Flare question(s)

Discussion in 'B&W: Film, Paper, Chemistry' started by Michael R 1974, Jul 31, 2012.

  1. Michael R 1974

    Michael R 1974 Subscriber

    Messages:
    6,200
    Joined:
    Feb 2, 2010
    Location:
    Montreal, Ca
    Shooter:
    Multi Format
    Consider the following setup for a 35mm roll film test for tungsten speed and development time, versus contacting:

    Camera with 85mm lens aimed at a brightly lit white card target. The focal plane is about 12" from the target. The lens is focused at infinity. The card is illuminated by flood lights such that metering the card at EI 100 gives a reading of f32@1/15s, so that exposures can be made from zone I (f32@1/250s) to zone XV (f4@1s). The white card is cut to size so that it fills the frame but does not go beyond. It is surrounded by black cardboard. Further, I made a lens insert (fits into the filter threads) made of black cardboard which masks everything outside the frame (picture a black lens cap with a rectangle cut in it).

    So, is this high or low flare? Presumably the masking work reduces camera flare substantially, but I'm still unclear on whether or not the absolute intensity of the target makes a difference. For example, suppose you can expose a frame for zone I in two ways - 1) high intensity as above, f32@1/250s or 2) a target 10 stops less brightly lit, f16@1s. Assuming this is a uniform target, is there a difference in flare between those two exposures? Camera flare? Lens flare?

    And what about the same questions but assuming instead of a uniform target, each exposure is of say a reflectance grey scale tablet, or a backlit transmission wedge (ie rather than a uniform tone in each frame there are multiple adjacent tones)? Does flare play a different role in that case?
     
  2. Bill Burk

    Bill Burk Subscriber

    Messages:
    4,614
    Joined:
    Feb 9, 2010
    Shooter:
    4x5 Format
    Your test is low flare... Think of it this way... The whole picture is same tone... Say a photon for flare reasons doesn't hit the point on the film where the lens theoretically aimed it. Another photon is just as likely to randomly miss where it was supposed to go... and is likely to take the place of the first one... So since the light bounced around at random it didn't change the intensity of the test target.

    Now put a backlit transmission step wedge in the place of the flat white target and you create a high flare situation. Lots of photons from the clear patch don't hit where the lens aimed them, they bounce and land in a dark patch where you were hoping to have nearly zero light.
     
  3. Michael R 1974

    Michael R 1974 Subscriber

    Messages:
    6,200
    Joined:
    Feb 2, 2010
    Location:
    Montreal, Ca
    Shooter:
    Multi Format
    Bill, for the first situation - what about camera flare (let's assume less masking)? For those two equivalent zone I exposures, wouldn't there be more flare in the high intensity exposure versus the low intensity exposure? If the intensity of light reaching the film plane is higher, isn't there more reflection off surfaces inside the camera? Actually now I've just confused myself even further, because I'm wondering if stopping down the lens is different than reducing the exposure time - from a flare perspective.

    Regarding the transmission wedge, so this means that people testing by photographing a wedge taped to a window are getting significant flare effects? Does the intensity of the backlighting make a difference?

    What about photographing a reflection target (say a zone grey scale)?

    And another question - is my lens cap mask redundant since the target in my test is surrounded by black paper? Or does it indeed reduce camera flare even further?

    Thanks for participating. Flare is tough for me to understand (not the impact on testing, but rather how to know when there is flare and when there is not).

    Hopefully Stephen will write in also.
     
  4. Bill Burk

    Bill Burk Subscriber

    Messages:
    4,614
    Joined:
    Feb 9, 2010
    Shooter:
    4x5 Format
    I guess you have reduced flare coming from outside subject with masking and compendium hood... But that only made a naturally low-flare test even lower in flare. Your masking means changing aperture affects flare "less" because masking already ruled out the "sun hitting the camera chamber wall" kind of flare. And that kind of flare would be reduced in a real camera scene with a smaller aperture.

    But the test wedge shot in camera is still high flare. The intensity doesn't matter because time would be reduced proportionally. This might be affected by aperture changes.

    A reflection target would include "normal" flare, depending on the design and layout. One shot of a stepped grayscale could tell you the speed. But a reflection target would require multiple shots to test higher values. But you already know a one-tone target requires multiple shots.

    Many in-camera test target users are satisfied with tests that include flare. Stephen would say it is harder to understand the film response to light when you add variables into a test that could be excluded with a contact test.
     
  5. Michael R 1974

    Michael R 1974 Subscriber

    Messages:
    6,200
    Joined:
    Feb 2, 2010
    Location:
    Montreal, Ca
    Shooter:
    Multi Format
    Interesting - so a reflection target is probably a reasonable way to do things for average conditions. Most of my actual shooting conditions at night are likely low flare so a low flare test is useful. I'd like to compare the curves I get under various testing conditions. Contacting could serve as a "baseline" (which is why I started the other thread about using an enlarger to expose film).

    It's also interesting to consider the size of the lens image circle. I forgot to add the 85mm lens I was using in this particular setup is a tilt/shift. When I initially tried testing with the white card - without masking the lens, there was considerable flare inside the camera leading to significantly higher densities toward the upper and lower edges of the film, compared to completely even densities with the masking in place. This is important for testing with view cameras, and also underlines the usefulness of an adjustable compendium hood in the field.
     
  6. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    Michael,

    Like Bill said, this is definitely a low flare condition. Not only because of the careful masking, but the subject itself. 80% of veiling flare comes from the subject. A single tonal target should produce negligible flare. Another important element in a test like you described is the exposure placement on the curve.

    A flare factor of 2 means the shadow exposure is doubled. For a 125 speed film, the shadow exposure for the statistically average scene will fall around 0.0032 mcs. A stop of veiling flare will bring that up to 0.0064 mcs. Now, if you are placing the exposure at the metered exposure point, then the 0.0032 mcs barely makes a dent in the midtone exposure of 0.064. In camera Zone System type tests meter the target and then stop down. This doesn't change the camera image. All it does is shift everything down the film curve.

    Here it is in graphic form. The camera image curve in quadrant 1 represents a full range subject and not a uniform target. The camera image would be more of a dot if that were the case. The flare at the metered exposure point can quite possibly equal almost zero.

    2 Quad - Zone exposure test example.jpg

    Absolutely. Both of those situations increase the potential flare , but perhaps more importantly, you also now have exposure values that are low enough to be influenced by flare.
     
    Last edited by a moderator: Jul 31, 2012
  7. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    From Theory of the Photographic Process, "Jones and Condit found that the magnitude of the camera flare depends on the distribution of luminances in the scene and in the areas surrounding the part of the field of view being photographed. When a large part of the field of view consisted of bright clouds, sky, water, light colored sand, or white buildings, the camera flare was much greater than it was when the dark foliage, dark buildings, or other objects of low reflectance occupied most of the field."

    A high proportion of lighter values in the frame will increase flare as will the higher illuminance that comes with increased luminance ranges, but only if the shadow exposure remains at the same value. The back-lit wedge test will produce a certain degree of flare, but if the intensity of the light source is increased, all the values from the step tablet are also increased. There isn't a change in the the exposure distribution and there shouldn't be a change in the amount of flare. On the other hand, if the intensity of the background is increased when shooting a reflection gray scale while the intensity of the light illuminating the scale remains the same, flare will be increased. Just like any back-lit scene will have a higher degree of flare than a front-lit scene.

    One of the difficulties would be in determining the amount of flare present. Is it average? Too low? Too high?
     
    Last edited by a moderator: Aug 1, 2012
  8. Michael R 1974

    Michael R 1974 Subscriber

    Messages:
    6,200
    Joined:
    Feb 2, 2010
    Location:
    Montreal, Ca
    Shooter:
    Multi Format
    So there are really many variables here (possible support for the contacting scenario - as you have said before). But I've learnt something in your last post - ie in photographing a range of tones, either by transmission of reflection, the amount of veiling flare remains constant with changes in the absolute level of illumination (assuming appropriate masking). Do I have that right? I think both you and Bill have tried to get that across to me several times before, but I didn't get it until now. Quite an important point I was missing. I apologize for being slow on this stuff. To be honest I have a hard time following the graphical/quadrant representations.
     
  9. Bill Burk

    Bill Burk Subscriber

    Messages:
    4,614
    Joined:
    Feb 9, 2010
    Shooter:
    4x5 Format
    You're doing alright... And I learned from your real test with your flat target masked off -- that you improved matters by eliminating the part of "Jones and Condit ... camera flare" that "depends on the distribution of luminances in the ... areas surrounding the part of the field of view being photographed"

    The four-quadrant graphs are pretty cool really. The first is subject (lower right) perfect 45-degree ideal and superimposed flare (departs from perfect mostly in the shadows), second (going clockwise) is the film curve you are familiar with, third (not shown) is paper... flipped from normal so that you can continue the dotted lines connecting what you get and what you got... the last quadrant (also not shown) shows the ideal 45-degree again with the final result of everything superimposed.
     
  10. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    I don't know if this helps since I'm not sure what part of the graph is tripping you up. Maybe we can use this as a jumping off point?

    Two Quad - example A.jpg

    The advantage of the camera image / flare curve is that it enables you to evaluate exposure separate from but interconnected with the film curve. I believe many people tend to confuse sensitometric exposure, which defines the film curve, with the camera image, which I like to picture as being superimposed over the film curve. The multiple quadrant approach makes it easier to distinguish all the elements.

    For a given scene, the camera image doesn't change shape. It just moves up and down with changes in the camera settings. And for a given film test, the film curve is completely static.

    Pick one of the guidelines starting at the subject luminance at the top of Quadrant 1. This represents the luminance of the subject. Following the line down until it intersects with the camera image curve. The f/stop and shutter speed settings on the camera determines what the illuminance value will be. From there, based on the film test, we know where that particular point of exposure will fall on the film curve.
     
    Last edited by a moderator: Aug 2, 2012
  11. Michael R 1974

    Michael R 1974 Subscriber

    Messages:
    6,200
    Joined:
    Feb 2, 2010
    Location:
    Montreal, Ca
    Shooter:
    Multi Format
    Thanks for this. I'm going to go through it over the weekend in detail.

    I appreciate all the help and the postings you and Bill have been doing on flare, tone reproduction etc. I'm learning a lot about something I highly doubt most photographers are aware of.
     
  12. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    It's a shame there's not more interest in the subject. I think it's the key to making all the puzzle pieces fit.
     
  13. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    Here is an example of the effect different levels of flare have on the negative log-H range and negative density range. It shows a no flare, a one stop, two stop, and three stop curve camera image situation. There is no change is film processing, camera exposure, or scene luminance range. The only variable is flare. Notice how it's the one stop flare curve where the shadow falls on 0.10 over Fb+f and not the no flare curve? The no flare shadow exposure actually falls one stop lower on the film curve.

    Two Quad - Multi flare.jpg
     
    Last edited by a moderator: Aug 5, 2012
  14. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    Michael, you seem to have had a change in your opinion on the importance of flare. What is your perspective and why do you think there's a lack of interest in the subject?
     
  15. Michael R 1974

    Michael R 1974 Subscriber

    Messages:
    6,200
    Joined:
    Feb 2, 2010
    Location:
    Montreal, Ca
    Shooter:
    Multi Format
    Stephen - sorry for the delay - will write more on this in the coming days.
     
  16. andrew.roos

    andrew.roos Subscriber

    Messages:
    568
    Joined:
    Jan 4, 2011
    Location:
    Durban, Sout
    Shooter:
    35mm
    The amount of flare does not depend on the intensity of the light, but on the contrast range - not only of the scene being photographed, but potentially of everything that can cause light to fall on the front lens element. Flare is normally apparent in high intensity light simply because this enables a much wider contrast range - for example, direct sunlight falling on the lens (but outside the image area) gives very high contrast compared with deep shadows in the image itself, which means that only a tiny proportion of the direct sunlight need to be "misdirected" onto the film in order to give visible flare. You would get the same amount of flare if photographing under moonlight with the moon in the place where the sun was and the exposure adjusted to keep the image shadows in the same zone because the relative contrast between the moon and the shadows would be the same as the relative contrast between the sun and the shadows. (I'm discounting reciprocity failure here).
     
  17. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    In addition to the luminance range's effect on the degree of flare is the luminance distribution. If flare was dependent on the luminance range alone, it would be predictable. Alas, it's not. Two scenes with the same luminance range can have vastly different flare factors depending on the ratio of lighter tones to darker tones. A white card with a small black square will have a higher flare factor than a black card with a small white square. In landscape photography, the amount of sky in the frame or just outside the frame usually has the largest influence on flare.
     
  18. Bill Burk

    Bill Burk Subscriber

    Messages:
    4,614
    Joined:
    Feb 9, 2010
    Shooter:
    4x5 Format
    And aimed into the sun, without a lens hood... The rays from the sun add flare on the surfaces of the lens elemnents (and filters)... And the rays can bounce off the side of the lens barrel or body of camera... adding more or less flare depending how well these surfaces are "painted" black.
     
  19. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    The type of flare from aiming the camera towards the sun tends to be more localized creating a ghost image of the diaphragm opening, as opposed to the evenly distributed veiling flare that is ubiquitous in any optical system.
     
    Last edited by a moderator: Aug 18, 2012
  20. Bill Burk

    Bill Burk Subscriber

    Messages:
    4,614
    Joined:
    Feb 9, 2010
    Shooter:
    4x5 Format
    The flare that I had in mind can be seen with the shutter open and back open and aim the camera so the sun is focused on the interior wall or bellows of the camera. Gears/levers in the camera can reflect... Black paint is rarely perfect... Even the tube of a telephoto lens can reflect. This can cause an evenly distributed, or at least widely spread, flare - and that is the kind you can resolve with a lens hood.
     
  21. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    I agree.
     
  22. Michael R 1974

    Michael R 1974 Subscriber

    Messages:
    6,200
    Joined:
    Feb 2, 2010
    Location:
    Montreal, Ca
    Shooter:
    Multi Format
    A rectangular adjustable hood, yes. I wish there was some kind of masking hood you could buy for view cameras. Ideally it would be a hood that not only did the standard job of reducing lens flare, but also had a set of adjustable rectangular blades in it. The idea would be that once you are done applying the camera movements, you then adjust these blades to exclude everything beyond the perimeter of the ground glass, thereby effectively eliminating camera/veiling flare, which can be substantial with lenses that have a lot of coverage. I've cobbled together something sort of functional along these lines for use on the tilt/shift lenses I use on my 35mm camera, but 4x5 is another story. My understanding is John Wimberley has constructed such a device for his view cameras.

    By the way Stephen and Bill, sorry I keep meaning to post more to this thread but I'm slower than usual lately due to too many simultaneous experiments (flare/film testing, pyro etc). One thing I'm hoping to post to this thread soon is a comparative speed/development test for a given film under no-flare and flare conditions. I think this could be a useful discussion point (Stephen and Bill you already know this stuff of course, but perhaps I can generate some broader interest here).
     
  23. Stephen Benskin

    Stephen Benskin Member

    Messages:
    1,511
    Joined:
    Jan 7, 2005
    Location:
    Los Angeles
    Shooter:
    4x5 Format
    Effectively reducing extraneous camera/veiling flare. Around 80% of veiling flare comes from the subject.
     
  24. Michael R 1974

    Michael R 1974 Subscriber

    Messages:
    6,200
    Joined:
    Feb 2, 2010
    Location:
    Montreal, Ca
    Shooter:
    Multi Format
    Good point. I should have been more precise about that.