Using Filters for Protection - pros & cons

Back to Home Page

Like many amateur photographers, I am in the habit of buying a 'protection' filter, to avoid getting dirt and splashes on the front elements of my valuable lenses. This note discusses the trade-offs inherent in this approach.

It is important to recognise that a filter can only subtract from the performance of the lens; in other words, there will always be some loss of optical performance. In some cases, the subtraction may prove beneficial - a colour filter removes some wavelengths,which may otherwise be over-emphasised in the final image, or a polarising filter removes some reflections, which may help the overall composition.

When we use a filter for 'protection', however, we need to consider whether the benefits outweigh any adverse optical effects of the filter. Usually, we choose a filter which has minimal effect on the transmitted light, such as a colourless UV filter. This filter may have little impact on the transmitted light but it also has two reflecting surfaces and these can seriously affect image quality in some situations. It is well known that coating or multi-coating reduces the amount of light reflected from glass surfaces. The following examples show how significant this can be, when taking photographs in high-contrast conditions.

The images on the left show the effect of reflections from the surfaces of different types of filter. All these photos were taken with an Olympus EP-1 camera fitted with a 17mm f/2.8 lens.

A bright light source (desk lamp) was placed in the lower part of the subject area and the top image shows that, apart from some flare around the lamp, the lens itself reproduces the scene pretty faithfully.

If an uncoated clear glass filter is placed over the lens, reflections from the flat surfaces of the glass produce a strong 'ghost' of the lamp in the upper part of the image.

The second image in this row was taken through a Hoya Skylight 1A filter. This filter still produces quite a strong ghost image, though it is not as bright as from the uncoated glass.

A greater improvement is obtained by the use of multi-coated filters. The results in the two examples shown here are very similar; in one I used a Hoya HMC Skylight 1B filter, while the other was a Jessops 'own brand' MC-1B filter

It is clear that the multicoated filter has reduced the effect of reflections considerably but they are not totally eliminated and the final result still does not equal the performance of the lens alone, with no filter.

These results, taken together, show that even the multi-coated filters introduce some ghost images, which are not apparent when the lens is used alone. In critical conditions, such as stage photography, where bright lights may intrude into the photograph, there should be a visible advantage if no filter is used. In general photography, however, it seems unlikely that the ghosting observed under the extreme conditions of my tests would be significant, especially if multi-coated filters are used.

On balance, therefore, in the case of my own use for natural history photography in the field, often in wet and muddy conditions, I feel that the benefits of adding a multi-coated filter as lens protection outweigh any potential image degradation due to 'ghosts' - but see below....

Fast Telephoto Lenses

I recently acquired a 300mm f/4 telephoto lens, specifically to photograph birds in flight. This lens takes a large (77mm) filter and I initially fitted a Hama UV filter ('borrowed' from a wide angle lens), to protect the front element. My initial trials with the lens were rather disappointing and, in particular, I noticed some strange bands appearing in out-of-focus parts of the image. These disappeared when I removed the filter and I also noticed a substantial improvement in overall image quality.

'Banding' in out-of-focus areas of an image when using Hama UV filter
(Nikon D300s with 300mm f/4 lens with 1.4X TC)

A high-magnification lens is especially susceptible to any imperfections in the filter and, since the filter is large, maintaining perfect flatness becomes a severe manufacturing challenge. To confirm my suspicions of the effect of the filter, I took a series of photos of a test chart, using in turn the relatively inexpensive Hama filter, no filter, and an expensive B+W F-Pro UV filter. I took all these photos under identical conditions, with the camera on a tripod and with flash illumination. The lens was fitted with a 1.4X teleconverter to give an effective focal length of 420mm and an aperture of f/5.6. The shutter speed in all these photos was 1/250s.

The results show that the Hama filter has a significant effect on the sharpness of the image whereas the B+W filter provides results that are indistinguishable from the lens used alone. In tests with smaller lenses of shorter focal length, I failed to detect any significant loss of sharpness when a range of filter types was used.

My conclusion is that, in the case of large high-power telephoto lenses, it is preferable to use no filter at all. If (like me) you have a strong desire to 'protect' the front element, then a top-quality filter, such as B+W, must be used. Even then, I would recommend comparing images with and without the filter in place, to ensure that the results are satisfactory.

In the field, I first noticed the adverse effect of the filter on out-of-focus parts of the image, so I carried out more tests with a square-grid test chart, deliberately shot slightly out-of-focus. Close examination of the results, below, reveals the 'banding' (especially where the vertical and horizontal lines cross), which I had previously seen in the field when using the Hama filter.


Mike Flemming, Jan. 2012

Back to Home Page