Narrowband Filters: What SHO and HOO Mean

Dual-band, tri-band, 7nm versus 3nm, and why the Hubble palette makes hydrogen green. The decisions behind a narrowband image, in order.

Narrowband is the biggest single step up available to most people, and it is usually explained backwards, starting from palettes and only at the end, if at all, getting to why any of it works.

It works because emission nebulae do not emit light across the spectrum. Specific atoms emit at specific wavelengths, and those wavelengths are narrow. The rest follows from that.

Three lines

Hydrogen alpha, 656.3nm. Deep red. By far the most abundant, and the reason most nebulae photograph red. If you own one narrowband filter, it is this one.

Oxygen III, 500.7nm. Blue-green. Traces hotter, more energetic regions: planetary nebulae, supernova remnants, the cores of star-forming areas near young hot stars.

Sulfur II, 671.6nm. Deep red, close to hydrogen alpha in wavelength but far rarer in the sky. Faint and slow, and the reason narrowband projects take many nights.

A filter that passes only a narrow window around one of these lets your target through and blocks almost everything else. Streetlights, moonlight and airglow emit very little inside a 3nm window at 656nm.

So under narrowband, a city sky and a rural sky are much closer than you would expect. They are not identical, but the gap narrows from hopeless to slower, and moonlit nights become usable. There is more on what this does and does not rescue in what you can photograph from a light-polluted sky.

It does nothing for galaxies or star clusters, which shine in broadband light. No filter separates a galaxy from a streetlamp.

Which filter, for which camera

This is where most of the money goes.

One-shot color camera or DSLR: a dual-band or tri-band filter. A color sensor already has red, green and blue filters over its pixels. A dual-band passes hydrogen alpha (landing on the red pixels) and oxygen III (landing on green and blue), so a single exposure captures both. You get two channels from one night.

Monochrome camera: individual filters, one at a time. Shoot hydrogen alpha for several hours, swap, shoot oxygen III, swap, shoot sulfur II. It is slower and more work, and it gives full control over the result, because every pixel collects light rather than a quarter of them.

A one-shot color camera with a dual-band filter gets far more done per clear night. A mono camera with a filter wheel produces better images per hour of integration and costs more in money, nights and patience. If you are moving up from a beginner setup, take the dual-band route first.

7nm, 5nm, 3nm

Narrower blocks more sky, which sounds strictly better. It is not.

Narrower helps when your sky is bright, the Moon is up, or you have plenty of clear nights to spend.

It costs you in three ways people tend to find out after buying. It cuts signal from the target too, so exposures get longer. Bandpass filters shift toward the blue as light hits them off-axis, so on a fast telescope (f/4, f/3) a 3nm filter can shift off the emission line entirely at the edges of the frame, giving you a dark ring and less signal than a wider filter would have. And they cost considerably more.

Rough guidance: 7nm for f/5 and slower under a moderate sky. 5nm as the general-purpose choice. 3nm only if you are under a bright sky, at f/5 or slower, and know you need it. Check the manufacturer’s stated focal-ratio limit; the good ones publish it.

Then the palette

You now have two or three grayscale stacks. Nothing about them is inherently colored; each records how much light arrived at one wavelength. Assigning colors is a choice.

HOO. Hydrogen to red, oxygen to both green and blue: two channels, three outputs. It gives the red-and-teal look and is close to what the eye would see if it were sensitive enough. It is the natural choice for a dual-band filter, and the place to start.

SHO, the Hubble palette. Sulfur to red, hydrogen to green, oxygen to blue. Needs a mono camera and three filters.

This is the part that confuses most people. In SHO, hydrogen goes to green, and hydrogen dominates almost every nebula. So a straight SHO combination comes out overwhelmingly green. Every gold-and-blue Hubble-palette image you have admired got there by pushing that green into yellow and gold afterwards. That is a standard, expected step rather than a fudge, and nobody’s raw SHO stack looks like the finished picture.

Other combinations such as HSO, foraxx and dynamic blends exist and are all equally legitimate. None is more real than another. False color applied consistently, and labeled, is normal scientific practice.

What this costs you

Two things to know before committing.

Stars go strange. Stars emit broadband light, so a narrowband filter blocks most of it. What survives is whatever falls inside your passbands, and for a hot blue star and a cool red star that can be very different fractions. The result is small, oddly colored, often magenta stars. The usual fix is to shoot short broadband exposures separately for the stars and combine them with the narrowband nebula, which is where star reduction and starless processing comes in.

Everything takes longer. You are throwing away 99% of the light reaching the sensor. Sub-exposures triple or quadruple, and a full SHO project takes several nights rather than an evening. See how long your subs should be.

For southern skies

Northern-hemisphere target lists dominate the internet, and they leave out some of the best narrowband subjects in the sky. From southern latitudes the Carina Nebula is brighter and larger than anything the north has, the Running Chicken and the Gum Nebula are both strong emission targets, and the Magellanic Clouds sit high with the Tarantula Nebula inside one of them.

If you image from Australia, South Africa, Chile or southern Brazil, you have the better half of the narrowband sky, and the standard lists are not written for you.

The short version

  • Emission nebulae emit at narrow wavelengths. Streetlights mostly do not. That is the whole trick.
  • Color camera: dual-band. Mono camera: individual filters. Start with dual-band.
  • Narrower is not always better. Check your telescope’s focal ratio against the filter’s limit.
  • HOO for two channels, SHO for three. Both are choices, and neither is more real.
  • SHO comes out green before you correct it. That is expected.
  • None of this helps a galaxy.

Try it on your own data

Akastroid runs its processing on your own frames and logs each step it took.

Download Akastroid