Narrowband filters - Nm choice - Forums

Since getting such a positive response from my "Jump in" question, I've been reading as much as I can on narrowband astrophotography.

Now I have to choose among several filters - 3nm, 5nm, 6nm etc. I understand that in more light polluted skies, one should choose a wider nm band. My question is in Bortle 4 skies, what is good enough? Am I going to see a great deal of difference between 3nm and 6nm in the quality of the subs as an example?

Hank

The narrower the bandpass, the more contrast you get … the more of the extraneous light you block.

"Great deal of difference in quality" is entirely subjective, and I don't have both a 7nm and 3nm filter to give a comparison sub

I understand that in more light polluted skies, one should choose a wider nm band.

It is the other way round. In more LP skies you want a narrower band. A narrower band will almost always benefit you (other than for very fast systems, meaning faster than f/3 or so, which require specialized filters). With mono, there really isn't a reason, other than cost, to use wider bandpass filters, whether at light polluted sites or otherwise. Even at Bortle 4, a 3nm filter will benefit you in gaining SNR faster for faint objects. If you can afford it, go with a 3nm filter set, regardless of Bortle class.

You are backwards on when you use wider or narrower. The more light pollution, the more benefit you get from narrower filters. I get very good results with a 12nm dual band filter and even had some good results using a 48nm dual band UHC filter from a bortle 5/6 (veil, north america and elephant trunk nebula in my gallery are 48nm). Narrower usually makes smaller, tighter stars and most people advise to get the best you can afford. I'll be buying 7nm because I can't afford 3nm but the remote scopes I use which are all in ideal bortle 2 observation areas all use extremely expensive 3nm filters so they must provide some benefit, I just can't quantify how much.

3nm to 4nm will give you grand detail. However filters depend on your budget, the lower the number the stupidly expensive they are. If your starting off go mid range and get use to the data. I went from 7nm to 3nm and oh my the detail difference was noticeable. I was glad I started at a higher nm though, as this got me into the process of dealing with narrow band. Enjoy the detail that will present it self, when you get going.

Joe Linington:
I just can't quantify how much.

One way to look at this - neglecting sources of noise like dark current and read noise and focusing exclusively on shot noise. There will be two kinds of shot noise - the shot noise from the sky background and the shot noise from the photons of the DSO itself. At a light polluted site, the shot noise from the sky background dominates. So going with a 3nm filter over a 12 nm filter should show a very significant enhancement in SNR (and hence improvement time). But at darker sites, the shot noise from the DSO itself is a larger proportion (though may not still dominate), so the benefit of a 3nm filter will be less, but still there. The bottom line is that it is always beneficial to go with a narrow bandpass filter, but the amount of benefit will depend on what you are imaging and where you are imaging from.

The narrower, the better, irrespective of sky quality. With 3nm , you can fight the moonlight, especially Ha. I have Antlia 3nm SHO set (36mm) and am perfectly happy. Antlia filters are one of very few ways to get reasonable 3nm without selling a kidney

I moved from 7nm to 3nm and the difference in contrast / detail was pretty amazing. They also handle bright stars much better - Oiii halos are all but non-existent.

The more light pollution you have the more the smaller bandwidth will benefit your imaging.

Jim Lamb has a pretty good video on his transition from 7nm to 3nm filters, you can find it on YouTube.

I have found it depends on the skies and filters. I, as many use Broder Ha and narrower OIII . I have used 12 nm on Hyperstar to avoid light fall off due to Wide angled light cone. My skies were around Bortle 6 then. Now they are Bortle 7. I have used 5nm Astrodons under Bortle 7 and the Ha and SII was clear and showed no gradients. The OIII was noisy. I switched to a OIII 3nm and the noise decreased dramatically.

I image at F4.5, 3.6 and F3. So, I needed filters that would accommodate those fast ratios. I when with the Astronomik MaxFR 6nm filters for Ha and SII. As mentioned above, I needed 3nm OIII that would work at F3. I choose the Chroma 3nm adapted to F3 filter. Expensive, but the alternative is poor data.

Hank, unlike you I am restricted to narrowband imaging at smaller nm filters due to Bortle 7 skies. With your Bortle 4 skies, you have options. When I go to my dark Bortle 2 site, I use 12nm only as added Ha and posible OIII to RGB and LRGB images. At Bortle 2, I noticed the my Ha data showed no more resolution than the Luminance data on M17. However, even at Bortle2, Ha data is crucial on such objects as the Lobster Claw due to its low magnitide and covered by Milky Way stars.

Wider nm Ha also geather Ni data.

Lynn K.

.

Lynn K:
Wider nm Ha also geather Ni data.

Lynn, I've heard this argument in favor of 5nm H-alpha filters over 3nm. But how many DSOs actually have nitrogen emissions of significance? And is the small benefit from gathering some of these from some DSOs worth the loss in contrast and higher imaging time on most DSOs? The nitrogen emission theory was why I purchased 5nm H-alpha over 3nm, but today, I'd purchase 3nm across the board.

I can't show you a comparison between 5 or 6nm filters as I've been using 3nm for 14 years. I can show you what a single unprocessed, uncalibrated image looks like using the 3nm Ha, NII, and OIII filters. Keep in mind that these were taken using an SBIG ST-10 at almost 90%QE and a TEC 140 f/7. I'm sure using a CMOS will give difference results. These are single 1200 second exposures. As for what the NII filter are mostly used for, and this is from the company that made then, PN's, there is more NII in planetary nebulas then in any other DSO. So unless you plan to image a lot of PN's, getting an NII will not help you much. The results of using these three filters can be seen hear.

Helix Nebula / NGC 7293 ( Bob Lockwood ) - AstroBin

And the filters. Single 1200 sec uncalibrated. From Bortle 3/4 skies

Arun H:
Lynn K:
Wider nm Ha also geather Ni data.

Lynn, I've heard this argument in favor of 5nm H-alpha filters over 3nm. But how many DSOs actually have nitrogen emissions of significance? And is the small benefit from gathering some of these from some DSOs worth the loss in contrast and higher imaging time on most DSOs? The nitrogen emission theory was why I purchased 5nm H-alpha over 3nm, but today, I'd purchase 3nm across the board.

Arun, I too have read verious views on benefits of narrower nm filters. I have a limited field of view and few Clear nights. So, I have become a real fan of fast scopes, and am limited to filters that will work with fast optics. The Astronomik MaxFR have worked well, and have received great reviews. But, they only go as narrow as 6nm. The F3 version of the Chroma OIII is, I belive, special oder.

The question I am asking myself is since I used the Astrodon 5nm OIII with a Sony CCD chip that had poor OIII QE, and am now using a CMOS Sony IMX571 with high 80s OIII QE, would my S/N improve enough to use the 6nm Astronomik MAXFR. The reson I am proposing such a switch is to be better parfocal. I want to totally automate my sessions, and there is quite a bit of focal difference between the Astronomik and Chroma. Going all Chroma will cost $2600 for a F3 optimized Ha & SII 2".

Lynn K.

As already pointed out narrower band passes are going to give you better contrast. It's just the nature of how light works. Ha emits light at only one very specific wavelength, and the narrower the band pass the less pollution you'll have in there from the rest of the red part of the spectrum.

HOWEVER, here are some super niche things to consider:

If you go with a narrow band pass, you are pretty much locked into a floor in terms of your focal ratio. Don't expect to use something like a 3 or 4 nm filter on a faster instrument (say below F/4) and expect efficient results. A 7nm filter, for example, is more forgiving in this regard because the band pass is so large you can still get decent transmission percentages at lower focal ratios. Not a lot of people image under F/4 so like I said, this is very niche.
Do you like imaging galaxies? Do you like adding Ha to your galaxies? A wider band pass might be for you! Some galaxies are moving fast enough away from us that they are red shifted to a point where narrower band pass filters wont pick up the Ha line anymore.

In terms of cost, you may wish to splurge on the Oiii filter and cheap out (relatively speaking) on Ha and Sii. Oiii is sensitive to everything under the sun and loves to bless you with halos that I know everyone loves to deal with during processing. So, its not uncommon to see expensive Oiii filters and cheaper Ha and Sii filters in people's gear section here.

At B4 it just really comes down to how much money you want to dump into filters. Wide or narrow band passes, your sky will provide you with quality images either way. It just depends on how much contrast you want.

as someone who recently moved from the ZWO 7nm to the Antlia 4.5nm Edge and now using the Optolong 3nm, I saw the biggest jump from 7 to 4.5. for the money, the Antlia Edge 4.5nm were really amazing. I bought the new Optolong 3nm because I was battling area lighting, but I think I could have been happy with the Antlia's in the end. The Optolongs are definitely the entry level to the 3nm, so it could be that Astrodons and Chromas will blow my mind one day. For now though, I'd go back and tell myself to stick to the Antlias until I had a solid reason not to. just my 2 cents

Thank you to everyone who replied to my nm filter question. I now have a better understanding on how to go about selecting and using filters

Do you like imaging galaxies? Do you like adding Ha to your galaxies? A wider band pass might be for you! Some galaxies are moving fast enough away from us that they are red shifted to a point where narrower band pass filters wont pick up the Ha line anymore.

I was curious as to exactly what the effect will be of the expanding universe and associated redshift on how much H-alpha radiation will actually be transmitted by a 3nm H-alpha versus 5nm H-alpha.

For a Chroma 3nm bandpass filter, the %T drops to ~50% at about 1.3 nm from the 656.3nm peak. For a 5nm bandpass, this increases to ~2.3nm.

A redshift of 1nm at the H-alpha wavelength equates to a recession speed of 457 km/sec (v=c*DL/L=300,000*1/656.3 km/sec).

Using a Hubble constant of 70 km/s/Megaparsec and Hubble's Law (d=v/H), we can see that, for the H-alpha transmission to be 50% of peak with a 3nm filter, the galaxy would have to be ~8.5 Megaparsecs away, which is about 25 million light years.

For a 5 nm filter, this increases to 45 million light years.

So the effect on nearby galaxies like M31 and M33 will be minimal.

But once you get to galaxies like M81 and M82 (about 12 million light years) or the Leo triplet (35 million light years), these differences can start to get significant and it might be preferable to use wider bandpass H-alpha filters, especially at dark sites.

Arun H:
Do you like imaging galaxies? Do you like adding Ha to your galaxies? A wider band pass might be for you! Some galaxies are moving fast enough away from us that they are red shifted to a point where narrower band pass filters wont pick up the Ha line anymore.

I was curious as to exactly what the effect will be of the expanding universe and associated redshift on how much H-alpha radiation will actually be transmitted by a 3nm H-alpha versus 5nm H-alpha.

For a Chroma 3nm bandpass filter, the %T drops to ~50% at about 1.3 nm from the 656.3nm peak. For a 5nm bandpass, this increases to ~2.3nm.

A redshift of 1nm at the H-alpha wavelength equates to a recession speed of 457 km/sec (v=c*DL/L=300,000*1/656.3 km/sec).

Using a Hubble constant of 70 km/s/Megaparsec and Hubble's Law (d=v/H), we can see that, for the H-alpha transmission to be 50% of peak with a 3nm filter, the galaxy would have to be ~8.5 Megaparsecs away, which is about 25 million light years.

For a 5 nm filter, this increases to 45 million light years.

So the effect on nearby galaxies like M31 and M33 will be minimal.

But once you get to galaxies like M81 and M82 (about 12 million light years) or the Leo triplet (35 million light years), these differences can start to get significant and it might be preferable to use wider bandpass H-alpha filters, especially at dark sites.

That is awesome. Thank you, I have my learning of the day done.

Arun H:
Do you like imaging galaxies? Do you like adding Ha to your galaxies? A wider band pass might be for you! Some galaxies are moving fast enough away from us that they are red shifted to a point where narrower band pass filters wont pick up the Ha line anymore.

I was curious as to exactly what the effect will be of the expanding universe and associated redshift on how much H-alpha radiation will actually be transmitted by a 3nm H-alpha versus 5nm H-alpha.

For a Chroma 3nm bandpass filter, the %T drops to ~50% at about 1.3 nm from the 656.3nm peak. For a 5nm bandpass, this increases to ~2.3nm.

A redshift of 1nm at the H-alpha wavelength equates to a recession speed of 457 km/sec (v=c*DL/L=300,000*1/656.3 km/sec).

Using a Hubble constant of 70 km/s/Megaparsec and Hubble's Law (d=v/H), we can see that, for the H-alpha transmission to be 50% of peak with a 3nm filter, the galaxy would have to be ~8.5 Megaparsecs away, which is about 25 million light years.

For a 5 nm filter, this increases to 45 million light years.

So the effect on nearby galaxies like M31 and M33 will be minimal.

But once you get to galaxies like M81 and M82 (about 12 million light years) or the Leo triplet (35 million light years), these differences can start to get significant and it might be preferable to use wider bandpass H-alpha filters, especially at dark sites.

RJF-Astro on Cloudy Nights crunched some numbers a while back for some popular galaxies. Not sure how good they are though.

RJF-Astro on Cloudy Nights crunched some numbers a while back for some popular galaxies. Not sure how good they are though.

That's a nice table. Using the actual redshift value of a galaxy is always going to be preferable to using the Hubble Law like I did, simply because there is scatter in the Hubble plot that is going to be more significant with close galaxies. The question, though, isn't just whether 3nm is suitable, but whether 5nm can be better.

For example, for M74, a 3nm filter will have a transmission of 80%, but 5 nm gives 97.5%. So while the 3nm will certainly work, the 5nm will be better, and probably preferable if one is imaging that galaxy from a dark site.

Ultimately, as you mentioned, it comes down to how much galaxy imaging you are doing and where you are doing it from. If you plan to do mostly nebula work, which is what most people here do, there isn't a question that 3nm will be superior, other than perhaps for planetary nebulae with significant NII.

Going further out and just to add as a point of tangential interest related to Arun's post above. ..

It turns out that the recession velociity of many galaxies in the Coma cluster (Z ~ 0.025) corresponds to a shift of about 16 nm which is convenient because you can then switch to an SII filter (672 nm) in order to pick up H alpha emissions. Using that filter I picked up IC4040 and NGC4848 as active galaxies in the coma cluster –most of the galaxies appeared not to be active - having been stripped of gas presumably.

Narrowband filters - Nm choice

Related discussions