Email: nexusmail at this Web site address
Latest Revision: December 11, 2003
DISCLAIMER: All this information is accurate to the best of my knowledge; if there are any omissions or errors, please let me know. This document is intended to be an overview, not the end-all-and-be-all of a given topic. If you want to find out more about a specific gadget, accessory, or thingy, consult the references listed in each section.
Light-pollution reduction filters (henceforth, LPRFs) do work -- on some things. They are not a substitute for a clear sky, but they can and do provide improved views of many objects. Furthermore, filters can be of use even in very dark skies, not just ones that are significantly affected by city lights.
Filters work by rejecting light in certain frequency ranges. As you may already know, many artificial light sources emit over fairly narrow wavelength ranges. Chief among these are the wavelengths associated with LPS (low-pressure sodium), HPS (high-pressure sodium), and mercury vapor lamps. Sodium lamps produce light centered on the sodium D line at 589 nm (yellow), with HPS covering a much wider range than LPS; mercury lamps produce a number of lines throughout the spectrum, particularly below 450 nm and above 550 nm. All LPRFs are designed to block the wavelength regions associated with these sources.
As you might have guessed, there's a catch. The LPR filter has to not only reject the undesired light, but also admit the desired light from your nice, friendly Messier objects (or whatever).
The most favorable circumstance occurs with emission nebulas; these are things like planetaries (e.g., the Ring and Helix Nebulas), supernova remnants (M1, the Veil), and other large diffuse nebulas that emit, rather than merely reflect, starlight (other good examples: the Orion Nebula, M42; the Eta Carinae nebula; and the Swan Nebula, M17). These nebulas emit light in only a few wavelengths, predominantly those of oxygen III and hydrogen alpha and beta.
By a fortunate coincidence none of these wavelengths is near significant* light pollution lines. OIII and H beta are located near 500 nm well away from most interfering lines; H alpha, at 656 nm (red) is less important visually but important in photography and sometimes in very bright nebulas like M42. Consequently, LPR filters are also designed, in addition to having very low transmittances for lines associated with light pollution sources, to have very high transmittances for these nebula lines. Thus LPR filters tend to work well on emission nebulas.
* For the technically...shall we say, "thorough," I should point out that HPS lights do put out a line very close to the OIII lines. This line is comparatively minor and does not normally affect the performance of a LPRF designed to admit the OIII region.
As already mentioned above, filters work by rejecting the light from various undesirable sources. In general, filters are designed to admit a fairly narrow region of light, called the passband (or bandwidth), around the desired lines. The rest of the visible spectrum, in general, is heavily blocked. As a result, if an object produces much of its light outside the passband, it will tend to be attenuated strongly, much as a light pollution source would be.
Sources of this sort include stars, and anything that is either composed of stars (galaxies, clusters, multiple stars) or reflects starlight (reflection nebulas, e.g. M78 and the blue portion of the Trifid Nebula).
The degree of attenuation, of course, depends on the object and how much of its light is being rejected. In particular, it is physically possible for an object like a galaxy to be attenuated less than a light pollution source, which should lead to its being enhanced. In practice, however, this does not occur often. Much of the time, in fact, the filter actually makes things worse. This is not a hard and fast rule, but it does hold enough of the time to merit strong mention.
In any event, enhancement of these objects is not as great as it is for nebulas. This last statement holds for all currently available LPRF's.
For many years, the dominant company in this market was Lumicon, which had (at last count) at least 5 different LPRFs, all with different technical specifications and performance on various types of objects. After a brief closure in late 2002, Lumicon re-appeared (apparently under totally new management by this company). Most large telescope/accessory manufacturers or resellers, such as Meade and Orion, offer some LPRFs. Some companies that specialize in filters generally also make LPRFs.
A couple of the big manufacturers or distributors:
The most important technical specifications for a filter are:
For determining general usage, the first parameter is the most important, as this detail tends to dominate the filter's effectiveness (or lack thereof) on various objects.
LPR filters are characterized as being either broadband or narrowband, depending on how wide the bandwidth around the desired lines is. These descriptions are a bit vague, but for practical purposes, a bandwidth of 30 nm, measured at the "half maximum" or 50% transmission level, is a good cutoff for distinguishing the two types. In other words, if a filter has a bandwidth appreciably smaller than 30 nm, it can be considered "narrowband". Here are some figures taken for some popular filters; bandwidth values refer to the width of the passband at 50% transmission.
|Filter Brand Name||Type||Bandwidth|
|Orion Sky Glow||Broad||85nm|
|Lumicon Deep Sky||Broad||68nm|
|Celestron Type A||Broad||47nm|
|Lumicon UHC (Ultra High Contrast)||Narrow||27nm|
|Lumicon OIII (Oxygen III)||Narrow||11nm|
|Lumicon H-Beta (Hydrogen Beta)||Narrow||8 nm|
(figures taken from Astronomy, Feb. 1991, p. 77. The specs for the Lumicon and Orion filters, at least, appear to be much the same in 2003.)
Extremely narrowband filters, such as the Lumicon OIII and H Beta, are sometimes called "line" filters, since they only admit one or two emission lines and reject nearly everything else. These are the ultimate for enhancing dim objects -- provided said objects happen to emit an appreciable amount of that particular line.
The capabilities of various filters can be summarized as a function of object type:
|Rating||Quality||Summary of filter effect|
|1||Very poor||This object generally looks much worse with the specified filter than without.|
|2||Poor||The object generally looks noticeably worse with the specified filter than without|
|3||Fair||The object normally looks about the same or possibly a little better|
|4||Good||The object normally looks appreciably better|
|5||Excellent||The object normally looks dramatically better|
|6||Outstanding||The object normally looks dramatically better even in conditions of very low light pollution (e.g., dark rural
Here are how various types of filters commonly available today rank on this scale:
|Object Type||Broadband Filter||Narrowband Filter||Line Filter|
|Planetary Nebulas||3-4||5-6||6, depending on filter*|
|Other Emission Nebulas||3-4||5-6||6, depending on filter*|
|Stars, star clusters||2||1-2||1|
* Some emission nebulas have emission profiles that favor certain line filters over others. In general, planetaries, like M57, M97 (Owl Nebula), and the Helix Nebula, tend to be rich in OIII emissions, so an OIII filter will rate "6" with almost all planetaries. By contrast, some non-planetary emission nebulas emit relatively more H-beta than OIII so for those the H-Beta would rate "6" and the OIII would rate lower (possibly as low as "1"!).
This can be summarized more compactly. Compared to broadband filters, narrowband and line filters provide:
You would thus want to use something like the UHC, OIII, or UltraBlock for viewing dim planetaries (e.g., the Helix) under difficult conditions, but using one on a galaxy or globular cluster is not a swift move!
For a very detailed comparison of the performance of 4 popular filters (Lumicon Deep Sky, UHC, OIII, and H-Beta), read David Knisely's filter comparison article.
If you have only enough money for one filter, I suggest getting a good narrowband version, like the Orion UltraBlock or Thousand Oaks LP-2 (Narrowband). One of these will optimize viewing of nebulas under all sky conditions.
If you have enough to buy two, a narrowband filter plus a line filter would be a good combination. The OIII is the most versatile of the line filters and is very highly recommended; the H-Beta tends to be a specialty filter (see below for more details). For objects where the OIII filter isn't so great, you'll often get good results with the regular narrowband filter.
You might also consider a "complementary pair" -- one broad, one narrow -- but in this case I'd recommend trying some filters first (e.g., at an astronomy club meeting) before committing to a purchase. This is particularly true for broadband models, since the benefits are more subtle. In fact, many people find the broadband versions disappointing for applications other than photography.
Read the articles in Astronomy, Feb. 1991 and Sky and Telescope, July 1995; although a little old, the overall information they contain is still valid. They describe the experiences of several experienced observers who tried many popular filters on various objects.
Some filters are "specialty" filters, which work well on a small (but possibly important) class of objects. The most notable example of this is the Lumicon H Beta filter, which strongly enhances a few emission nebulas (notably IC 434, the emission nebula surrounding the Horsehead in Orion), but does not work as well as other narrowband filters on most others. In fact, the H Beta is superior to the usual narrowband filters on a very small class of objects, unlike the OIII, and so is not a good first purchase. The H Beta is the one to get after you've tried most of the other varieties for a while.
You can also get something loosely called a "comet" filter, which passes emission lines seen strongly in some (though not all) comets. Filters like these are also not recommended for a first purchase.
Light pollution isn't limited to cities. There are naturally occurring sources of light (no, I mean besides the Sun and Moon, etc.), most notably something called "airglow" or "auroral" glow. This is fluorescence produced by the air itself (more precisely, by molecules in the upper atmosphere). As with many light pollution sources, airglow predominatly occurs at a small number of frequencies, particularly at 465, 558, 630 and 636 nm. Since many filters block these regions, or at least reduce it significantly, they can be helpful even in very dark skies with little artificial light. The line at 558 gets munched by all filters; most narrowband filters will also eliminate the one at 465 nm as well. Judging from transmission tests, e.g., those found in the Feb., 1991 issue of Astronomy, filters that strongly attenuate the red (630 and 636 nm) airglow lines include the Orion Ultrablock and (to a noticeably lesser extent) the Lumicon UHC and DeepSky filters.
Dim nebulas seen from dark skies frequently benefit from filtration because of the removal of this natural airglow. This is most noticeable on large, low surface brightness nebulas such as the Veil Nebula and many planetaries, e.g., Jones 1 or any of those little obnoxious things labeled with a P-K number in Uranometria.
Another tip: Unlike ordinary color filters, which work by absorption, LPRF's are highly reflective. As a result, observing with a LPRF requires some extra care in screening out stray light. Light that enters the eyepiece from the outside, e.g. around your eyes, will be reflected back by the filter and interfere with viewing.
Since LPRFs work on the same principles as interference filters, the exact position of the bandpass depends on the angle the filter's surface makes with the incoming light. (You can demonstrate this for yourself by taking a LPRF, looking at a light through it, then watching its color change as you tilt the filter.) This is not a problem for normal use, with the filter inside the eyepiece, but if you're using it to "blink" for nebulas (i.e., rapidly moving the filter in front of and away from the eye end of the eyepiece), you'll want to make sure the filter is parallel to the eye lens.
Furthermore, if you're "blinking" with a wide-angle eyepiece (a Nagler, say), light emerging from the edges of the field of view might experience enough of a path length difference at the filter than light coming from the center. This applies only when the filter is used after the eyepiece, as in blinking; there is no problem when the filter is mounted normally. In theory this could lead to a change in the position of the bandpass, making nebulas harder to see at the edge of the field of view. I haven't observed this effect myself; owners of wide-field eyepieces and filters are encouraged to try this and comment.
Finally, remember that filters never make things brighter; they make everything dimmer, to varying degrees. What happens, of course, is that things like nebulas are scarcely dimmed at all, whereas the sky background (containing light from artificial sources, airglow, etc.) is greatly darkened, thus leading to improved contrast. However, the resulting image may still be fairly dim, especially in a smaller telescope. Patience is advised. Before hunting dim, challenging objects with a filter, practice on easier targets first. Try a big bright nebula like the Lagoon or M42 to get a feel for how much the filter improves things.