Without getting heavy into the math, light pollution is a signal, not noise. Your target is also signal.
Lets say that you get 100 points of signal (ADU) from the light pollution. And let us say that the pure target signal looks like this on part of the imaging chip.
113, 110, 114, 117, 116
With light pollution it would be
203, 210, 214, 217, 216
Can we recover the signal from the target? Yeah, no problem, just subtract 100 from everything. We are back to
113, 110, 114, 117, 116
It is NOT the signal portion of light pollution that causes a problem. It is the noise portion. The noise goes as the SQRT of the signal. So if the signal from the light pollution is 100, then the average noise is 10.
Now where am I going with this?
Light pollution filters work by blocking part of the light spectrum where you expect to get some of your light pollution (low pressure sodium, mercury, etc.). But to do so, they are also blocking some of the signal from your target. And to make matters worse, much of the light pollution is not in the portion that gets blocked (incandescent lights, LED lighting, etc.)
It is often the case that the signal you lose from your target, by going with a light pollution filter, is more than the noise you incur by allowing the light pollution, (which you won't be totally blocking anyway).
And so by using a light pollution filter, you can actually decrease your S/N ratio.
Now there does come a point with faint enough targets in bad enough light pollution that the light pollution filter does actually help. But you won't know that point without doing some experimentation.
I have heard some people mention around here that they thought that using a light pollution filter actually hurt their images, and they were very possibly right. And these are people that aren't imaging in Bortle 3 and 4 zones with minimal light pollution.
Manufacturers of light pollution filters are unlikely to explain this. They are out to sell you filters.
Now that said, there are targets out there that will almost certainly benefit from them. These are the SAME targets that are used for narrow band imaging. Much of their signal is in very narrow bands and when they design the light pollution filters, they avoid notching out those bands. So for example, if you are taking a picture of the North American Nebula, or the Veil Nebula, you will want that light pollution filter on.
But if you are taking pictures of Galaxies or reflection nebulae like the Pleiades, they may or may not do you any good, and in fact could be harmful.
Moral of the story... Don't assume it is going to help unless you are on a narrow band target. Try it and see if it is actually making a positive difference.
Edited by Madratter, 26 January 2020 - 12:22 PM.