M9 - Is this true?

Discussion in 'Leica and Rangefinders' started by rory_rege, Feb 2, 2010.

  1. Michael Reichmann's recent essay on Leica cameras contains the following statement:​
    For decades many rangefinder camera photographers have gotten by with hyperfocal focusing. In other words, pre-set the camera to a distance and aperture that provides depth of field sufficient for the type of shooting that one is doing. That's what the DOF lines on the lens are for – right? Then, the camera can be used almost as a point and shoot, with no need to be concerned about focusing.
    But with the advent of digital this has become problematic. High resolution sensors require a considerably smaller circle of confusion than does film, especially when larger prints are made. In my experience one needs to stop down at least two stops beyond what is indicated on the engraved lines of a Leica M lens for optimum sharpness if hyperfocal focusing is used with an M9.​
    Re the last sentence, have owners of M9s, or presumably M8s, found the same thing? If so, I would expect that the same is true if one is focusing on a plane rather than for a zone. My one reservation about buying this camera is the apparent performance at higher ISOs, and if Reichmann's statement is correct, it complicates that question.
    Reichmann's full essay is here: http://www.luminous-landscape.com/essays/leica-open-letter.shtml
     
  2. Even with film, stopping down one extra stop beyond the one indicated by the DOF scale seems necessary with today's ultra-sharp films and lenses. With most digital cameras, I don't think the resolution is so high, even at 10 or 12MP, to impose a more exacting requirement compared to film. However, with the M9, at some 18MP, Reichmann may very well have a point--but only if the camera is on a tripod, and only at optimum aperture! The loss in performance caused by diffraction when stopping down too much could offset any potential advantage from increasing the DOF. Certainly by f/8, if not f/5.6 or f/4 (depending on the lens) maximum definition has been reached.
     
  3. The other thing that concerns me, if this is true, is that it would seem to compromise the usability of Leica lenses at wider apertures in circumstances where one wants limited apparent depth of field.
     
  4. Reichman's definition of hyperfocal focusing is wrong; what he describes is zone focusing.
    If two prints are going to be viewed from the same distance, then the larger one will have less DOF than the smaller one if shot at the same aperture, so you would need to stop down more to have the same DOF in the larger print. However, this isn't a function of the resolution of the capture medium. If it were, you would have to stop down more when using slow, fine-grained film than when using faster, coarser-grained film in order to have the same DOF.
     
  5. Depth of field indicators were often based on acceptable sharpness in a resulting 8 x 10 print. Leica, I believe, actually were/are more stringent than this. Nothing has changed really in the transition to digital - except that now people are used to pixel peeping at 100%, so imperfect focus can be noticed more easily (which I think is what Reichmann is referring to). You still have to decide what gives an acceptable-to-you depth of focus.
     
  6. The depth of field markers were optimistic for film cameras, and even more so for digital as they are based on the negatives being enlarged much less than is currently the case.
    Depth of field is a bit of a misnomer, it's really depth of acceptable sharpness.
     
  7. "the larger one will have less DOF"​
    Mike, the DOF hasn't changed.
    I think you mean the "perceived" depth of field at the fringes of the focus range anticipated. At higher magnification, the fringe focus becomes more noticeable, whether print or computer screen.
    I feel that Reichmann's comments are to point out that high resolution film is always in the camera, i.e. 18MP sensor. Because of the convenience of the medium, the possibility of enlarging any given image is greatly enhanced.
    If it's important to have true/sharp range of focus (DOF), then I'd use the hyper-focal lines just one in from the taking aperture. Certainly two in is a safer bet.
    Rory to answer your question, no it isn't true, not the way I feel you're taking it...
    00VfRk-216733684.JPG
     
  8. Gus, all DOF is perceived DOF--it's a concept about perception. Any time we assign specific numbers to DOF (such as on a DOF scale), those numbers are based on properties of the lens (aperture, focal length, and focal distance) and assumptions about viewing the final image (size of the image, distance from which it's being viewed, and how sharp is "acceptably" sharp). If you change any of those things, you change the DOF.
     
  9. Completely agree with Mike Dixon's statement above - DOF is a perception based on "acceptable sharpness". Only one plane is exactly "in focus"; in front and behind that plane "sharpness" gradually and continuously decreases. Acceptable sharpness is based on a assumed diameter for the "circle of confusion" and DOF tables and markings are based on that number; different numbers are being used, like 1/30mm or 1/length of format diagonal. Read somewhere that Leica uses a smaller COC of around 1/25mm - but have no confirmation for that.
    Based on my experience, even if you miss focus by just a tiny bit (well within DOF though), it shows in images were pin-pointing focus is crucial (portraits for example) - and it shows the more the larger you magnify the image.
    Long time ago Merklinger published a series of article suggesting an approach different from "hyperfocal focusing" - makes for interesting reading.
     
  10. Frankly from exprience on elmarit28 and cron 35 the DOF marks are badly optimistic for film, you need to close at least one F stop more than what the marks on the lens show. So if focus is more critical with senor this is becoming nasty!
     
  11. "Acceptable sharpness" is a good description, but actually what we are considering with DOF is "acceptable out of focus". The DOF scales are too optimistic in my mind and closing down at least one diaphgram stop is necessary and two is better. Problem: This can often mean stopping down into the region of negative diffraction effect (of the diaphgram blades).
     
  12. He's wrong to attribute it to digital. Dof is dof, and one can choose one's coc as desired.

    Whether dof scales on lenses are good enough depends on the lens. But pixel peeping an m9 image or a scanned film image should show the same results on the same lens, as far as dof is concerned.
     
  13. In the literal sense he is completely wrong. The DOF produced from, say, a 35mm lens - mounted to a film or digital camera - is identical to that of one affixed to a MP, M8, M9, Canon DSLR, a Hasselblad, or whatever.

    However, this is what I believe he is referring to - since a digital film plane is perfectly flat, a particular lenses propensity to front or back focus becomes more prevalent. In this case, assuming the lens has not been adjusted or corrected to account for this, then closing the aperture a stop or two will provide for some additional forgiveness.

    But the fact is, no matter the case, when you're zone focusing you are at the mercy of your "best guess."
     
  14. From wikipedia it seems to me that CoC should be smaller for most digital, as it is effectively a measure of acceptable focus, factored for enlargement. Therefore, it will be considerably smaller for 35mm than MF, as the enlargement required for the same print is smaller by a factor of two or so. Essentially, if a point is focussed such that it's fuzzy outside circle is 0.01mm wide on the plane to make this visible on an enlargement you might need to go up by an enlargement factor of 10x, which will give you a 3.5m wide 35mm print, but a 7m wide 6x7 MF print. If however you went for the same print size, the fuzziness would be visible on a 3.5m 35mm print, but not on a 3.5mm MF print. Given that most digital sensors are smaller than the size of 35mm film, they should generally have a smaller CoC, as fuzziness will invariably cover a greater percentage of their capture area (lenses focussing ability being equal). This is what I assume drives the compact camera markets habit of having such small, long focussing lenses, to maximise DoF, and minimise the effect of a tiny CoC. However, I think the M9 is a full frame sensor? If so this doesn't really apply...
     
  15. Wrong point1:" hyperfocal focusing. In other words, pre-set the camera to a distance and aperture that provides depth of field "--that is not hyperfocusing. Hyperfocusing must involve infinity, othewise it is zone focusing, not hyperfocusing.
    Wrong point2: High resolution sensors require a considerably smaller circle of confusion than does film.
    Given the same amount of enlargement from same size sensor/film, COC has nothing to do with film/sensor resolution
     
  16. Front focus or back focus means little when you're shooting real-world 3d objects. Of course, the focal plane will be a little behind or in front of your intended focus point, but something will still be sharp, and the zone of acceptable sharpness is still the same.
     
  17. The Circle of Confusion depends on the print size and visual accuity, etc. So, yes, it has changed with digital photography.
    http://www.luminous-landscape.com/essays/Leica-M8-Perspective.shtml
    .
     
  18. The circles of confusion, for full frame digital and 35MM, are identical 0.03 MM.
     
  19. When I first started using a 35mm SLR in 1972 I remember being told not to trust totally the DOF scale. "Always at least one stop in", ie. set on f8 when stopped to f11 etc. Also if the distant background is important then focus should be set to infinity even if you lose some foreground sharpness. At that time album size prints were only 3.5 X 5.5 in. so if you never made enlargements the regular DOF scale was fine.
     
  20. CoC is computed based on std viewing distance of an 8x10 print; obviously if one prefers to view closer or larger then the CoC should be adjusted accordingly. But CoC doesn't change just because of digital. You can scan film at 4000 dpi and peep if you want.
     
  21. stb

    stb

    If it were, you would have to stop down more when using slow, fine-grained film than when using faster, coarser-grained film in order to have the same DOF.​
    Mike, that's exactly what happens. The circle of confusion is smaller with TMX than with Tri-X. The enlargement factor works exactly as you describe too. So the resulting DOF on a print is a combination of lens characteristics, subject distance, sensor of fim resolution and enlarging factor. All in all, almost impossible to accurately predict. The general rule of thumb of one extra stop seems to work OK for me. On my Mamiya it is more like 2 extra stops.
     
  22. The only part i can agree with is MR's statement "photographers have gotten by with hyperfocal focusing".
    "Gotten by" as in "better than not really focused" because they never really learned to F-O-C-U-S their cameras quickly and well. Practice focusing and you will get better. Focusing is almost always better than not.
     
  23. Hyperfocusing must involve infinity, othewise it is zone focusing, not hyperfocusing.​
    That is wrong. Hyperfocal is merely the focus point on the lens that provides the maximum depth of field. If you are to set the hyperfocal going by the dof scale on the lens, then you don't infinity focus. Every aperture has a hyperfocal point, it's just that dof scales on lenses are outdated and the hyperfocal point will usually be found far away from that which the lens indicates. The scale was developed in the '20's, and now we have much better emulsions, better lenses, and most people don't even make proper prints with enlargers anymore.
     
  24. Not according to Wikipedia:
    Hyperfocal distance
    From Wikipedia, the free encyclopedia
    In optics and photography, hyperfocal distance is a distance beyond which all objects can be brought into an "acceptable" focus. There are two commonly used definitions of hyperfocal distance, leading to values that differ only slightly:
    Definition 1: The hyperfocal distance is the closest distance at which a lens can be focused while keeping objects at infinity acceptably sharp; that is, the focus distance with the maximum depth of field. When the lens is focused at this distance, all objects at distances from half of the hyperfocal distance out to infinity will be acceptably sharp.
    Definition 2: The hyperfocal distance is the distance beyond which all objects are acceptably sharp, for a lens focused at infinity.
    (Emphasis added)
     
  25. Wikipedia??
     

Share This Page