I've had my Sigma 85mm out under the stars twice so far, the first time without a high quality UV/IR filter and the second with one. I usually have a high quality UV/IR filter living over the front of my lens' as a protection against the elements. A hardy piece of glass that is worth $100-150 but makes sure that nothing damages the real nice glass that comes with your high quality lens
Bit of rain, dust, splashes of mud; remove lens, clean and put back on. No harm done.
After the 4+ hour ordeal of calibrating, debayering, registering and stacking my Rho Oph shots, most of which taken on the second night, I couldn't help but notice that the stacked image lacked the sharpness of the first night and displayed some weird star shapes.
This morning I went and stacked 8x120s from both the first night (no filter) and second night (with filter) and have since come to the conclusion that for astro work, NEVER using a filter.
I did attempt to get these two comparisons down to <200kb but by that point the quality was too bad to even tell the difference between them.
First Night (without filter)
Second Night (with filter)
Both of these are shot @ F/2.8 with a Nikon D7200 (APS-C with 3.91 micron pixels and no anti-aliasing filter)
I think for every day usage the filter is worth while, the sharpness difference is possibly not perceivable when shooting anything other than stars. Still, it is something to think about.