Narrowband filters isolate a few nanometers around specific emission lines — hydrogen-alpha at 656.3nm, OIII at 500.7nm, SII near 672nm — blocking nearly everything else, including most light pollution. Broadband filters pass a much wider slice of the visible spectrum and only notch out known light-pollution wavelengths. Narrowband suits emission nebulae; broadband (or no filter) suits galaxies and star clusters.
What each filter type actually blocks
Our mono vs. OSC comparison already introduces the broadband-LRGB / narrowband-Hα-OIII-SII vocabulary, in the context of what goes into a mono filter wheel. This article goes a level deeper — into what each filter type is physically doing to the light passing through it, and where the specific numbers actually come from.
Narrowband: isolating emission lines
A narrowband filter is an interference filter tuned to pass only a few nanometers of bandwidth centered on one specific wavelength, and to block essentially everything on either side of it. Emission nebulae glow at a small number of very specific wavelengths — hydrogen-alpha at 656.3nm, OIII (doubly ionized oxygen) at 500.7nm, and the SII (singly ionized sulfur) doublet at 671.6nm/673.1nm — Chroma Technology, via OPT Telescopes, Agena AstroProducts and High Point Scientific product pages and a narrowband filter is built to pass just one of those slivers. Because artificial light pollution is broadband or falls on its own separate set of emission lines (sodium- and mercury-vapor lines, LED spectra), a filter tuned tightly to a nebula's own line rejects nearly all of it as a side effect, without needing to know anything about your specific sky.
Broadband: passing everything except known light-pollution bands
A broadband (light-pollution) filter takes the opposite approach: rather than isolating one line, it passes most of the visible spectrum and notches out only the narrow bands where common artificial lighting actually emits. Astronomik's own transmission-curve documentation for its CLS filter gives two wide pass windows, roughly 450–540nm and 640–690nm, with about 92% measured transmission at 486nm/496nm/501nm and about 97% at 656nm Astronomik. That is a fundamentally different shape than a narrowband spike — it is designed to keep star colors and continuum light (the kind galaxies and reflection nebulae actually emit) largely intact while dimming the specific wavelengths sodium- and mercury-vapor streetlights add to the sky.
A CLS-type filter is a reasonable broadband example to reach for — check current price on the Astronomik CLS at Agena.
| Filter type | Named example | Bandwidth | Key lines passed | Best-suited target | Typical mount format |
|---|---|---|---|---|---|
| Broadband (CLS) | Astronomik CLS | Two wide windows Astronomik | Most of the visible spectrum, minus known light-pollution lines | Galaxies, star clusters, reflection nebulae | Clip-in or 1.25″/2″ threaded |
| Tri-band | Optolong L-eNhance | Multi-window, includes Hβ Optolong | Hα, Hβ, OIII — exact passband not published by Optolong | Emission nebulae, including Hβ-rich targets | Clip-in or threaded, mono or OSC |
| Dual-band | Optolong L-eXtreme (7nm) | 7nm FWHM Optolong | Hα 656.28nm, OIII 500.7nm Optolong | Emission nebulae under light pollution | Clip-in or threaded, mono or OSC |
| Dual-band | Optolong L-Ultimate (3nm) | 3nm FWHM Optolong | Hα 656.3nm, OIII 500.7nm Optolong | Emission nebulae under light pollution | Clip-in or threaded, mono or OSC |
| Mono narrowband | Chroma Hα (3nm) / Optolong Hα (7nm) | 3nm or 7nm FWHM, one line per filter Chroma Technology; Optolong | One emission line per filter — Hα, OIII or SII individually | Emission nebulae, maximum per-line control | Filter wheel (mono only) |
Which targets each suits
The fit follows directly from the mechanism above. Emission nebulae — regions of ionized gas glowing at those same few specific wavelengths — are exactly what a narrowband filter is tuned to isolate, so narrowband gets the most out of them. Galaxies, star clusters, and reflection nebulae work differently: they shine by continuum starlight (or, for reflection nebulae, starlight scattered off dust) spread broadly across the visible spectrum rather than concentrated in a handful of lines, so there is no narrow line for a narrowband filter to grab onto.
This is also where our mono vs. OSC guide already makes the shared-ground point worth repeating here in one line: for broadband targets, neither narrowband nor a dual-band filter buys you much light-pollution relief Urban Astrophotography — that fight is won with darker skies or narrower-band broadband techniques, not with a filter of any kind.
None of this is a genuine dispute between sources — every manufacturer and guide we checked agrees narrowband and broadband solve different problems. The real disagreement in this space is narrower and more technical, and it’s covered in full below.
Narrowband bandwidth: 3nm vs. 6–7nm
This is the genuine, unresolved dispute in this article — distinct from the narrowband-vs-broadband category question above, which nobody actually disputes. Within narrowband itself, filters ship in a range of bandwidths for the same line — a 3nm Hα filter and a 7nm Hα filter both isolate the same 656.3nm line, just with a tighter or looser window around it — and how much that width actually matters is argued out on Cloudy Nights, Stargazers Lounge and AstroBin, not settled by any manufacturer.
A few underlying technical facts aren’t in dispute and are worth separating out first. Narrower filters are more sensitive to focal ratio: Chroma’s own technical notes state that 3nm filters generally need f/4 optics or slower to avoid a blue-shift in the transmission line that degrades efficiency, though Chroma also sells a “fast” 3nm variant optimized for f/2.8–f/3.6 systems Chroma Hα 3nm imaging filter product listing, SKU CT-27065-2M, via Agena AstroProducts — check current price on the fast 3nm Hα variant at Agena. And the general bandwidth-vs- contrast relationship is qualitative, not a fixed multiplier: Chroma’s Technical Library describes its narrowest sets as “designed for maximum contrast and background suppression” against wider “balanced sets that offer strong contrast while maintaining higher signal throughput” Chroma Technology, without publishing a numeric bandwidth-to-exposure relationship anywhere.
“The intensity of the H-alpha light does not change significantly whether an Astronomik H-alpha filter is installed or not… the background signal in images is usually limited by the sensor’s dark current rather than light pollution, so further suppression of the sky background by reduction of the FWHM does not bring out more details.”Astronomik
In other words: the on-band signal from the nebula itself barely changes with bandwidth — a narrower filter mainly buys you a darker background, and that benefit shrinks once your noise floor is set by your sensor’s own dark current rather than by skyglow. Optolong publishes no numeric bandwidth-vs-exposure relationship at all, which leaves this a qualitative tradeoff across every manufacturer we checked, not a quantified one.
You may run into a specific exposure-time multiplier cited online for how much more exposure a 5nm filter supposedly needs versus a 3nm filter to match it. That figure does not appear anywhere in Chroma, Astronomik or Optolong’s own technical documentation — we checked each directly — and cannot be corroborated against any manufacturer source. Treat it as an unsourced community estimate, not a technical fact, and don’t use it to plan an exposure sequence.
Position A — narrower wins under light pollution. Practitioners on Cloudy Nights, Stargazers Lounge and AstroBin argue a tighter filter (3nm) suppresses more sky background, tightens stars, and raises contrast under light-polluted skies. Position B — wider is more forgiving and loses little under dark skies. The same forums also carry the counter-argument: a 6–7nm filter is more tolerant of fast optics, generally cheaper, and gives up comparatively little once you are actually imaging from a dark site. What both agree on: this is a community heuristic dispute, not a manufacturer-settled question, and the underlying f-ratio sensitivity and bandwidth-vs-contrast facts above are shared ground neither side actually contests.
Mono filter sets vs. single-shot dual/tri-band filters
The same narrowband idea splits into two very different physical setups depending on which camera type it’s feeding. A mono narrowband set is a group of individual, single-line filters — Hα, OIII, SII, each its own separate piece of glass — loaded into a filter wheel and shot one at a time, giving full control over each line’s own bandwidth (including going as narrow as 3nm per line if you want it). A dual- or tri-band filter takes the opposite trade: one filter isolates two or three lines at once in a single exposure, and it mounts on mono or OSC cameras alike, not OSC exclusively — Optolong’s own product documentation describes the L-eXtreme as built for use with DSLR, one-shot-color, and monochrome CCD/CMOS cameras Optolong. It’s most often paired with OSC, where a single exposure straight off the Bayer array is the whole appeal: Optolong’s L-eXtreme passes Hα and OIII together in a 7nm dual-band design Optolong, and L-eNhance adds Hβ as a third band Optolong; Ontario Telescope. Either way, you trade per-line bandwidth control for a single-filter, single-exposure workflow.
Which camera type is worth building around in the first place — mono’s per- line control and sensitivity edge versus OSC’s simpler workflow — is the full subject of our mono vs. OSC comparison; this article stays on what the filters themselves do, not which camera you should buy them for.
Fitting a filter into your imaging train
Two things worth knowing before you actually order one, both owned in full elsewhere. Every filter you thread into the optical path adds roughly a third of its own thickness to your required back focus, per ZWO’s published rule of thumb ZWO — the arithmetic, the worked examples and the tolerance discussion all live in our 55mm back-focus standard guide, and we won’t re-derive it here.
On mono rigs guiding through an off-axis guider, where the filter wheel sits relative to the OAG matters: practitioners on Cloudy Nights describe putting a narrowband filter ahead of the guide pickoff as a guide-star budget problem, since the OAG then has to find a star in whatever light survives a few nanometers of bandpass Cloudy Nights — our imaging-train assembly order guide covers that placement question, and the rest of the physical ordering, in full.
Which filter should you buy?
That depends on your sky’s light pollution and your camera type — two variables this article has deliberately left out, because they turn a filter-mechanics question into a buying decision, and that decision is its own article. Our light-pollution filter selection guide maps Bortle class and camera type to specific filter picks, Bortle-by-Bortle, with named products at every tier. If you already know from the sections above that you want a mono narrowband set or a dual-band filter, the products linked in this article are real starting points — the guide linked above is where the sky-darkness and camera-type math actually happens.
FAQ
Is a 3nm or 7nm narrowband filter better?
There’s no manufacturer-settled answer. A 3nm filter suppresses more sky background and is favored by practitioners imaging under light pollution; a 6–7nm filter is more tolerant of fast optics, generally cheaper, and gives up comparatively little under genuinely dark skies. The one manufacturer-sourced quantitative anchor, Baader Planetarium’s own SNR derivation, puts the gap at roughly 1.36× for its own 3.5nm-vs-6.5nm pair under light-pollution-limited conditions Baader Planetarium — treat that as one data point, not a universal rule, since it carries Baader’s own stated assumptions.
What’s the difference between a dual-band filter and a single narrowband filter?
A dual-band filter (like Optolong’s L-eXtreme) isolates two emission lines — typically Hα and OIII — at once, in a single exposure, and mounts on mono or OSC cameras alike, not OSC exclusively Optolong. It’s most often run on OSC, where one exposure captures both lines straight off the Bayer array. A single narrowband filter isolates only one line at a time and is loaded into a filter wheel on a mono camera, one line per exposure, for full control over each line’s own bandwidth.
Can I use a narrowband filter with a one-shot color camera?
Yes — dual- and tri-band filters like Optolong’s L-eXtreme and L-eNhance are commonly run in front of an OSC camera’s Bayer array and pass two or three emission lines at once, and the same filters also mount on mono cameras Optolong. True single-line narrowband filters are more commonly run through a filter wheel on a mono camera, but the filter glass itself doesn’t require mono — see our mono vs. OSC guide for the full camera-type decision.
Do narrowband filters help with galaxies?
Not much. Galaxies emit broadly across the visible spectrum rather than concentrating light in the specific lines a narrowband filter isolates, so neither narrowband nor dual-band filtering buys galaxies much light-pollution relief Urban Astrophotography. Broadband filters, or no filter at all under a dark enough sky, are the better fit for galaxies, star clusters and reflection nebulae.