5nm vs 3nm narrowband filters

[troypiggo]

I'm in the process of swapping my Astronomik NB filters over to Astrodons. I recently posted here about halos/reflections in the Astronomiks (http://www.iceinspace.com.au/forum/showthread.php?t=86869). I've sent them test shots, and await their replacement filters so the problem will go away. But they're 13nm bandwidth.

But that being said, I always intended to go narrower bandwidth on the NB filters. Been doing some reading on the 5nm vs 3nm recommendations:

Ha: the 3nm filter cuts out the NII bandwidths as well, while the 5nm does overlap a little so you do gain some NII signal in some planetary nebula etc. Since I don't want to get a special NII filter, I'm thinking of going with the 5nm for Ha

OIII: I understand that OIII is more susceptible to moonlight. So the 3nm is recommended over the 5nm for this reason? But are there cons to doing this versus keeping it 5nm to match the Ha? I'm leaning towards the 5nm for cost and matching reasons.

SII: Not much I could find on this, so thinking 5nm to match the Ha as well.

So my dilemma is more about the OIII, and whether to get the 3 or 5. Any issues or comments to add?

I can't justify getting 3nm for all filters.

[Poita]

One thing to keep in mind is that 3nm = less light and longer exposures.

[troypiggo]

That's what I thought, but then I read this:
http://www.astrodon.com/Orphan/astrodonfaqnarrowband/#h19

[Poita]

Interesting. Integration times with the mallincam are longer with the 3nm than the 5nm for me. Maybe it is the brand of filter??

[RickS]

I wonder if it depends on the object you're imaging? If it's something which emits a lot of Ha then you wouldn't expect to see much difference in exposure time between a 3nm and a 5nm Ha filter. If it emits more evenly across a broad spectrum then perhaps you get a significant amount of additional light through the wider 5nm window?

I have both 3nm and 5nm narrowband filter sets but I've never really spent time comparing them. Perhaps one moonlit night I'll give it a go. Assuming the sky ever clears...

Cheers,
Rick.

[Poita]

That makes perfect sense, many objects I view with it are not purely Ha so to speak.

[troypiggo]

From what I read on that Astrodon link I gave above, it's because of the high transmission of those filters. For both 3nm and 5nm the transmission is same percentage at the relevant wavelength.