By William Schneider
Most photographers have a sense of what flare is, especially specular flare. Specular flare is distinct enough that even non-photographer recognize it. This kind of flare is so prevalent that some software programs have a filter to add it in varying degrees to any picture as a special effect.
Depending upon the individual lens used and the position of light sources in the frame, specular flare can range from enchanting to downright ugly. The common cross-star filter attached to the front of an otherwise good lens produces rays from each light in the scene, and it's considered a pleasing form of flare. However flare can get ugly too. Below is an slide made using an older 35mm ƒ/1.8 Minolta lens. Look in the lower left of this photo to see a grotesque flare shape that detracts from the quiet scene.
The odd-shaped specular flare in the lower left detracts from this photo.
More insidious is veiling glare. With it, there are no obvious specular shapes in photos, just a lowering of overall picture contrast. It often goes unnoticed. Shadow contrast is especially hit hard. Dark tones are swamped with unwanted light from flare, artificially lightening them and reducing contrast.
Look in the dark tones for the effects of veiling glare.
In the old film days, using UV filters on the front of the lens for protection was hotly debated. Some said that it made no difference in picture quality, others swore that filters hurt the picture. What was often missing in these arguments is whether the filters were optically coated or not, and what kind of films were being used. B&W films have a tremendous ability to record information deep in the shadows where veiling glare lurks, but slide films can't reach down as far, and that matters. In short, if the sum of veiling glare and shadow exposure is below the film or sensor threshold, the effects won't be seen. I suspect that modern digital cameras fall somewhere in between B&W and slide films in how they record the deepest shadows.
The lens used for this 4x5 photo has low flare, although there are still indications of it around the street lights. To minimize flare and maximize contrast in the shadows, no filter was used and a lens hood protected the lens from a nearby street light above the frame.
I once did some experiments to see how Tri-X negatives fared in three different scenarios—using a lens hood to shade the lens from the sun, using a bare lens, and attaching a non-coated UV filter. (A non-coated filter reflects about 4% of the light that hits each surface. A multi-coated filter will reduce reflections to around half a percent, an 8x improvement.)
Shooting into the sun in a flare light torture test.
I tested two different Nikon lenses—one was a Nikon 45mm ƒ/2.8 P "pancake" lens that has only four elements in three groups. Having few glass/air surfaces, it should be naturally flare resistant. The other lens tested was a Nikkor 50mm ƒ/1.4 AIS, with seven elements in six groups, and obviously more prone to flare. I photographed the setup below, exposing for the gray card sitting atop the cardboard box containing the flare target. The flare target was a printout of type in various shades of gray from 0% (paper white) to 95% (very dark gray) all against a black background.
The first comparison negatives from the 45mm pancake lens in this test clearly showed a difference between using a filter and the bare and hooded cases. With this flare-resistant lens, very little change was noted between using a hood or just a bare lens as you can see in the animation below. That the lens hood didn't make much difference is not really a surprise here.
Animated comparison of cropped negatives from the 45mm pancake lens.
The next picture shows the flare performance of the 50mm ƒ/1.4 lens using the same configurations of UV filter, bare lens, or lens hood.
Animation showing scanned Tri-X negatives from the 50mm f/1.4 lens.
It's easy to see that the non-coated filter diminishes contrast in the negative with both lenses, and that the lens hood offers the most improvement for the most flare-prone lens. I suspect that zoom lenses, having many more elements than my 50mm ƒ/1.4, would benefit the most from a lens hood.
I also suspect that the increasing dynamic range of newer digital camera sensors will re-ignite the filter vs. no filter argument. As newer cameras begin reaching deeper into the shadows for information, the effects of flare will become more evident.
Filters are useful for protecting lenses, but substantially-made lens hoods offer a lot of protection while improving performance. That’s something no filter can do. Wide-angle lenses are special problem because their lens hoods are too shallow to provide much protection. A filter is wise here, especially if the filter has an anti-reflection coating to reduce flare effects. Quality optical coatings do help.
Use protective filters when needed, but recognize troublesome situations to know when to remove them. Shade the lens from oblique lights using lens hoods, or even your hand if you can. Don't forget the old trick of positioning bright lights in a scene behind branches, people, etc. to reduce all types of flare.
Using a low-flare lens, removing filters, and hiding the direct light source are ways to reduce both specular and veiling flare. Note the substantial detail in the shadowed side of the tower.
No lens or camera is completely free of flare, but learning what you can and can't do is part of understanding your equipment.
ADDENDUM: As I mentioned in the story, I don't think that it's wise to go to one extreme or the other regarding filter usage. There are times where the pluses and minuses should be weighed. I'd use a filter, for example, when shooting a muddy motocross race close to the track, or in blowing sand and surf.
If there's an expectation of danger to the lens itself, by all means, use a filter. You'll also have to factor in how you treat your own equipment in the course of working. In short, use reason.
Several comments were made by users who have coated filters. Not all filters coatings are equal. In the photo below, I arranged a collection of filters in increasing coating effectiveness from left to right.
At the far left, I placed a piece of plain glass for comparison. I also placed a stainless steel ruler in the scene to show that the reflected illumination (from a white cathedral ceiling lit by a softbox) was very even.
The cheapest uncoated filters reflect as much light as the plain glass. The middle filter has a single coating. At the far right is a B+W multicoated filter showing exceptional performance. However, be prepared to spend a some money on a filter of this quality.
Bill Schneider teaches in the School of Visual Communication at Ohio University.
Additional reading: Mike's take on the filter vs. no filter question.
Note: Links in this post may be to our affiliates; sales through affiliate links may benefit this site. More...
Original contents copyright 2012 by Michael C. Johnston and/or the bylined author. All Rights Reserved.