depth of field has a principle dependency on three things:

    - subject to camera distance
    - a reference image magnification
    - physical size of the lens opening

    While there are other dependencies (such as pixel pitch of a digital sensor) these are of much lesser importance.

    Note that focal length is a factor only in the last one: F/numbers are lens openings normalized to focal length for exposure evaluation purposes, they do not indicate absolute size of the lens opening, so focal length must be used in conjunction with f/number to determine the physical size of the lens opening.

    Format size has an influence on the magnification required to make a reference image, and on the focal length to achieve a particular field of view.

    ---

    So, let's look at some hard numbers presented by DOFMaster and the FoV calculator I mentioned above as examples.

    Consider a 50mm "normal" lens on 35mm format set to f/2, and set the subject distance to 10 feet.

    The Field of View presented by this lens and format is 39.6, 27.0, and 46.8 degrees in horizontal, vertical and diagonal respectively.

    The depth of field at these settings will be

    Near limit 9.33 ft
    Far limit 10.8 ft
    Total 1.45 ft

    In front of subject 0.67 ft (46%)
    Behind subject 0.78 ft (54%)

    Hyperfocal distance 136.9 ft

    given a Circle of Confusion, the magnification constant, of 0.03 mm.


    If we put the same 50mm lens on a 4/3 System camera, you can immediately see the effect on Field of View ... 20.4, 15.4, and 25.4 degrees becomes the new FoV, and setting it to f/2 and 10 foot distance we see the DoF change too:

    Depth of field
    Near limit 9.65 ft
    Far limit 10.4 ft
    Total 0.72 ft

    In front of subject 0.35 ft (48%)
    Behind subject 0.37 ft (52%)

    Hyperfocal distance 273.6 ft

    given the new Circle of confusion constant of 0.015 mm, smaller because the smaller format requires more magnification to achieve the same reference print size. Note that the total DoF has dropped to about half of what it was with the 35mm format camera.

    ---

    Considering, however, that we switch to a 25mm lens on the 4/3 System camera, which approximates the 50mm field of view on 35mm format, we obtain the following results:

    Field of View 39.6, 30.2, and 48.5 degrees in horizontal, vertical and diagonal respectively. (Notice that the 3:4 proportion format provides the same horizontal FoV but a little more vertical and diagonal FoV compared to 35mm's 2:3 format proportions.)

    Holding the same subject distance of 10 ft and f/2 lens opening nets:

    Depth of field
    Near limit 8.73 ft
    Far limit 11.7 ft
    Total 2.96 ft

    In front of subject 1.27 ft (43%)
    Behind subject 1.7 ft (57%)

    Hyperfocal distance 68.4 ft
    Circle of confusion 0.015 mm

    Overall, at the same field of view, the 4/3 System camera will return twice the total depth of field compared to the 35mm format camera.

    ---

    This leads to a pair of simple rules to keep in mind:

    1) If you hold focal length, focus distance and lens opening constant, a 4/3 System camera will net approximately HALF the depth of field and field of view of a 35mm format camera.

    2) If you hold the Field of View, focus distance and lens opening constant, a 4/3 System camera will net approximately double the Depth of Field.


    I hope that clarifies the Depth of Field and Field of View issues in a useful manner.

    best,
    Godfrey
     
  10. Just a small point f1.4 is still f1.4. Lens is still 50mm. only field of view changes to be equivlaent to 100mm on 36x24 frame
     
  11. Did nothing I wrote above make sense to you?

    I agree: focal length is still focal length and aperture is still aperture. FoV changes because
    of the format change.

    Magnification is an intrinsic part of DoF calculations, as seen in the CoC constant. As a
    consequence, DoF also changes because of the format change due to the difference in
    magnification to reference output size image.

    Godfrey
     

Share This Page