I haven't asked the journal's permission to reproduce this article, so I hope they think it's OK... In general, they're a much appreciated source of photographic information in Scandinavia.
There are legion brands of filters. And the materials they are made of can be glass, plastic or gelatine. Sometimes we even put two or more filters before the lens. How does this affect image sharpness?
To test this, we asked Per Nordlund at the Hasselblad factory to do MTF meterings for us. In addition, we wished to know whether the filter surfaces were absolutely plain, or whether they had any theoretical focal width. We also asked him to look for astigmatism.
Many of us leave a skylight- or UV-filter on the lens at all times. Partly to remove UV light, and partly to protect our precious lenses. -But how do they affect sharpness?
The longer the focal width, the stronger the negative influence a poor filter will have. So we chose to test the filters mounted on a Nikkor F 2.8/180mm. That should leave us somewhere in the middle between normal lenses and the longest lenses people normally use. Shorter lenses will do better, while longer lenses should not be loaded with more than one filter under any circumstances.
The UV- and skylight filters we have tested are Hoya, Nikon, B+W, Cokin plastic and Cokin glass, Vivitar, Rowi and Hansa. Our positive result is that there are no differences among brands as long as we stick with glass filters. These filters were almost totally flat. The theoretical focal width was around 5 km, having no effect on the image at all.
Plastic filters, OTOH, were not nearly as good. Cokin's plastic skylight filter is NOT recommended at all. Lens sharpness (MTF) falls dramatically with this attached to your lens. The filter also suffers from astigmatism, which make horisontal and vertical lines fall into different focal planes (-hey; don't ask me what they mean by that....Jostein). In addition, the filter is far from flat, having a focal width of about 100 m. This gives a focus shift at the film plane of about 0,3 mm with a 180mm lens. Under this circumstances, this is a rather high value.
The anti-reflex treatment of the different brands is very variable. A simple, visual comparison reveals that multicoated filters are superior. The B+W, Nikon and Hoya are all multicoated. From Cokin, only the glass versions are, and from Rowi, only some filters (A pro-series? -Jostein) are. The filters from Hansa, Vivitar, Cokin's plastic and Rowi's standard series have poor or nonexistent coating.
Should one go for multicoated filters? We think so. A lens doesn't get any better with this kind of filter, but OTOH, it doesn't get much worse either. In contrast, an uncoated or poorly coated filter adversely affect image quality.
We also tested ND/gray filters by MTF. Here, we compared three types of filters; multicoated glass filters (Hoya and Nikon), Kodak Wratten gelatine filters, and Cokin's plastic filters. The glass and gelatine filters yielded almost identical results in the MTF procedure. None of them gave any noticeable loss of sharpness. The MTF curves for the lens looked almost exactly the same whether or not there was a filter attached. Cokin's gray filter are made of plastic, and this filter gives a noticeable loss of sharpness. There is also astigmatism. Our advice is: use glass or gelatine filters ONLY if you want a ND or gray filter.
B+W use brass in their filter rings. This makes the filters a bit more expensive, but they are much easier to unscrew than rings made of aluminium. To our knowledge, all filter brands but B+W use aluminium rings.
From an article in the Swedish journal FOTO (Aktuell fotografi), no. 3, 1992
Translated by: Jostein Øksne