[SamD (Sam)]

Putting some numbers to what Rick and Barry say, I get about 7e-/min skyglow at zenith with my 7nm Baader Ha (15km from Brisbane, no moon, with 314L+, and at f/7.5). This skyglow number is similar to what Craig Stark reports in his 314L+ pdf review at a similar f ratio.

If I doubled up from a 7nm to a 14nm Ha filter and follow the SNR calculations through, I would expect to need double the integration time to get the same SNR, or half the integration time if I went down to 3nm (assuming you don't chop off any signal like NII).

Even with NB I find that I'm better off to image when moonless (ie lower integration times). Less of a problem of course for the brighter NB stuff, but for the deeper stuff and a reasonably dark suburban sky I find no moon is the way to go.

Another factor of course is that the narrower filters are pretty expensive !