New 20D needs lenses

Discussion in 'Digital SLR' started by Dale, Dec 26, 2004.

  1. Dale

    nick c Guest

    Musty, don't be too concerned about the 17-85 lens seemingly being a
    little on the soft side at 17mm. The lens is a very good lens. I do my
    own computer processing and I have yet to find a problem with the use of
    this lens.

    I had the 17-40MM F4L lens before I bought the 17-85 lens IS lens and
    yet I still bought the 17-85mm lens and I'm quite happy having both lenses.
    When I bought my 1D MkII, I used both the 16-35 F2.8L and the 17-40 F4L
    lenses and I found the 17-40 F4L lens to be better than the 16-35 F2.8L
    lens. I did not like the noticeable distortion of the 16-35mm F2.8L lens
    at 16mm. I had reservations about not getting a fast lens because I have
    always bought fast lenses but the 17-40mm F4L is a super lens and I no
    longer have any reservations about the lens not being a fast lens.

    nick c, Jan 3, 2005
    1. Advertisements

  2. Dale

    paul Guest

    Pardon my butting in with ignorant comments but I just realized that a
    fast lense (or ability to simulate with high ISO) gives both the ability
    to get narrow or deep DOF. Am I understanding this correctly?

    I'm still trying to grasp all the possibilities.
    paul, Jan 3, 2005
    1. Advertisements

  3. Dale

    Musty Guest

    Deep DOF uses small apertures (high f number), so this is correct. A fast
    lens has the advantage that it gives you shallower DOF due to larger
    apertures. Having a small aperture, is not any issue with most lenses I
    imagine (all you need is f/22). So in essence a "fast" lense provides a
    wider range of depth of field. Also keep in mind that DOF is a function of
    focal length so the smaller the focal length (eg 17mm vs 28mm) will give you
    larger DOF.
    Musty, Jan 3, 2005
  4. Dale

    paul Guest

    OK so a telephoto lense limits your depth of field. So if I'm shooting
    macro, I would want a short lense right? But Telephoto enlarges things
    so it counteracts that benefit in terms of DOF. So if I used a wide
    angle lense (short?) that would mean I'd have to be be 1/2 inch from my
    flower but I could have decent depth of field vesus sitting 2 feet away
    with a telephoto lense.

    Am I getting the general idea? What goes into an ideal macro lense?
    paul, Jan 3, 2005
  5. Dale

    Musty Guest

    I am not an expert on macro photography, but I dont believe you need a great
    DOF for macro shots. For example compare two scenarios:

    1) Shooting a landscape where you want to capture maybe several miles of
    distance - here you will need massive DOF

    2) Shooting a flower macro more where you want to have "acceptable" focus
    probably over an inch or two - This is not a large DOF, its a tiny DOF

    So generally macro shooting is done not with wide lenses (since you want to
    magnify and limit DOF). If you look at Canon EF lenses that are designated
    "macro" here are some examples:

    EF 180mm F3.5L macro
    EF 65mm F3.5L macro
    EF 100mm F2.8 macro
    EF 50mm f2.5 macro

    So you can see that focal length is not wide by any means (50 to 180mm). The
    main features of a macro lens would be as follows (I would think):
    - Long enough focal length to "magnify" the subject
    - Large aperture (small f) to limit DOF (you want to focus that one flower
    or insect)
    - Small focusing distance (eg the 50mm can focus as close as 23cm) which
    would give awesome magnification with 50mm focal length.

    Musty, Jan 3, 2005
  6. Dale

    Ken Guest

    Don't confuse telephoto/wide angle lenses with DOF characteristics. For a good
    explanation see -
    Ken, Jan 3, 2005
  7. Dale

    Frank ess Guest

    At macro scales, lighting can be a major consideration: unless you are
    otherwise equipped, common solutions are more easily applied with larger
    lens-to-subject distances.
    Frank ess, Jan 3, 2005
  8. Dale

    paul Guest

    Excellent explanation. However, under limited lighting, the wide angle
    should be able to use a smaller aperture (higher f-stop) and get better
    DOF. Or 'apparent DOF' I'm not sure because the distances &
    magnification are different.
    paul, Jan 3, 2005
  9. Dale

    paul Guest

    I'm not sure but the relative DOF issues are a problem with macro.
    Usually you end up only getting a 1/2-inch slice of focused flower in
    the middle.

    I don't understand what allows you to get close.
    paul, Jan 3, 2005
  10. Dale

    Jeremy Nixon Guest

    Why? If the lighting is the same, the exposure will be the same regardless
    of whether the lens is wide angle.
    Jeremy Nixon, Jan 3, 2005
  11. Dale

    paul Guest

    OK so maybe I'm not understanding. I'm just thinking that wide angle has
    a larger piece of glass so it lets in more light but I don't know.
    paul, Jan 3, 2005
  12. Dale

    Jeremy Nixon Guest

    It doesn't make a difference to your exposure settings; if you need a
    setting of 1/60 at f/8, you'll need that regardless of focal length.
    f/8 is a different size at different focal lengths, freeing you from
    needing to worry about adjusting your exposure according to focal
    Jeremy Nixon, Jan 3, 2005
  13. Dale

    paul Guest

    Maybe we are talking about something different. If I zoom in from 28mm
    to 200mm & stand back to get the same frame, I need to lower the f-stop
    from 8 to 4.6 at the same speed (minimum to hand-hold), plus now I need
    an even faster speed since I've magnified the shake.

    Or since I'm talking about framing the same shot maybe it doesn't
    matter. I just tested & I think it makes a difference. I get to use a
    higher f10 at 28mm & therefore have more depth of field. I think.

    So for macros I'd do best with a shorter lense that allows me to get a
    few inches away from the subject than a telephoto lense a foot & a half
    away. Unless there is some other factor or I'm otherwise confused.
    paul, Jan 3, 2005
  14. Dale

    John Francis Guest

    No you don't. If the object you're trying to photograph is correctly
    exposed at f/5.6 in the 200mm shot, it will be correctly exposed at f/5.6
    in the 28mm shot.

    If you're using some form of center-weighted or multi-segment metering,
    though, it's quite possible that the extra field of view of the wide
    angle lens is getting more sky in the background (even if the foreground
    objects are framed correctly), and this can make the camera suggest f/8
    for the aperture.

    Try using a spot meter (or an incident meter) instead of an averaging meter.
    John Francis, Jan 3, 2005
  15. Dale

    Jeremy Nixon Guest

    You shouldn't, if it's the same scene, especially with the same framing.
    If the background is fooling the light meter in one version vs. the other
    (because the wide angle is getting more sky in there, for instance) then
    you may actually need to compensate for that.
    Don't trust the meter... it lies.

    It's probably changing the recommended exposure due to different light
    in the background. At wide angle it may be drastically underexposing
    the shot due to backlight.
    Jeremy Nixon, Jan 3, 2005
  16. Dale

    Musty Guest

    That is not always practical. For example an insect might get scared and fly
    or run off if you are too close, so a more tele-style macro lens is handy
    when getting close to the subject may not be possible or practical (eg a
    dangerous scorpion). This is why Canon offers both a wider (but not "wide")
    and tele macro lenses. Even if you can get REALLY close, nothing beats a bit
    of tele to get very detailed close ups. Consider those insect photographs
    where they take the shot of just the eye.
    Musty, Jan 4, 2005
  17. True.

    However, if you use different lens systems, and one of them has
    _many_ more lenses than the other one, you may find _slight_
    differences in brightness due to lens transmissivity.

    Assuming a 20 lens glass, with 0.5% loss at each glass-glass or
    glass-air border, you'll have 0.995^40 == 0.818 == 81.8% of the
    light left (i.e. about 1/3rd stop less) compared to an ideal lens.

    Wolfgang Weisselberg, Jan 4, 2005
  18. Dale

    John Francis Guest

    If you're using lenses with that many elements, you'll probably
    have rather less than 0.5% loss at any surface. The most I've
    got in my personal selection is 18 elements in 16 groups, for a
    total of 34 surfaces. But some of those elements are coated,
    and I'd be surprised to find total losses as high as 15%.

    On the other end of the scale, I don't have any lenses with less
    than seven elements, although there is a lens I'd consider that
    only has five elements. So the difference is more likely to be
    at most 25 air-glass or glass-glass surfaces. I'd consider it
    astonishing if there was as much as one sixth of a stop difference,
    which is for all practical purposes almost undetectable.
    John Francis, Jan 4, 2005
  19. Hopefully! But you also get loss with each mm of glass that the
    light has pass through.
    34? I must be dumb today ... I don't get that number.

    I counter with a few more elements. Lots of _heavy_ glass.
    Don't drop it! :)
    We are talking about 1/100 loss of light on a lens -- I'd thought
    that quite realistic.
    Just for kicks:
    Canon's EF 35mm f/2 has 5 elements, the EF 70-200 IS L USM has
    23. (18 elements == 36 surfaces difference).

    If you go overboard, add a 1.4 and 2.0 converter (and better
    use a high-end body, if you want any autofocus at f/8), for
    5 and 7 additional elements, for a 30 elements (60 surfaces)
    worst case difference.
    Canon seems to feel different, at least for TV cameras:
    | F-stop aperture marks are based on a theoretical mathematical
    | formula which when transferred to different lenses can
    | sometimes be inaccurate up to a full f-stop error.

    | This inaccuracy is actually much more likely with zoom
    | lenses as they have a more complex series of different glass
    | elements that move around inside to give the different focal
    | lengths. These glass elements can combine together to decrease
    | the actual transmission factor of the light passing through the
    | lens quite considerably at different focal lengths.
    says about a lens: "also about 1/3 stop less transmission as is
    typical for zooms".

    Look at (search for
    transmission) or
    for some more info.

    And here
    an actual transmission factor is listed as (no worse than) 0.77,
    aka loosing no more than 23% of the light (on an 1000mm f/10
    *mirror* lens). That would be about 1/2 stop ...

    Wolfgang Weisselberg, Jan 4, 2005
  20. Dale

    John Francis Guest

    Consider a very simple example; two elements in one group.
    That's 3 surfaces; one air/glass, one glass/glass, and one glass/air.

    Ah, but those Canon TV cameras have an astonishing zoom range;
    something around 100:1 It's not exactly surprising that at
    some part of that range some of the light doesn't manage to
    get through all the elements, but ends up in the sidewalls.
    (It's also hard to understand how surface-to-surface losses
    could be dependent on just where the element is in the path)

    Again, that's not because of surface losses; it's because even
    those constant-aperture zooms sometimes manage to paint the
    inner walls of the lens barrel with some of the light that
    should really be passed on through the optical chain.
    Again, talking about light falloff in the corners; an artifact
    of the way the cross-section of the light beam is restricted
    by the geometry of the lens, especially at very wide apertures.
    Hard to tell, but this too could be talking about the same issue.
    It's certainly not going to be talking about surface-to-surface
    transmission losses; this is, after all, a mirror "lens". My
    guess is that what it is talking about is really falloff; with
    no aperture control, you're always using the lens at full aperture.
    John Francis, Jan 5, 2005
    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.