Hello Michael,
the requirement for a filter depends what type of variable star you are observing. I observe Cataclysmic Variables, in particular Dwarf Novae in outburst. I usually do unfiltered photometry to get the best signal-to-noise ratio. I’ll use a V-band comparison star sequence. Observing some eclipsing systems, where timing the eclipse is critical rather than the actual mag, can also been done unfiltered. For much other VS work a filter is highly desirable in order to be able to compare your photometry with other observers – and in the case of coloured variables (like Miras) a filter is essential. Whilst many amateurs use a V-band filter, some use a B, R or I filter. These are specific wavebands, based on the Johnson-Cousins system, which are designed for photometry and are thus different for standard coloured filters for tri-colour imaging (and different from UV or IR cut-off filters).
I am not sure where one can buy 2-inch V filters. The company Astrodon lists them, but they are unmounted. Maybe someone else can advise.
Best regards,
Jeremy