When choosing a filter, don’t be misled by the names fog and double fog. The names of those filters are totally wrong. From the name you would think that double fog has a double effect. Double fog gives you a sharp image; all it really does is give you a glow around the light sources. Regular fog is foggier and not as sharp. Regular fog has more of a diffusion effect than a double fog.
Low-contrast filters really deal with contrast. They do not glow on the highlights. I have very little use for these filters. For some reason I have not learned to use them. I hardly ever use the traditional diffusers. I do not like to take my pictures out of focus. And what those diffusions are actually doing is throwing things out of focus.
Caleb Deschanel, ASC
I don’t like fog filters. I used LCs from time to time and there are these frost filters made by Wilson in England—they are another kind of diffusion; they basically soften the edge of highlights. They don’t affect the low end [the black] as much as low contrasts do. They sort of soften the highlights. My tendency is to not use much diffusion at all. I prefer keeping things fairly clean.
Haskell Wexler, ASC
I use low-contrast filters when there is heavy contrast on an exterior. Instead of flashing I would just throw in a #2 or #3 or #4 low-contrast filter. I use diffusion filters on people when required. I would use Mitchell diffusion A, B, and so on.
I find nets useful because you do not have to worry about the kicks on the glass. They flare differently from the fog filters. Say, on a lamp the fog would flare out, when the net would crisscross.
It is important to know the differences among the filters made by different manufacturers. I know for example that Harrison fog filters are stronger than Tiffen fog filters.
Harrison and Harrison also make a couple of useful filters not mentioned by these cinematographers: the Scenic fog filter and the Black Dot texture screen.
The Scenic fog could be described as a fog graduated filter. You can situate it to obtain a light fog effect in the foreground and a denser fog in the upper part of the frame. This would resemble a natural situation.
The Black Dot texture screen represents another kind of diffusion filter; it gives you diffusion without the milkiness and the halos around the light sources.
To soften the outer areas of the frame while keeping the center image clear, Tiffen makes the Center Spot filter. The same company offers a series of star filters that create star effects on any light glint in the scene. Actually Tiffen offers a very extensive and innovative line of filters, many of which became great favorites of the cinematographers community. The Pro-Mist filter creates a gentle halo around the highlights, while an 812 filter adds warming color to flesh tones. These two filters are also manufactured in combination, the result being a “warm” version of Pro-Mist.
One of the oldest diffusion devices is a net. Traditionally, cinematographers assemble a personal set of nets in various densities and shades, ranging from millinery veils to an assortment of ladies’ pantyhose.
Philip Lathrop, ASC
When I use black nets on the lens for female close-ups, I use them in combination with Mitchell diffusion filters. Double fog is better than the low-contrast filters because you do not get out of focus as easily. LC looks to me always out of focus. Single fog is more for a fog effect.
James Crabe, ASC
White gauze is very effective just by virtue of its being white. Some white gauzes were actually backlit for the incredible complex silent movies where the great glamour shots of Gloria Swanson needed this kind of flare. It is pretty difficult now with the zoom lenses that are used as often as they are. In old days the matte box could be extended as far as it can go, to be just off the shot and keep a lot of that spurious light from hitting the lens.
The problem of spurious backlight is the reason why black nets are preferred to white ones for the normal, not flared, net effect. Other color nets are also used.
Vilmos Zsigmond, ASC
Pantyhose is a very beautiful thing, actually. It is like using a scrim. It creates a very beautiful effect. It gives you sharpness and it gives you softness at the same time. The only disadvantage to pantyhose is that it has a funny effect on light sources, like a cross-star effect that many times you do not like to use.
It is important to remember with all these filters that the image seen through the camera is more diffused than the image that will appear on the film. More accurate evaluation is possible when looking through a combination of the viewing glass and diffusion or fog filters.
James Crabe, ASC
The ground glass, even though it is pretty accurate most of the time, does often play tricks on you. If you look through a camera that is heavily diffused, the image usually looks more heavily diffused than it will when you see the final result, so you have to be pretty confident about what you are doing. Often people will look through the camera and say, “You must be kidding, I cannot even see him,” whereas in actual fact the diffusion is less severe.
One should also remember that the effect of the same fog filter will differ depending on the focal length of the lens and on the f/stop at which the lens is set.
James Plannette, gaffer
At 25mm a fog filter does not have as much effect as it has at 250mm. At 25mm you may use a one-fourth double fog, while at 250mm you will use a one-eighth double fog and it will look the same. Therefore, if you cannot change filters during zoom (as with a sliding graduate filter), it will have much more effect at 250mm. With sliding diffusion you could theoretically slide it during the zoom-in from one-fourth to one-eighth. You really have to keep in mind the effects that filters have at different focal lengths.
At the same time, if lenses of a similar focal length are used for a close-up and then for a wider shot, the close-up should be given more diffusion than the wide shot. These are very general rules; it is essential to shoot tests for the particular circumstances.
The diffusion effect will also change with the f/stop. Stopping down decreases the effect of a diffusion filter, particularly in the range from f/8 onward. Fog filter results are similarly affected. The wider the lens opening, the more “blossoming” effect the fog filter will produce. Exposure also needs to be considered. Underexposed parts of a scene will be less affected by diffusion than overexposed areas.
Fog, double fog, and low-contrast filters do not require more exposure. On the contrary, images with lowered contrast, especially in already less-contrasty subjects, may appear as overexposed, and therefore some minor exposure reduction is sometimes needed. When using low-contrast screens, exposure reductions of one-half to one and one-half stops may sometimes be required to prevent a flare effect. A Black Dot texture screen requires approximately one stop more light to compensate for its density. A professional diffusion filter, which usually consists of a glass with a finely ground pattern laminated onto an optical glass, does not require any exposure correction. On the other hand, a homemade diffusion device like a clear glass smeared or sprayed with petroleum or glycerine, or nets like pantyhose, should be tested for an exposure compensation. The easiest exposure test consists of holding such a “filter” in front of an exposure meter. But only shooting tests on film will allow a proper evaluation of all the characteristics of such a device.
One of the most important reasons for the rather wide use of fog filters and nets is their ability to desaturate color. Cinematographers and directors are usually not interested in obtaining highly saturated “postcard” colors.
Jordan Cronenweth, ASC
Some colors need more desaturation than others, but that is a selective thing that has to be done by wardrobe, set dressing, paints used, colors that you allow the sky to be. I think that blue, almost unlike any other color on film, calls attention to itself immediately when you see it on the screen, and oftentimes to the point of distraction. I find red to be second to blue in calling attention to itself.
Robert Wise, director
To me more often than not, color is too rich, t
oo full of color, too unreal. The very nature of photography is not quite realistic. I have found that if you wanted to get something really low, really down, really degraded, you had to dress it and paint it and treat it, and knock it down. To the eye it seemed to be too much, but it was right to the camera. I very much favor, if possible, desaturating it, knocking down some of the extremes. But it depends on the film that you are doing, of course.
Working with Smoke
Working with fog filters is relatively easy when compared with real smoke or fog situations. Smoke works in depth. It affects light sources as well as the surfaces on which light falls.
Backlighting makes the smoke visible in this scene from E.T. the Extra-Terrestrial. (Allen Daviau, ASC, cinematographer)
(© Universal Pictures, a division of Universal City Studios Inc. Courtesy of MCA Publishing, a division of MCA Inc.)
Allen Daviau, ASC
The secret in learning to work with smoke is that you must get to do some tests. Reading with a spot meter, you try to find some average reading. You are lighting air. Your exposure index is based upon the reflectivity of the smoke in the air. This is what makes it very difficult even with a spot meter to read. And to be truthful, the most valuable tool is your eye, judging the density of the smoke by eye and evaluating when the smoke is ready. You say, “Smoke it up!” and the smoke starts swirling around and building up density. You are waiting for it to stop moving and finally you are ready to go. You roll the camera and it may be still moving, but finally you say, “OK,” and the director says, “Action!” just when the smoke stops moving. In my tests for E.T., I bracketed the exposure in half stops. I was higher up in my printing scale than I thought I was going to be. We printed at #40, 44, 30. This way if the smoke turned out to be less dense and therefore less reflective, there would be still enough exposure to print in the middle of the scale.
Image Manipulation in the Laboratory
The film laboratory represents the next place where the image can be manipulated through processing and timing. The print made from the negative or reversal, whether it is a projection print or an intermediate, is timed to have densities and color balances that match the aesthetic concept of the cinematographer. Today, when the negative is mostly “timed” as the digital intermediate, the photochemical timing is more applicable to film tests.
PROCESSING
Film development requires stringent quality control to assure consistency from day to day. On the request of the cameraman, a film can be forced (pushed) in development to make up for an underexposure. It means that the film will require an extended development time to produce more density in the image. Today, with fast emulsions, pushing is used less frequently for gaining additional film speed. One-stop pushing, on Eastman negatives, should have hardly any effect on the look of the film, but extensive pushing will raise the grain and deteriorate the black.
Allen Daviau, ASC
I don’t believe that pushing inherently gives a higher contrast. In fact, in many circumstances I feel that pushing film deteriorates the black. I try to avoid pushing film except in a true emergency. Particularly since we have high-speed stock. Pushing is an option, but in general I am rather a purist, I don’t like to push the film, I do like a healthy negative.
The opposite of pushing is pulling the film in the lab. Here an overexposed negative is underdeveloped (developed for a shorter time), which results in a lower contrast and weaker blacks.
TIMING AND PRINTING
The widest range of image manipulation is offered at the printing stage. The printing machine exposes the film stock to a light that can be precisely controlled in both its intensity and its color quality.
The light hitting the film is composed of three primary or three complementary colors, depending on the printer. Therefore, one lab may call its lights RGB (red, green, blue), whereas another will work with CMY (cyan, magenta, yellow). Regulating the amount of individual colors used in printing is called timing (or grading in England), hence the timer (or grader).
The printer machine has fifty or more light intensities, called lights, to choose from. Eight lights equal approximately one lens stop. The ideal printing lights for a correctly exposed scene differ from lab to lab and from cinematographer to cinematographer.
With Eastman negatives the printing light for the yellow or blue designation is always lower than the other printing lights, giving for example a set of lights for a given scene as 30–30–23 instead of 30–30–30. Usually the range of printing lights from 25 to 40 represents a well-exposed negative.
An underexposed, thinner negative will require lower-number lights, although it is called “printing up.” On the other hand, a well-exposed, dense negative requires higher lights and is called “printing down.” Printing up means printing lighter; printing down means printing darker. We can state here the relationship between exposing the negative and timing the print as follows: expose the negative well and print it down for rich color and rich black.
John Alonzo, ASC
If you want rich blacks, you expose to your key light and you fill your actor as full as you want and then print him down. As you are printing him down, the background will get rich and velvet black. But if you do not want rich black, then you add light to it so that you have a soft gray or a soft black but not a jet black.
You underexpose the negative and print up for less contrast and more grain. Too much underexposure causes the blacks to get milky and lowers the contrast. Too much overexposure also lowers the contrast. It is important to remember that when we talk about under-and overexposure, we understand it in terms of the whole frame. It is quite normal that parts of a scene are under- or overexposed, yet the scene prints in the middle of the scale. These extremes are often purposefully designed by the cinematographer for the creation of a certain mood.
Vilmos Zsigmond, ASC
In many paintings people’s faces are not always painted in full light. You remember it and you say, “Well, I will underexpose,” but it is not really underexposing; it is underlighting selectively. Many cameramen develop a technique of underexposing everything; they do that for the look of the picture. But the problem is that darkening everything to the same degree from the beginning of the picture to the end would become boring. It is also very dangerous. For example, if the f/stop is not working correctly on the lens: you are already on the very edge, on the threshold when underexposing, and now a half stop more underexposure caused by a lens malfunction can be disastrous. I do not like to work that way. I like to stay safe in the middle. My dailies always print in the middle of the scale. Around #30. Whatever I wish to appear dark, I will light dark. But still, it is perfectly exposed. Everything will be printed at the same light level.
To obtain the desired look of the print, a cinematographer must establish good communication with the timer as early as possible, preferably during the testing in the preproduction period. For financial reasons, elaborate timing is usually done only on the edited film, when answer prints are being made.
During the shooting, the lab provides one-light dailies. In reality these dailies are usually printed at about four lights, which ideally the cinematographer had a chance to test with the lab. These separate lights are usually for day exterior, day interior, night exterior, and night interior. Apart from financial savings, one-light dailies allow the cinematographer to judge exposures against constant values.
Vilmos Zsigmond, ASC
I try to create the mood with lighting, filters, and so on, but if I cannot do it, I will sometimes ask the lab to print some scenes colder or warmer, say for the sunset tones. I will ask them to do this for the dailies.
I used to establish separate printing lights for the day and for the night scenes. Now I want them to be printed using the exact same light. I use less lighting at night and still print on the same light (#30), so night looks darker than day. Having one printing light, the lab makes fewer mistakes.
I change my ASA rating on the light meter when wo
rking on a night scene. You have to learn to underlight night and to light the day normally. For example, if a daylight scene needs 50 FC with f/4, then for night I will need only 25 FC for f/4.
I like my night interiors warmer, and sometimes if I don’t want to use gels, I may establish with the lab that my nights will be printed warmer. On Heaven’s Gate we had a scene in a canvas tent, so we had the tent painted to have it warmer.
To work with the lab effectively, we have to eliminate as many variables as possible by introducing standards and reference points. One such reference point is the gray scale, and many cinematographers provide their timer with a gray scale for each lighting situation.
Allen Daviau, ASC
I always shoot a gray scale in each situation. If I am all day in an interior, though, I may shoot the gray scale just once. It acts as a control for me to see how the lab is maintaining its consistency in printing color. Even though my printing light is locked, some dailies’ supervisors appreciate seeing the gray scale. It lets them know that everything is working correctly at the lab, that nobody made a mistake on the printing light setting. I find that the gray scale is just one more guarantee of consistency. If I shoot the gray scale in a normal color temperature light source in each lighting situation, even though every light in this scene may have gelatin on it, the fact that I have shot that gray scale in white light gives a point of reference for the timer. I give him a simple instruction: “Time color to gray scale.” By this I mean that he shouldn’t base the density reading on the gray scale totally, although I will try to keep this density as close to my key light situation as possible, but to evaluate the use of the three colors in the printing light in such a way as to give me a true gray scale. And then I will not talk to him about the color at all. I will say, “Evaluate the color by this gray scale and then don’t be alarmed if there is a warm source or a cool source or anything else in the room.”
Film Lighting Page 14