reflected light vs incident light metering

Discussion in '35mm Cameras' started by Joseph Meehan, Jul 12, 2004.

  1. It is suggesting that with the metering equipment available and used in
    their sample produced results they liked better using reflected metering
    than incident. I am not surprised.

    Depending on your knowledge, the equipment you are using and the
    subject, either may be better.
    Joseph Meehan, Jul 12, 2004
    1. Advertisements

  2. Joseph Meehan

    sreenath Guest

    Hi All,

    I was reading Kodak Color handbook (a very old collection of booklets
    from Kodak, possibly late 60's), where there is a long treatment of
    various techniques.

    There is a statement in that book that surprised me:

    "Field studies have shown that pictures made using reflected metering
    appear to be more pleasant than those made with incident light

    I may not be repeating exact words, but this is the idea.

    How is it that pictures made using reflected light metering are "more

    sreenath, Jul 12, 2004
    1. Advertisements

  3. Joseph Meehan

    Alan Browne Guest

    Since there are a variety of approaches, techniques and devices
    that do both, the generalization is a bit odd and probably
    related to some marketing effort rather than any truth ... which
    would be subjective in any case and would, incidently, reflect
    the times (p.i.)
    They can be a damn sight less pleasant when badly done!
    Alan Browne, Jul 12, 2004
  4. Joseph Meehan

    Gordon Moat Guest

    Considering that it was an old book, the views at that time may have
    been very different. Hand held light meters are much better now than
    they were in the distant past. A modern comparison of Incident .vs.
    Reflected metering is at:

    <> The images show the
    comparison nicely.

    <> Coming Soon!
    Gordon Moat, Jul 12, 2004
  5. Joseph Meehan

    George Guest

    It doesn't surprise me in the least (let the flames begin). If you think
    about it, incident metering makes little
    sense in either vision or photography (another cue for dissenting opinions).
    What gives an object color, shape, and texture is the light that it REFLECTS
    back to your eye or your film/digital sensor. Take this to the extreme, and
    consider that you want to photograph a black hole (so dense, even light
    doesn't escape...remember that light behave both like a particle and a
    wave) REALLY DOESN'T matter how much light falls on it, none is coming
    back so it just appears to be a black area with no shape, color, or texture.
    I'm sorry, but the human eye and photographic processes ARE NOT sensitive to
    how much light FALLS on an object, only on how much comes back to the sensor
    (eye, digital sensor, film, etc.).

    So, careful (so as not to be deceived by a non-representative area)
    reflected metering yields results more like what you see and attracted you
    to the subject in the first place. Incident metering can yield results
    unlike what you are seeing so if the scene is unremarkable, you might prefer
    those results.

    George, Jul 12, 2004
  6. Joseph Meehan

    Alan Browne Guest

    George wrote:

    You're correct that a sensor can only detect what is being thrown
    at it; however for a given scene if a given amount of light falls
    on it, an incident reading of that light indicates what will be
    detected by the recording sensor (at that setting). Some objects
    may reflect an amount of light that falls below (or above) the
    latitude of the sensor, and they will be unrecorded or burned
    out... but that's an issue with film (sensor) not the meter.
    That is a very good reason to use incident. It cannot be
    deceived by the non-representative subject or metering object.

    The meter in the camera (exception: Nikon F5, D70 RGB metering,
    not sure about D100 and the D2/D1x, ) is calibrated for white
    light reflecting off of an 18% grey surface. Hence, what you
    point the meter at is affecting the accuracy of the reading. You
    can always bring along a grey card but that is no less
    inconvenient than an incident meter.

    The popularity of reflected metering (in-camera) is mainly due to
    convenience (and possibly cost)... It is always more convenient
    to meter through the viewfinder, and reduces the equipment count
    and load. However, there aren't always representative objects to
    meter which means some judgement is required to get the desired

    Both are of course useful ways to meter, if well understood, but
    the most consistent results come from incident metering, not
    in-camera spot or weighted metering. An exception here is if
    using the in-camera meter to make a measurement and determine the
    correct setting, that that setting (or reciprocals of it) be used
    in manual mode for all subsequent shots in the same lighting and
    close to same orientation.

    Alan Browne, Jul 12, 2004
  7. In theory, incident metering eliminates the problem of how to photograph
    snow, or grass or shadow etc. Your meter will never be off because of the
    subject and each part of the subject will be recorded on the film in at the
    expected density.

    The theory sort of falls apart due to the limits of the film. It can't
    always record all light levels.

    Reflective averaging makes sure the average density of the subject is
    grey, which is usually right, but not always.

    Now if you use the zone system you get around this, and can use it to
    your creative advantage, if you have the skill and time.

    In the real world, there is no best, only what works.
    Joseph Meehan, Jul 12, 2004
  8. I mildly disagree with that page; not a big deal to me, but it _is_ written
    by a company promoting its incident meters.

    My very humble opinion is that no meter readings should be followed
    mindlessly. If one is using a reflected meter, one must think about where
    the gray should fall in the scheme of the values present in the scene. I
    think reading various items in a scene with a reflected meter gives me
    better control over which item I place as my main 'correct' reading. It may
    be that I would want a black plate black, but I may not want a shadow which
    would be rendered 'correctly' black to _be_ black in the final exposure. (I
    shoot slides, so my among my concerns are the limited lattitude of the film
    and blown out highlights.) I think the test shot on that page is rigged.
    :-> It doesn't reflect the things _I_ take pictures of, incidentally.

    "Mercilessly recording all things as medium gray," as that Web page says is
    really the photographers' "Mindlessly reading all things a medium gray,"
    not the meter's fault. Mindlessly following an incident meter will not
    guarantee that the result is what I _want_.

    I've read the original post and its quote, and I have no clue what the
    context was for that quote. I also question the source's use of "field
    tests" or whatever the phrase was. I remember Kodak surveying photographers
    for years about what they wanted, and the replies always were 'accurate
    colors.' Then Fuji showed people images and asked which they
    preferred. Accurate colors lost every time to the "bright golden haze on
    the meadow," as Rogers and Hammerstein wrote back in the 50s.

    I may not even want the accurate colors the incident meter mindlessly
    promises. And the "highlight and shadow areas [that] will fall naturally into
    place" may be outside my film's latitude.

    Shrug. People get to make their choices, and I have no idea what tests
    showed that reflected readings were more pleasing (I suspect the tests were
    skewed, but who knows), but using any meter without considering the
    range of readings, the reflectivity of the items in the shot, and what the
    photographer wants the subject of the image to be drives me to post my
    mild rant.
    Phil Stripling, Jul 12, 2004
  9. Joseph Meehan

    Loiskelly1 Guest

    George wrote-
    OK, let's consider it. If an EV of 8 falls on the subject, an incident meter
    would call for some combination of exposure equivalent to f2.0 at 1/60. A
    reflected light meter would call for the maximum exposure possible. In both
    cases the photo would record the maximum black that the medium allows.
    Now if we were to photograph a pair of black slacks, rather than a black hole,
    things would be quite different, as well as much easier. Under the same
    lighting conditions as above, the incident meter would still call for the same
    exposure, and the pants would be rendered as black. A reflected light meter
    would, however, call for a much higher setting, rendering the pants
    Loiskelly1, Jul 13, 2004
  10. Because at the time, Kodak was trying to sell new cameras
    with built-in reflected light meters.

    I'll bet you could easily do studies to prove that no experts
    can tell the difference between photos made with reflected versus
    incident metering. Think about it: can you tell the difference
    two photos, shot of identical subjects in identical open shade
    lighting conditions, both at f/5.6 and 1/125, where one
    photographer used incident metering and the other reflected?
    Neither could I.

    There are many situations where incident and reflected metering
    will cause the meters to recommend exactly the same exposure.
    In situations of extremely white or dark subject material,
    the meters may disagree, but an intelligent photographer will
    compensate the readings to come up with very nearly the same
    camera settings either way. Neither type of meter should
    be followed blindly, but if you're going to follow a meter
    blindly, it's usually better to follow an incident meter.

    Virtually all studio photography is done with incident
    metering. I've never seen an astrophotographer use an
    incident meter. Maybe that study polled people who
    thought comets and nebulae were more pleasant than
    Playboy Centerfolds.

    Richard Cochran, Jul 13, 2004
  11. Joseph Meehan

    George Guest

    Incident metering wouldn't detect the light that will be detected by the
    recording sensor UNLESS
    the recording sensor is located at the subject pointed toward the camera
    position. Correct? And
    locating your camera at the subject and pointing away from it would be quite
    a different picture.
    George, Jul 14, 2004
  12. Joseph Meehan

    Alan Browne Guest

    Not sure what you're getting at, but perhaps my writing above is
    not clear.

    The incident meter sensor is used at the subject position (or at
    least in the same field of light, eg, if the subject is on the
    far side of the park, then measure where you are if in the same
    light (account for local reflections too) pointing the dome at
    the camera).

    The in-camera sensor reads the light reflected as if it were off
    of a standard reflector (18% grey card) from the subject postion
    (or at least whatever is being metered). So if your subject
    across the field is big enough and 18% ish grey enough, you'll
    get a good reading. Otherwise you need to compensate for the
    non-18%-grey of the in camera meter.

    This compensation is simply not needed with an incident metering
    of the light field. In the simplest use, just point the meter
    dome at the camera and that reading will be correct.
    Yes, becasue the purpose of the in camera meter is "reflected"
    metering, and the purpose of the incident meter is (funilly
    enough) incident light metering.

    Put a clean, white styrofoam coffee cup over the lens from the
    subject position and point it at the camera position. That
    reading will be correct... the cup acts similarly to the white
    dome on an incident meter. Try that v. a grey card reading and
    you will be quite close (1/2 stop or less).

    Alan Browne, Jul 14, 2004
  13. Joseph Meehan

    sreenath Guest

    I am the original poster of this message. I have with me the book and
    here are the actual sentences stated in the book.

    The book is Kodak Color handbook - Materials, Processes Techniques.
    Printed in 1950.(Not 60s, as I had wrongly stated in original post.
    Sorry about that)
    This is a collestion of a few smaller booklets.

    This booklet is named : Color photography, outdoors

    Page no 53-54. The section has heading
    "Precautions in the Incident-Light Method"

    "The incident-light meter measurement disregards the effect of
    distance on shadow detail, or more accurately, the relative extent,
    importance, and darkness of shadow detail. In general, the smaller and
    darker shadow details in distant views are not resolved by eye or
    camera and merge with lighter areas. When the subject is close, more
    details and darker shadow areas are resolved. The practical conclusion
    is that the exposure indicated by incident-light measurements for
    distant scenes or those without shadows should be reduced.

    "Field tests of incident- versus reflected measurements for color-film
    exposure in diffused daylight, particularly on overcast days have
    indicated that reflected-light results are more pleasing. The
    incident-light technique is, however, extremely convinient for
    measurement of lighting contrast in various parts in the scene.

    "there is also question of pointing the incident-light receiver in
    right direction. More care appears to be needed in this matter with
    flat cells than with convex ones.........

    "Although good results can be obtained with incident-light
    measurements outdoors, it is questionable whether the results are any
    better than those obtained with reflected-light measurements in
    equally experienced hands"
    Perhaps this better qualifies the statement contained in the original

    Thanks for all the discussion,
    sreenath, Jul 14, 2004
  14. Joseph Meehan

    Alan Browne Guest

    This last para seems to be in the marketese of the time: note
    the "in equally experienced hands." statement.

    Yes. It was BS then, and it is BS now. (That's not a reflection
    on you, but on Kodak or whoever really wrote it.)

    Reflected metering systems (in camera) have gone through a lot of
    changes since the text was written (late 60's you say), for the
    better ... and yet are still not as reliable as a properly taken
    incident reading....unless all your subjects happen to be 18% grey.

    Alan Browne, Jul 14, 2004
  15. Joseph Meehan

    Barney Guest

    If the plate is black and a shadow is black, how do you get them to be
    different on the same shot when shooting slide film? Or are you talking
    about taking two different pictures of different subjects? If so, then I
    think you would over expose by a certain amount to make black a little
    lighter regardless of the type of meter one uses.


    No, however it does make a good point with a setup such as that AND if
    someone really did not want to think about how to expose correctly using a
    reflective type metering system.

    Barney, Jul 15, 2004
  16. Joseph Meehan

    Gordon Moat Guest

    I think Phil has used a spot meter a few times. If one understands the Zone
    System, or similar methods, and takes several spot readings of a scene, then
    a greater degree of accuracy can be obtained. However, all variables need to
    be considered, all the way through to the final print. A spot meter is also
    one of the quickest ways to ruin lots of film, since it takes more
    understanding and experience to use one properly. A spot meter is essentially
    a reflected light meter, but with a very tightly controlled measuring area.
    Any metering system just provides a starting point for exposures. Knowledge
    can dictate that certain scenes or subject might be better off with slight
    underexposure or slight overexposure. Most in camera meters (reflective) are
    quite good for the majority of photography situations, though there are times
    they can be wrong for a situation. A handheld incident meter is slower, and
    measures the light falling upon a scene or subject. While it might seem that
    a handheld meter could be more accurate, that is not always the situation,
    and again it is important to know when to deviate from the reading.

    I think a problem Phil had with the link I provided is that it came from
    Sekonic. I know a similar web page from Gossen, but then again, they are just
    another company. I do not know of any independent web sites that explain, or
    show images, of this comparison, or I would have provided a link. I also have
    no affiliation with either Sekonic, nor Gossen, and leave it up to the reader
    to determine those products qualities.


    Gordon Moat
    A G Studio
    <> Updated Galleries!
    Gordon Moat, Jul 15, 2004
  17. Yeah, sorry. I wasn't clear about that. Sorry to have confused you. The
    plate photo is an example, but a 'black' shadow is another example of
    having black but not wanting detail lost.
    Uh, not the smirk preceding the pun on reflecting the things I take
    pictures of. Incidentally, reflect ... pun.
    Exactly what I want people to avoid. I realize, however, that my point of
    view expresses only my point of view and not the view that people see out
    their windows. _Many_ times a point and shoot is all that's required, with
    no human thought at all. I have no objection to people mindlessly snapping
    pictures to their hearts' content. I mildly object to a blanket statement
    about one kind of metering being always superior to another.
    Phil Stripling, Jul 16, 2004
  18. I think a problem Phil had with the link I provided is that it came from
    Not a particular manufacturer, but _any_ maker of incident meters. I think
    any maker of either type of meter will post 'tests' which show their meters
    are better. My view is that no matter which type of meter I use, I'm better
    off if I think about what I'm trying to photograph and set the exposure
    accordingly. _Often_ a little thought leads to the conclusion that the
    meter's right.
    Phil Stripling, Jul 16, 2004
    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.