The effect of flare on film exposure and development systems
On a different thread Mr. Sandy King suggessted to another person (Kirk Keyes) that a thread on flare as it pertains to exposure and development. Mr. Keyes passed.
I find this to be a topic of much interest. I believe that exclusive of the use of filters, it is the hardest variable to predict and to control. I already am aware of the BTZS technique of making a general allowance for flare.
How do you allow for and control the flare variable in your personal system of exposure and development?
Claire, my "system" for film testing is to use the camera and lens combo I normally reach for, and do tests in camera not in the darkroom. While I know this is not as accurate as using an enlarger to control variables, my basic idea is that flare is inherent and part of my exposure inside the camera. Since film tests are the result of in-camera exposures, flare is included as a normal variable and dealt with in development.
I know there are gaps of logic here, but it works for me. As far as flare from an actual shot, I use a lens hood whenever I think it may affect the image. tim
Even though flare is something that is taken into account in film development times, I think it's better to control it at the source by shading the lens properly with a compendium hood when possible, or with the best lens shade available when it's not possible to use a compendium.
I think that we have to get used to the idea that we cannot predict everything about any photograph we are about to make. In years gone by, many of them for me, the first question asked when you announced that the pictures came in was "How did they come out?" Things have change a lot, but not when it comes to flare. A camera with ground glass and a good dark cloth is a help, but not absolute. We all know that the general effects of flare are to reduce contrast and to make a short toe film act like a long toe film, but how much? It's a good thing we who do our own have the ability to adjust what we could not predict.
Originally Posted by Claire Senft
Primarily, there is the choice of film stock and the use of flare-prevention techniques in taking. In a photograph that includes an overcast sky, there will unavoidably be an increase in flare, which will affect shadow densities most of all.
Films intended for outdoor/multipurpose use (e.g., old Super-XX) have a short toe, which tends to fight the flare to a certain extent. Film intended for studio use, where flare is much lower (e.g., old Portrait Pan, Ektapan) generally have a longer toe. Using such films outdoors under high-flare conditions tends to yield rather weak, soft shadows.
The number of films available has been severely reduced over the last 15 years or so. The variation in curve types has also been reduced.
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
It's always a good idea to control excessive flare, but 80% of flare comes from the subject itself. A white card against a black background will produce a very difference flare value than a black card against a white even though both scenes have the same luminance range.
The only way to realistically deal with flare is to use averages. For an average scene, flare is considered to be 0.34 for larger format lenses and 0.40 for smaller format lenses. The difference is mostly from the smaller number of lens elements used in large format lenses. The general rule of thumb is for a 0.10 difference in flare for each stop difference in luminance range. (2.2 is considered average luminance range). Kodak's aim CI numbers confirm the average flare values. They used to use CI 0.56 for normal and changed it in the early 90s to CI 0.58 to (I believe) better represent the predominance of 135mm users.
There are two different camps on the issue of how to calculate flare's influence when determining aim contrast indexes. One is to use the average flare value for all luminance ranges. This is known as the fixed flare method. The other approach is to use an approximate flare value for the various luminance ranges (see previous paragraph) known as the variable flare method. The equation Kodak uses is Neg DR / LSLR - Flare. The difference between the two methods becomes really obvious as you move to the more extreme luminance ranges.
Flare isn't always unwanted. Film speeds would be approximately one stop slower without it. Film processing times would also be shorter (sometimes too short). It also helps keep processing times from being too long with scene's with high luminance ranges (pushes) and too short for scenes with short luminance ranges (pulls). But what about the influence of flare on the quality of the image? Tone Reproduction theory suggests that average flare really doesn't cause too many problems. It seems that we are O.K. with compression of tones in the highlights and shadows as long as the midtones are produced slightly more contrasty than the original subject. In fact, our eyes compress the shadows normally to begin with. We not only don't mind slight compress of the extreme tones, but if the values of the print perfect match the values of the original subject in distribution and density, the average person wouldn't consider it to be a good photograph.
Mr. Beskin, I concur with your analysis that flare can be as much a help as a problem. I did get lost when you said NEG DR/ LSLR which I understood to be the density range of the negative/ ?. What is meant by LSLR. I freely admit my ignorance of what may be understood by all others.
My main question of course was not aimed at lenshoods but rather achieving with more precision and accuracy in controlling its effect its effect on the negative at he time of exposure and in the determination of the development..
Sorry I wasn't more clear. LSLR is short for Log Subject Luminance Range. BTW, the Negative Density Range can also be thought of as the paper's Log Exposure Range (LER) and it's value comes from determining the LER in a paper test. An example using the equation is:
Originally Posted by Claire Senft
1.05 / 2.2 - 0.40 = 0.58
From there, it's really easy to calculate the entire range of CIs for your conditions. Below is a table of CIs for different LSLRs using the various flare methods. The question of which is the best method depends on your perspective. No Flare is good for contacts (ie dupe negatives, etc) but not for much else. Variable flare is technically the most accurate, but because of the large difference of flare depending on the tonal distribution from any given scene, is there enough benefit using the variable flare method to justify the additional effort to calculate it? Is the fixed flare good enough for government work? Kodak seems to thing so because they use the fixed flare method. I've also added compromise method that I recommend called the practical flare method (PFM)
Dev__LSLR___No Flare___Fixed Flare - 0.40___Variable Flare____PFM
Hope this helps.
Attached is a graphic representation of the above equation, and a complete tone reproduction cycle. Notice how different the resulting values are in the tone reproduction quadrant (upper right) comnpared to the original values. This print would be considered of good quality.
Last edited by Stephen Benskin; 12-02-2007 at 12:32 AM. Click to view previous post history.
Ornello I wish I could read your attachment, the pixellation has pretty much ruined the typefaces. Maybe break the two pages into two images? FWIW I'm told a gif file is superior to a jpeg for graphic images such as this, making a smaller file. Jpegs do better on variable tones for whatever reason.
Gary, the attachments are fine if you view them full size. Shrunk to fit the window (by MS Exploder?) they are unreadable as you say.
-- Ole Tjugen, Luddite Elitist