Less is More: Narrowband Imaging

Joe Shuster

Newbies to observational astronomy quickly learn that the full moon is a daunting adversary that blots out contrast and detail in the faint fuzzy objects we pursue. If you're getting your feet wet in astrophotography, you've probably noticed that the bright moon does an even nastier job on your photos. The light pollution in your sky is amplified by the sensitivity of the camera. Moisture in the air reflects the moonlight and gives a smoky look to the sky through the camera. And nasty gradients start harmlessly at one end of your image and increase until they fog the details of your image.

All in all, the moon is a much bigger enemy to the imager than the observer. But what if there was a way to neutralize the moon and even neutralize the local light pollution we're constantly bemoaning? Wouldn't you like to be able to image even during the 13 brightest days of the lunar month from the 9-day old moon to the 22-day old moon? Well of course there's something you can do and involves doing something that is totally the opposite of what we normally do in astronomy: You want to throw away light. Huh? Yes, throw away light. Whah? Yep, get rid of a LOT of light. Let me explain...

Conventional progress in astronomy is all about collecting lots of light: We get big lenses and mirrors - more surface area means more light collected. In photography, we use those big optical surfaces with cameras that can stare at the sky for a long time to pile up the incoming light. When we're collecting light from a nebula or stars or dust the goal for better pictures is to use bigger optics, more sensitive cameras, and longer exposures. That's the philosophy from light pollution-free areas and (relatively) moonless nights.

When there's light pollution and/or a bright moon, the rules change. Bigger optical systems amplify light pollution - man- and moon-made. Long exposures on sensitive cameras do the same thing. And what's worse is that the sky mess isn't uniform. The area of the sky closest to the terrestrial or lunar light source is brighter than the areas that are farther away. The camera can catch this gradual brightening and dimming - a "gradient" - much better than your eye can detect. So collecting a big bunch of "bad" light makes your photos much poorer instead of richer.

The solution many astrophotographers use is to be selective - VERY selective - in our light "shopping". Discard all the bad light, and keep only the very good light. We'll skip the physics lesson, but the most popular lights to collect are Sulphur II (SII), Oxygen III (OIII), Hydrogen beta (Hbeta) and the most popular of all, Hydrogen alpha (H-alpha). These specific wavelengths of light correspond to different atomic states of different elements. To collect only specific light requires a filter and a small detour to explain terminology.

Normal astronomical filters1, like an infrared blocker or primary color filters will allow a fairly large part of the optical spectrum to pass through the filter. The part (or "band") of the spectrum is wide, so these are called "broadband filters". On the other hand, when we want to be snobbish about the light we want to collect we choose a filter that only allows a thin part of the spectrum onto the camera. So these specialized filters are called "narrowband filters". Narrowband imaging is quite popular these days because it allows astrophotographers to take photos regardless of the phase of the moon and it can substantially eliminate the effects of your local light pollution. No more hiding out during bright phases of the moon and no more chagrin at your local light pollution.

In this article, I want to concentrate on H-alpha imaging, first because it's so popular (in the universe and among astrophotographers) and second because it's important to understand the limits of H-alpha imaging before you get too excited about it.

Picking your targets: So what objects benefit from H-alpha filtering? The H-alpha light comes from excited hydrogen. Certainly stars have a lot of that so stars show well through an H-alpha filter. And other types of energy can excite clouds of hydrogen gas, so clouds surrounding star creation and star destruction activity are also candidates. What we see in objects like the Eagle Nebula, the Orion Nebula and the Lagoon Nebula are clouds of gas surrounding locations that represent forming stars. The energy of these stars excites the gas that then emits light. So these nebulae are a called "emission nebulae". Emission nebulae are VERY GOOD candidates for H-alpha filtering. The light they send out is broadcast in one or a few very narrow parts of the spectrum.

On the other hand, objects that are merely reflecting light usually have light that is broadcast along much larger parts of the spectrum. Reflecting nebulae are poor candidates for H-alpha imaging. The Running Man Nebula (NGC1977) and The Pleiades (M45) are famous reflection nebulae.

There are composite objects like the parts of the Veil Nebula (including NGC6992 and NGC6940), and the Trifid Nebula (M20) that have an emission nebula that illuminates a reflection nebula. You can use H-alpha filtering to capture part of the object, but the reflection part of the nebula will need other techniques to be captured.

Some galaxies have strong areas of H-alpha emissions (like M82 and M33), but overall, galaxy images don't benefit from H-alpha photography. Like galaxies, there are some planetary nebulae with portions of H-alpha light, but again, the reflection part of the nebula will not be seen through the filter.

The moon and other solar system objects don't benefit from H-alpha imaging. H-alpha light is significant in solar photography, but YOU MUST USE SPECIALIZED SOLAR FILTERS OR YOU CAN DAMAGE YOUR EQUIPMENT OR CAUSE SERIOUS PERSONAL INJURY. In this article we're discussing nighttime H-alpha images (otherwise the phase of the moon wouldn't matter!).

So, the primary targets for H-alpha imaging are the emission nebulae and the open clusters that include an H-alpha gas cloud (for example, M16 - the Eagle Nebula).

Camera matters. Not all cameras are good at collecting H-alpha light. Many dedicated astrophotography cameras have good or excellent sensitivity to the H-alpha wavelength (6463 angstroms). Consumer digital cameras sometimes contain a chip that is sensitive to H-alpha (as well as infrared) light, but the manufacturers mount a blocking filter to help get a good color balance for everyday photos. So the normal pocket digital camera won't do the job in H-alpha.

The popular DSLR's (e.g., Canon 350XT and Nikon D-70) have the same problem, but there are home modifications that can be made to the camera to remove the manufacturer's filter and allow the H-alpha light onto the chip. Sometimes, this renders the camera unusable for routine photography and sometimes you can correct the colors using external filters or special settings and processing. One company, Hutech Astronomical Products, even sells fully modified Canon and Fuji cameras to permit H-alpha photography.

By far, the best candidate for H-alpha photography is a dedicated astronomical camera. Cameras from SBIG, Starlight Xpress, FLI, Meade, Orion will almost always have decent H-alpha sensitivity. The easiest way to find out if your camera has good H-alpha sensitivity is to find the spectral response specifications or just use Google to search for examples of imaging in H-alpha. The exact sensitivity varies with the camera models. For example, my SX MX716 has great H-alpha sensitivity while its cousin the MX916 has fairly poor sensitivity.

Mount requirements. So you've decided your camera will gladly accept H-alpha light, you like the choice of imaging objects and you think you're ready to go. But hold on - the mount you will use needs to do a little more work than you are accustomed to. Here's why: In normal, multi-color light imaging (for example with a DSLR), you collect light from a wide spectral band - about 400nm wide. H-alpha filters come in bandpass widths of 3nm to 13nm. So you are only collecting 1/30th to 1/130th the amount of light. This means you need to take longer exposures (or many, many short exposures) to compensate. That image you captured with 10 exposures of 30 seconds (5 minutes total exposure) will need 30 to 130 TIMES more exposure - 150 minutes to almost 11 hours exposure. There are processing techniques so that narrowband imagers don't necessarily endure these marathon exposures, but the bottom line is be prepared for many and long exposures.

This requirement puts a lot of pressure on your mount. Getting smooth tracking in a short exposure is (somewhat) easy. But in a 4, 5, 8, 10 minute exposure? Only the best mounts can do unguided images that long. If you have an autoguider setup you will get a lot of use out of it during H-alpha and other narrowband imaging.2

Another mount-related aspect is the quality of your alignment. For sequences of short exposures any errors in your alignment won't cause too much field rotation. However when your exposure length get longer - 4 to 10 minutes - you might notice that the stars on the edge of the image seem to be more oblong compared to the round stars in the center. This is the effect of field rotation. A poor polar alignment (or a good alt/azimuth alignment) will show arcs of stars - small arcs in the middle and larger arcs as you get farther from the center.) If the polar alignment is good (but not perfect) you don't see the effect much in the middle of the image, but nearer the edge, the short arcs make the stars a little squat. The solution?? Put more effort into refining your polar alignment if you want the long exposures needed for good H-alpha (and other narrowband) images.

Filters. Narrowband filters, including H-alpha, are like typical colored filters. They come in 1.25" and 2" sizes (priced accordingly). The filters have standard threads to attach to most cameras. Like normal filters, narrowband filters can be used in color filter wheels and strips.

Each filter has a specific bandwidth and bandpass center. The H-alpha filters are centered on the 65633 angstroms or 656.3 nm. The bandwidth of the filters varies from 3nm to 13nm. The 3nm filters have a very, very narrow bandwidth and gets only the purest H-alpha wavelength and normally have a higher price. This bandwidth results in really small, sharp stars. On the other hand, the 13nm filter has a more "open" bandwidth and gives slightly larger (but still small!) stars and fractionally less contrast compared with the narrower filters.

The masters of narrowband imaging prefer the narrowest bandwidth forms. They feel that the precise exclusion of all other light makes for better photos. Of course, their images take about 4 times as long as images with the wider filter because they need to compensate for the narrower slice of the spectrum. Most of these experts have excellent mounts and automated controllers, so some extra hours of exposure aren't a burden to them.

One other consideration when selecting a bandwidth is "drift". The coatings used for narrowband imaging presume that your optical system operates within a specific range of f/-ratios. Often, the comfort zone is about f/4 to f/11. As you go outside that range, the filter's "center" will shift red-ward or blue-ward. If your filter has a narrow bandwidth and you use optics outside the comfort zone, the filter can totally miss the targeted light. The wider bandwidth filters have the same shift, but being wider they won't miss the goal light.

So if you plan to use optics outside the comfort zone (for example, very long focal length telescopes or very fast camera lenses), you should stick to the wider bandwidth filters.

Other filters, other applications - We only touched on the other kinds of filters for narrowband imaging. Each has a use for a specific kind of light that corresponds to specific astronomical environments and events such as star formation, star destruction, etc. You are welcome to explore the use of other filters via on-line resources, but I recommend starting your imaging with H-alpha to get the best choice of targets and ease of imaging.

Conclusion and Resources: So, do you want to get out and take images regardless of the phase of the moon or even if your sky isn't perfectly dark? Can you mount handle longer exposures? Can your camera record the fun parts of the spectrum, like H-alpha?

If so, buy a nice H-alpha filter and give narrowband imaging a try.

You can find great examples of narrowband imaging at Richard Crisp's website -
The Starizona website has a good discussion of narrowband imaging located at -
There is a Yahoo! group dedicated to narrowband imaging located at -

You should also check with astrophotography groups for your specific camera to see what tips for narrowband imaging apply to your equipment.

1Filters come in two forms: "pass" and "block". A pass filter allows specific light to pass. So a red filter allows red (but no other) light to pass. A block filter allows all light except a specific light to pass through the filter. An "IRB" blocks the infrared light, but allows all other wavelengths. Unfortunately, the common terminology for filters is not specific, so a reference to an "infrared filter" could mean a filter that blocks infrared light or a filter that passes only infrared light. For this article, the term "H-alpha filter" means a pass filter that only allows light near the H-alpha wavelength.

2There is an upside to the laughably long exposure requirements for H-alpha imaging. When you kick off a 90-minute sequence of images (under computer control), you can use that time effective for other tasks. Some people do observing with a separate set of optics. Me? I take a very nice, short nap!

3In New Mexico, most highways have a 3-digit designation, but there is one 4-digit highway: 6563 also known as the "Sunspot Highway". It leads to the National Solar Observatory at Sacramento Peak.

Published in the November 2006 issue of the NightTimes