Let me begin by stating that this article will be quite different from the usual content. Instead of reporting on some major news item, or some interesting piece of mathematics, or explaining some theory from modern physics, today's entry is the result of a simple experiment that I performed last night, which has very little importance but may help to answer a question that appears to be unanswered on the internet. 

As I have written in the past, I like to indulge in some very simple astrophotography. Unfortunately I am usually confined to areas with poor viewing conditions due to the presence of light pollution from the nearby city. As such I started pondering whether or not there was a good light pollution filter for a dSLR camera. Of course a quick online search reveals several quality filters for just this purpose - however they tend to cost over a hundred dollars for even the most basic products. Clearly the average amateur is not going to be putting out that sort of money for a very specialized piece of equipment.

And so I began to wonder whether any of the standard photography filters would reduce the level of light pollution. Is the light from street lamps and neon signs polarized differently from the light from stars? Does light pollution have a large component from fluorescent lights? What does light pollution in the near infrared spectrum look like? To my surprise, the first few pages of results on Google are simply debates between people who have stubborn opinions but have not actually gone outside and tested anything!

That gave me the idea of taking basically every type of filter I have in my collection, and taking exposures of the night sky through them. In each case I have taken a single photo rather than a stack, and where possible used either the same exposure time or a pair of exposures where the second one is adjusted to keep the overall light levels comparable. (One of the common criticisms of using polarizers at night is that they do reduce the available light, and therefore would result in larger star trails). I have also omitted post-processing and dark frame subtraction so as to focus entirely on the effects of the filter.

In short, it would appear that the linear polarizer has minimal effect at any angle, and as expected a circular polarizer only serves to extend the exposure time. Similarly, the fluorescent filter would appear to only alter the colours and doesn't really reduce light pollution. The infrared filter may have some effect, but at the cost of greatly extending the exposure time. It appears that there really isn't a substitute for a higher end specialty filter.

PS: After writing this entry, I learned that the large photo files were causing problems with the blog. As such the results of this experiment can be viewed here instead.