Many photographers tout the use of clear filters on their lenses to provide general protection from bumps and scrapes. The idea is if you're going to scuff your camera, you damage a $100 filter instead of the $1000 lens. But other photographers—myself included—tend to prefer working without protective filters. That's not to say I recommend against using a skylight filter to change the tone of a scene (which it will do, adding a slight warmth) but that doesn’t make me a fan of the "protect my lens with a filter" approach. Why? Because I want to minimize the glass between my sensor and my subject unless I’m explicitly trying to alter the image with a filter. I don't want to run the added risk of flare, dust and general image degradation on all of my images just because I eventually might damage my lens. Why mess with every single picture I take, even if it’s just a little bit?
A blog reader recently asked me to do a test to see exactly what difference there is between filtered photos and those made with nothing over the lens. It was a brilliant suggestion, so I did it. Here’s what I found.
For my first test, I didn’t use any cheap old filter. I used a name brand, high quality $100-plus skylight filter. A skylight filter is frequently used as if it were the same as a UV filter, and it’s often touted as such. But skylight filters are designed to remove excess blue, which means that they do cut light (not much, but some) and create a subtle color shift—unlike a UV filter, which is designed to only cut ultraviolet light. I photographed a test chart and a color checker using no filter, a dusty skylight filter, and then a spotlessly clean filter. What I found is first and foremost that a dirty filter makes a huge difference in pictures. Further proof that it’s crucial to keep your filters, and lenses, clean. (And reinforcing my own bias toward not using a filter, as it’s just another two glass surfaces that can hold dust and dirt.)
The next thing I discovered is that, yes, even a clean skylight filter alters the image. It changes the tone, making the scene warmer. This could often be a benefit, but for someone who shoots RAW, I want as much data—including color—as I can get. If I want a warm tint, I’ll add it in post or adjust my white balance in the camera. I don’t need to use a skylight filter for that effect.
The other thing I noticed is that in the unfiltered images there was subtly more detail in the black tones. This could be a function of the tenth-of-a-stop light-cutting power of the skylight filter, or it could be a bit of optical confusion causing a loss of detail. The bottom line, though, is that in the blacks of the unfiltered image I saw subtle details that were missing from the filtered file. It’s not a lot, but it’s there.
The second test I did was to try a UV filter on my lens, both in the studio on test charts and outdoors on a cloudy day. In both instances I must say I didn’t see much of a difference. (But I know from my previous test that if I start getting dirt on that UV filter it’s going to have a huge impact on my images.) UV and haze filters are really designed to filter out ultraviolet light that creates a hazy look in the distance of outdoor photos. Errant ultraviolet light can make an image appear less sharp and clear. So I strapped a clean, quality UV filter on a sharp 100mm lens and shot the same scene at a long distance. What I saw was an extremely subtle difference. The image without the filter was ever so slightly sharper and contrastier than the image with the filter—noticeable only in tiny text far off on street signs in the distance.
No UV Filter
With UV Filter
In the end, these differences are so subtle that I say it comes down to personal preference: do you prefer the comfort of knowing that nothing comes between your lens and your subject, or do you prefer the knowledge that your expensive lens is always protected? For me, even a tiny difference is enough to reinforce my preference for no added glass on my lens. After all, if I stacked three UV filters on my lens I’m sure I’d start to see a major difference, and with a dozen I’m sure the effect would be horrendous. That just reinforces the idea for me that any filter, even a clean and clear one, does alter an image. Then again, working without a filter increases the chances that a simple little bump will ruin an expensive lens. So do what you feel comfortable with and enjoy the benefit that comes with either decision.