Narrowband nebular type filters for improving contrast on certain types of non-stellar deep-sky objects, are discussed frequently in the newsgroup sci.astro.amateur and advertised in the catalogs of such manufacturers as Orion Telescopes and Lumicon, prominent manufacturers / dealers for numerous models; and the largest amateur telescope makers often sell their own "house brand" of eyepiece filters. It might be useful to offer our perspective on the use of filters, to provide a middle - ground between the unreservedly enthusiastic advertising claims of retailers, and the conflicting opinions of diverse amateur users. The author (now retired and not involved in the telescope industry) worked for two companies as a consultant in the testing and marketing of nebular filters; developed software for one company that served as a tool for determining usage; and has employed them extensively as an amateur astronomical observer since the 1980s.

Like "high end" audiophile gear, eyepiece filters of each type and brand have their enthusiastic advocates, as well as their firm denouncers. What is the general truth about them that few would really dispute? In trying to answer this question, we offer the following article, based on information originally published in our software programs REDSCOPE and EYEPIECE 2.0.

Warning: Not all eyepieces (especially very inexpensive or oddball ones) have appropriate and accurately cut threads for eyepiece filters! DO NOT FORCE the filters onto your ocular if in doubt! The good-to-excellent quality range 1.25" and 2.0" barrel eyepieces sold by the above-mentioned companies, as well as other fine astronomical specialty dealers, will provide the best results.

NEBULAR FILTERS FOR DEEP-SKY OBSERVING & PHOTOGRAPHY:

Four general types are of greatest use to amateur visual observers:

· Light Pollution Filters;

· General Narrowband Nebular Filters;

· Oxygen Nebular Line Filters; and

· Hydrogen-Beta Nebular Line Filters.

Astrophotographers often use this narrow-band filter to register the red wavelengths of hydrogen:

· Hydrogen-Alpha Nebular Line Filters.

Each of the visual filters will be discussed in turn, from information in our software program EYEPIECE, augmented by sources of information I later gleaned from the Net. The chart below of visible light wavelengths and filter performance will help illustrate the various nebular lines of deep-sky objects that may be viewed or photographed, and their transmission by specific filters which, simultaneously, cut off unwanted wavelengths.

The "visual filters" discussed in this article are designed to complement the performance of the dark-adapted human eye, whose response in dim light is largely to the green-blue range of the color spectrum (shown in the chart in this article on the "HyperPhysics" website.)

BROADBAND LIGHT-POLLUTION-REDUCTION (LPR) FILTERS:

This eyepiece or camera lens filter has a relatively minor effect on the bandwidth of visible light: it enables enhanced results in viewing or photographing images of celestial objects whose faint light is obscured by the stray photons from street lamps and from the dim but perceptible skyglow caused by excited atoms in earth's atmosphere.

The typical broadband LPR filters have high transmission (which should ideally achieve better than 90% efficiency) at all visual light wavelengths except in a region centered around 560 nanometers, which contains the frequencies of the yellowish light of sodium- and mercury-vapor streetlights, and airglow. Other wavelengths radiated by oxygen and hydrogen are little affected by an LPR filter, and views of gaseous clouds like the "Crab" (M-1), "Lagoon" (M-8), and "Great Orion" (M-42) nebulae are somewhat enhanced in detail and contrast; photos of these objects may be recorded with longer exposures before the skyglow "washes out" the background with stray light. Galaxies with extremely low surface- brightness (such as NGC-6822, "Barnard's Galaxy") may sometimes stand out better against the soft background light of skies afflicted with the glow of distant streetlamps. And reflection nebulae, such as M-20 or M-45, featuring the scattered broad wavelengths of starlight, may be enhanced, since the LPR filter will improve the contrast of bluish wavelengths to which the eye is particularly sensitive.

Since LPR filtering is broadband and relatively gentle, most of the light of stars, clusters, and galaxies which radiate a wide range of visible colors (and generally appear faint white or grey to the eye) will be quite clearly seen, often with contrast enhancement as well. Sky fogging of photographic exposures will be reduced, permitting exposures as much as three times longer without an LPR filter: our old DOS freeware program "Eyepiece" will help you calculate the exposure times and sky fogging factor for appropriate objects -- if your telescope focal ratio is in range for efficient deep-sky photography.

Some beginning visual observers have complained that this type of filter has a weak enhancement ability compared to the OIII or hydrogen line types. This author has found, however, that the LPR filter may be used at higher magnifications (which create a small exit pupil) than the narrow band filters: you may find that experimenting with magnification will yield a particular exit pupil size for a given object that improves the enhancement factor, allowing faint nebular details to show up with good contrast compared to using the LPR filter at low power.

Advantages: Greatly reduces the skyfogging of long-exposure astrophotography; somehat enhances contrast on many objects without extinguishing stars severely. Disadvantages: Gentle and subtle enhancement may not be noticed by beginning observers, at lowest magnifying power, or in all sized aperture scopes and sky conditions.

GENERAL NARROWBAND "HIGH CONTRAST" NEBULA FILTERS:

One of the most versatile of the contrast enhancing visual nebular filters, a narrowband nebular line filter -- now commonly called UHC, or ultra-high contrast, using the terminology originally devised by Lumicon -- has a narrower bandwidth than an LPR filter. Though the typical broadband LPR (light pollution rejection) filters described earlier will transmit all but a narrow band of wavelengths that reduces light pollution and skyglow, the general narrowband nebular line filter blocks a broad band of wavelengths while passing only a narrow range, peaking sharply just above 500 nanometers (in the bluish-green wavelength range of light). Some brands and models include also the long wavelengths of red light; other makes cut those wavelengths off, as most observers' dark adapted sight is not sensitive in the red region.

Those light wavelengths transmitted by the general nebular filter are radiated by ionized oxygen and atomic hydrogen, which fortunately happen to be in the blue-green light region to which the dark adapted human retina is most sensitive. Some extended nebulae, like the "Veil" in Cygnus (NGC-6992), are spread over such a wide area of the eyepiece field that they are often invisible to novice astronomers, or may never seem to resemble in an eyepiece view -- even employing the largest of scope apertures -- remarkable long- exposure observatory photos.

However, a narrowband nebular line filter, even under moderately light polluted skies, will enhance images of such nebulae with a clarity and contrast not otherwise obtainable. A number of different brands are available: if possible, obtain a filter with a high transmission factor in the bandpass region (at least 90% is desirable).

In heavily light polluted skies, the nebular line filter can make the difference between a nebula being utterly INVISIBLE or being unmistakably VISIBLE, even with scopes as small as 80mm aperture. Of course, the exit pupil should be in the proper range for effective filter use, as shown by our program EYEPIECE.

Advantages: Greatly enhances the contrast of many nebulae; some may be invisible without the filter. With certain objects, such as small diameter high surface brightness planetary nebulae, such a filter type may be used with a small exit pupil and help identify the object in a field of view that encompasses many stars. CCD imagers occasionally find that the filters are usable with appropriate objects. Disadvantages: With smaller scopes, stars are severely extinguished, and with any scope, use of a very small exit pupil may cause such extinction that objects lose visibility. Conventional film photographic use is virtually impossible, aside from the occasional bit of experimentation, since the "filter factor" causes great loss of light, leading to the need for extremely long and impractical exposure times.

OXYGEN NEBULA LINE FILTERS:

A contrast enhancing deep-sky visual filter (currently made by Lumicon Company, Orion, Astronomik, Baader, Thousand Oaks Optical, and perhaps a few other companies), the oxygen line filter has its high transmission region tuned to pass the light wavelengths radiated by doubly ionized oxygen near 500 nanometers: the so-called "forbidden lines" of radiation that can occur in the near vacuum of deep space but not under earthly conditions of higher pressure. The oxygen line filter dramatically reduces skyglow, while transmitting wavelengths emitted by ionized oxygen: planetary nebulae are particularly enhanced.

Observers of planetary nebulae, such as the large- diameter "Helix" or the small, nearly stellar IC-2003, will find the O-III filter virtually indispensible.

In addition, the oxygen line filter provides the highest possible contrast enhancement for viewing many types of large diffuse nebulae in sites that suffer from extreme light pollution. The oxygen line filter is, however, not designed for photographic use, since its "filter factor" of attenuation will increase exposure times beyond a useful range, and will block wavelengths to which most film types are sensitive.

Advantages: Sometimes the only way to identify very small diameter planetary nebulae in a complex field of view with many stars is to use an OIII filter. Gaseous nebulae that radiate the nebular lines of ionized oxygen will have greatly enhanced contrast with appropriate exit pupil. CCD imagers occasionally find that the filters are usable with appropriate objects. Disadvantages: Extreme extinction of image may occur with small exit pupils, especially in small aperture scopes. Essentially useless for conventional film photography, aside from those interested merely in experimentation.

NARROWBAND HYDROGEN-BETA LINE NEBULAR FILTERS:

The hydrogen-Beta filter (currently made by Orion, the Lumicon Company, Thousand Oaks Optical, and Astronomik), is designed to enable the visual astronomer to accomplish the almost hitherto impossible task of actually seeing the faint and elusive Horsehead dark nebula near the star Zeta Orionis in the constellation Orion.

In years past, numerous experienced observers (such as the late Walter Scott Houston) reported unfiltered observations of the Horsehead with small richfield telescopes, but in this era of urban sprawl and light pollution, the task is becoming difficult or even impossible.

The name "hydrogen- beta" refers to the Balmer Beta nebular emission line of light wave energy at 486.1 nanometers radiated by atomic hydrogen (excited atoms of hydrogen in the low pressures of deep space). This filter has a very sharp high transmission response to the blue-green light of this specific frequency: the Beta line is the weaker line of atomic hydrogen, the stronger being the h-Alpha or Hydrogen I line in the red color region, centered at 656.3 nanometers.

Unfortunately, the human eye can see very little of weak red light, so the rich reddish hues of a hydrogen nebula are largely undetected by eye through a telescope, though they register strongly on film. However, the weaker blue- greenish h-Beta line is much more readily detected by eye: thus, for example, the energy radiated by the nebula IC-434, located behind the "Horsehead" dark nebula, along our line of sight, will show up to the eye (under the best viewing conditions) as a faint greyish smudge, with the "Horsehead" being a dark (or darker) spot against the faint glow. This non-linear detection of color wavelengths by the dark adapted human eye is called the "Purkinje" effect, after its discoverer, the 19th- century Czech physiologist J. N. Purkinje.

To see the "Horsehead", the blue-green hydrogen-Beta wavelength from IC-434 can be detected, as can the dark cloud of interstellar matter arranged in front of it, if all "competing" skyglow is reduced. The authors of this website have observed the "Horsehead" in Moonless, clear skies of at least 5.8 to 6.0 limiting magnitude (preferably darker), with this filter employed in scopes as small as 3 - 4 inches in aperture, by following the recommendations of the program EYEPIECE for exit pupil size in carefully selecting an eyepiece for appropriate magnification, when the object is nearest the meridian and above the rough air at the horizon.

In addition to the "Horsehead", the hydrogen-Beta filter can also enhance the contrast of a few other extended objects like the "California" Nebula (NGC-1499) in Perseus, faint "H-II" regions in the arms of some bright spiral galaxies like M33, or the Cygnus objects the "Cocoon Nebula" (IC-5146) and the nebulosity of "Campbell's Hydrogen Star" (PK64+5.1, a planetary nebula surrounding star BD +30 3639 at 19h 32m 47.5s RA, +30d 24m 20s DEC.)

Advantages: Some observers find that the H-Beta filter is the only way to really see and appreciate the Horsehead nebula. Disadvantages: Useless for deep-sky film photography. Filter is costly considering its limited utility with only a relatively small number of objects, such as those mentioned immediately above. Furthermore, it works only at the largest exit pupils with very small aperture telescopes; otherwise, severe image extinction will occur.

FILTER SPECTRAL COMPARISONS:

Full Color Transmission Response:

If you could view the response of each filter with "daytime" (color or photopic vision) they would have response similar to our simulations, above. Though our frequency passband color charts for each filter type don't seem to look much different in the amount of green and blue light wavelengths transmitted, in the telescope the difference between a UHC, an O-III, and an H-beta type filter will often seem to be enormous, as many celestial nebulae transmit a very narrow range of critical frequencies. Cutting off or including them will radically change the image contrast. The purpose of the filter is indeed to DIM light that is undesired, and which does not carry 'information' from the object you wish to see.

When your eye is dark-adapted, you employ scoptopic vision, which uses only the retina's luminance-sensitive rods. In the transition zone between photopic and scotopic vision is a range called mesopic vision (explained in this article.) At a light level equivalent to twilight or dusk, both the rods and cones are active, with less color sensitivity but more perception of faint light. Depending on the object you are looking at in your telescope (very brilliant stars, strong dense nebulae, or extremely dim and faint objects) some aspect of each of the three types of vision may come into play: thus, some people can see traces of color in certain nebulae, if the view in their telescope is bright enough. The comparison below will illustrate the effect of filters seen with dark-adapted vision, a large telescope, and fairly bright celestial objects. Some color may be detected (as Bill Ferris notes in his summary of a 1955 scientific paper about a mesopic vision study, in a post to the Yahoo technical group "Amastro".)

Mesopic Vision Color Sensitivity:

The human eye's response to very dim light, when fully dark adapted and using scotopic vision, has essentially little or no color detection ability. Very dim celestial objects will be most visible when their light is radiated in the region shown below, comparing the color/photopic vision, top, to the night/scotopic vision, bottom.

Day (Photopic) versus Scotopic (Night) Vision Sensitivity:

Novice telescope users should remember that in dim light, almost every 'extended' celestial object (galaxies, nebulae) will look only gray to the eye. Though faint light color wavelengths are striking the eye's retina, it is sensitive at faint light levels only to luminance differences, not colors. So, the nebular filters do not generally change the apparent color of very faint objects -- to the eye -- though they will do so with color photography, as discussed below.

PHOTOGRAPHIC TESTS OF FILTERS:

Years ago, when one of this website's authors (Waldee) was the manager of a consumer science products dealership, we produced a "point of sale" display for the company, to try to illustrate the efficiency of various forms of eyepiece nebular filters, by photographing a light polluted region of sky where we conducted the visual tests of filters that will be described below. The prints and negatives for this display were long ago lost, so on the evening of 7-6-97, we repeated the test and provide the results for your inspection.

Many astrophotographers have tested the efficacy of filters with guided exposures in order to show the improvements and reduction of skyfogging; but our own specific test with a fixed camera tripod setup was intended simply to show the reduction of light pollution.

We constructed a small metal hood for the 50mm f/1.7 lens of our camera, that fitted closely over the lens, cutting off all light except through a 27-mm aperture that was fitted with threads for 1.25" eyepiece filters. This changed the focal ratio of the system to about f/4, still fast enough to attain skyfog limit in a few minutes with color-rich Kodak 400 print film.

However, a vignetting of the negative was caused by this adapter, so we have cropped the scans of the photos to contain the exposed region of each frame, and just a bit of the dark region around it.

The exposure time was standardized at 15 minutes in each case. The site was one of the viewing venues we employ for astronomical observing, at over 3,000 feet of elevation above sea level in the mountains north of Santa Cruz, California. Lights at the horizon are from the populous areas of Gilroy and Morgan Hill, and in the unfiltered picture are extraordinarily bright and intrusive in the long exposure! A number of other "unidentified flying objects" may be visible: more on these below.

No Filter in Use:

In this exposure, the brilliance of low- and high- pressure sodium vapor streetlights is terribly evident. In addition, at the right of the horizon, a trail is visible from the rising planet Jupiter, as are some fainter star-trails.

Though we did not take special note of them by eye while the picture was being exposed, the camera unerringly captured numerous "UFOs": trails of airplanes taking off from a south Santa Clara county airport! All in all, it is hard to imagine that any decent deep- sky observing can be done at this venue, but we manage to accomplish a lot when fog comes in to blanket the worst of the light pollution, by observing away from the strongest source of light, or by using filters on extended nebulae.

LPR Broadband Filter in Use:

Using a standard "light pollution rejection" filter [similar to a Lumicon DEEP-SKY"[tm] or Orion "SkyGlow"[tm]), which primarily suppresses the yellowish wavelengths of sodium- vapor streetlights (plus some high atmospheric airglow), this 15 minute exposure captures a lot of reddish radiation in the h-alpha range from the city lights, but loses most of the awful yellow pollution wavelengths.

Sadly, however, the effect by eye is not NEARLY so dramatic: in fact, it is rather subtle, provided just the right exit pupil is chosen for the best effect. Do not expect to view nebulae with extraordinary contrast enhancement with an LPR filter: but some improvement will be evident by eye. The increased photographic exposure time is the best service provided by this type of filter, though for visual use, it can provide a slight contrast enhancement on some globular clusters, galaxies, and on bluish reflection nebulae (such as the faint glow around some of the bright stellar members of the "Pleiades" (M-45), or part of the "Trifid" nebula (M-20). The author finds also that the LPR type filter works well in small aperture telescopes, at narrow exit pupils that are too small for UHC or OIII types and consequently cause dimming of all light. Often under such conditions an LPR type filter may produce effective contrast enhancement with exit pupils as small as 0.5 inch. The other filters, used at such high magnifications -- or in small aperture scopes, with very small exit pupils -- might create a "dead black field", nothing being seen; whereas the LPR filter might help, without causing the desired object to disappear.

UHC Filter in Use:

With the addition of a standard visual narrowband nebular filter, primarily centered on hydrogen wavelengths (similar to a Lumicon "UHC" [tm] or an Orion "UltraBlock"[tm] model), we see a less- strong transmission of reddish h-alpha wavelengths, combined with the greenish-blue light to which the dark adapted eye is most sensitive, resulting in a slightly less reddish cast than the LPR filter picture above; note that the star trails have nearly disappeared, along with much of the nasty skyglow.

This type of filter is not intended for photography, but is designed to help obtain high contrast on emission nebulae such as the "Lagoon" nebula (M-8), or the "Omega" nebula (M-17), both in the constellation of Sagittarius.

O-III Filter in Use:

The O-III filter created an effect in the photo that was similar to the UHC filter except for a slightly fainter registration of a somewhat more purplish color.

H-Beta Filter in Use:

The Hydrogen-Beta filter [in this case, a Lumicon "H-BETA"[tm] model] cuts off even more light, transmitting a very narrow region right at 486.1 nanometers, the precise hydrogen- Beta wavelength. As you can see in this photographic test exposure, very little light is recorded from the streetlights at the horizon, though the trail of Jupiter is just visible at the top right of the image. This filter -- though NOT for photography -- is very efficient at eliminating almost all light pollution and skyglow, but unfortunately works well visually for only a few kinds of emission nebulae: thus it should be a choice for the dedicated amateur who has progressed beyond the other types of filters, and who believes that the study of such objects as the "Horsehead" and "California" nebulae justify the cost.

However, it does work as advertised, and permits one to see the faint nebulae, provided that the observing conditions are propitious, and the telescope / eyepiece can provide the exit pupil necessary (shown in our software program EYEPIECE and in the data sheets for the filter.)

Again, these tests above are not indicative of the telescopic visual performance of the eyepiece filters, but instead demonstrate their general cutoff effects in reducing light pollution. The eye's sensitivity to faint light, and the film's response to visual wavelengths, are strikingly different.

VISUAL TESTS OF FILTERS:

The primary authors of "Eyepiece" -- Waldee and Wood -- have many years' experience with both visual and photographic use of the appropriate types of these filters, and have even devised and conducted an extensive series of double- blind tests of visual filters, with a group of four observers with varying degrees of viewing experience (age 14 to the late 40s), to evaluate their performance in moderately light- polluted skies.

Our methodology was to use a selection of filters with several identical brand and focal length oculars, which were inserted by a test administrator technician quickly into the focuser of our test telescopes -- while the observer waited with eyes closed, not realizing at any time which filter was being used -- so that only 2 to 4 seconds elapsed between filters being tried by the observer. The recorder of the observer's opinions thus did not know either which eyepiece was in use, which was determined by the test technician. This is not as precise as the method employed in a Feb. 1991 ASTRONOMY Magazine article in which a virtually instant change was made using a specialized mechanical switching contraption (similar to the Multiple Filter Selector made by Lumicon), but it is possibly closer to the normal process used in typical observing sessions, while eliminating preconceptions or bias, or the observer's expectations based on knowledge of which filter was about to be employed.

In addition, we had first consulted actual laboratory measurements of the transmission characteristics of each filter, and used a visual spectroscope display, employing a diffraction grating and matching optics, to study the light wavelength responses, to correlate the lab measurements to the telescope tests "under the stars."

Our tests indicated in general that: OPINIONS ABOUT VIEWING ENHANCEMENT WIDELY VARY.

Differences between brands were noted, but preferences were not always consistent. There is not necessarily a BEST filter for any specific object under all viewing conditions, or with all telescopes. Despite what you may read in certain posts or alleged 'studies' done by amateur astronomers, there is arguably no satisfactory test report that reliably indicates what filter to use on specific objects -- for a variety of telescopes, conditions, and observers -- that has a full description of protocol and complete testable data that may be examined, tried, and confirmed (or discounted) by anyone. In the present writer's opinion, all work in this area published so far for amateur astronomer use is merely anecdotal, accomplished without controls and proper blind or double-blind test methods; some reports do not even rely on more than one test observer. Thus, the individual personal opinions of diverse users provide the recommendations you are likely to encounter. Some of the observers are indeed very practically skilled, reliable, and careful; others may fall short of this standard of competence.

Furthermore, during our test of several different brands of similar bandpass filters, we discovered that there was a visual perception of significant differences due both to production line model variations in one given brand/type, and also to the shape of bandpass or bandstop response, differing in general among different brands of equivalent type/use filters.

The best filters are sold with manufacturer-derived printed response graphs, plotted for each filter sold. If you are able to visit an astronomical dealer who carries several competing brands, you may examine those graphs and decide which one looks most effective, based on the shape of the response curves. However, in our test, some observers liked the more natural effects of certain makes of narrowband nebular line filters (commonly referred to as UHC type) with gentle cutoff slopes; others appreciated the high contrast of the filters with the sharpest curves. So, merely checking comparative measured response from the printed plot might not be an indicator of your preference by eye, through the telescope.

Filters we compared, which were made in the late 1980s, may not be equal in design to production models sold in 2006, so the actual response curves measured by instrument, and perceived by observers, are not necessarily precisely relevant to today's products at telescope dealers. Even though the same models and brand names of filters are still being made, these products are sometimes "job-shopped" to a variety of different oriental filter makers, over time. The marketers of filters sold in America under their house-brand name may still be using a certain trademarked title; but there is no absolute guarantee that ones purchased over a twenty-year period (with astronomy products companies changing management, and even corporate ownership) that a "2007 oxygen filter" and a "1989 oxygen filter" are exactly equal in all characteristics. Therefore, it is not relevant to publicize precisely what we learned about competing products, nearly twenty years ago, as that condition may not have remained stable. We can say, however, that the more we tested, the more we observed interesting variations, not only in the products but also in the perceptions of the testers.

I am now entirely retired and run a home music teaching business, no longer being involved with telescope products dealers; so I have no particular "axe to grind" with respect to any particular maker, and do not benefit in any financial way from my amateur astronomy hobby. I came to the conclusion, partly from the earlier tests and also from later ones done after I had left the telescope products business, that the very particular, dedicated observer may want to try competing brands in the same observing session, either buying them (with the seller's agreement that the undesired filter may be returned for full credit or refund) or by obtaining the cooperation of a fellow observer with a different complement of filters. While, say, a "sharp" sloped hydrogen-line nebular filter might reduce light pollution, it could diminish faint details visible in the light of other wavelengths radiated by the deep sky object. So there is no guarantee that "Brand X" with steep slope will LOOK better on a given object, than "Brand Y" with a more gentle slope. Testing visually is always the best way to determine an individual user's aesthetic preference.

For instance: during our tests, we had earlier measured the bandpass characteristics and slopes of competing brands of equivalent hydrogen-line (ultra high contrast type) filters using a spectroscope, and knew which ones were "sharper" and which were "gentler". But, remarkably, there was no consistent preference, in the opinions of our testers, for one or the other. Our own surprising conclusion, on such multi-faceted objects as M-42, was that there was a strong justification -- for some demanding souls with a patience for making incessant comparisons -- for owning two different brands of ultra-high contrast hydrogen-line filter! The nebula looked significantly different with supposedly "identical use" filters, crafted by different makers.

Filter transmission percentage variations were not always detectable. Small differences in the transmission peak of a given brand of filter were not necessarily perceived under the conditions employed in which a 2 - 4 second lapse occurred between switching ocular / filter, using identical brand and type filters. In other words, viewers could not always reliably tell which filter had a 10% loss, or a 15% loss, at the precise wavelength of a nebula. However, we concluded from certain trends that the alternate "instant switching" method would possibly have shown such differences -- to some of our observers.

Yet, the eye - brain quickly compensates for slight visual differences of perception, and "readjusts" to the view presented. As in every product, 'you get what you pay for': but the most careful and experienced viewer is likely to be able to spot the differences in filters.

Exit Pupils

Exit pupil size was quite critical to viewing success. If the exit pupil was too big, the result was a poor enhancement of contrast and little improvement. Conversely, if the exit pupil was too small, the background was very dark, and the filter made little improvement, since the image was already too dim. While calculation "by the numbers" using our program or published filter recommendations is not always exactly predictable, it certainly provides a means of understanding the acceptable range of magnifications in a given telescope, over which a filter will work well.

Dr. Jack Marling's original research from the 1970s and 1980s about exit pupil size was tested, and confirmed, in our tests over the years. With his kind cooperation, we used his range of recommended sizes in our computer program Eyepiece. Below is a table of the values of exit pupil worked out by Dr. Marling for two different conditions of sky darkness: "suburban sky" (with light pollution) and "dark sky" (approaching a naked eye stellar limiting magnitude of about 5.5 or better.)

You may also find a chart of those exit pupil values, plus other data, on the Lumicon website as a PDF file. Of course there is no harm in deviating from those values if your preferences differ: experimentation is always useful.

In conclusion, the tests we did indicated that though certain controls had been established, and that double blind methodology was used, testing visual filter perceptual response variations was a very demanding and complex task. We derived results that were satisfactory and informative, though limited in scope. And an over-riding discovery was the lack of majority consensus during many individual tests. That's why I am encouraging the user to take his or her own voyage of experimentation and discovery!

Some corroboration of the results of our test of competing brands of filters, in which we had found that no consistent consensus was achieved, is found in the review of the Andover Corporation brand filters, by Ed Ting, on his comprehensive and fascinating "Telescope Review Web Site". On this page of his accessory reviews, Ed says (emphasis mine):

At a recent club skywatch, I handed both 19 mm Panoptics (with the OIII filters installed) to a number of experienced observers. Since it was dark out, they had no way of knowing which filter they were looking through. I trained the scope on the Ring or Dumbbell and told them to switch eyepieces as often as they liked, and to tell me which one (if any) they preferred. This took about two hours. The results? Two preferred the Andover, two preferred the Lumicon, and one was neutral. All the observers said the differences were slight and that they would be happy with either unit. I guess it's a tie.

I can assert that our test team also had many split decisions; but the preferences of the observers were often quite strongly expressed. If money is not scarce, there may be a justification to own different brands of identical types: there is often a distinct difference noticed in filter slope, but as often as not, it helps some objects, while not being as effective on others (and the results vary slightly with exit pupil and telescope aperture size.) I can say, however, that as an audiophile and musician myself, the differences between brands of nebula filter were not nearly as noticeable to the eye as -- say -- the sonic characteristics of competing phono pickup cartridges or loudspeakers (or pianos!) were discernible by ear.

FILTER SIMULATIONS

For our early-1990s version of the Waldee-Wood "Eyepiece" program, we created a series of simulated views of what various nebulae look like with and without nebular filters. Recently we have recast those images in a more realistic form in webpages. Some of the simulations of nebulae are better than others, due to limitations of the processes available to us fifteen or more years ago, but with some of our nebulae comparisons, such as M-20, M-27, M-42, M-97 (above), and the Horsehead, we feel that the simulations are reasonably effective. Click here for the main menu for the webpages featuring our Eyepiece/Filter Simulated Views.

FILTER ILLUSIONS?:

Recently in late 2006 and early 2007, the author read numerous comments on "high end" observers' forums that were very skeptical of observations of nebulosity made by viewers using filters. During the years that the author was involved in the testing and marketing of nebular filters, in the late 1980s, no one we knew had any worries or could offer any caveats. Filters were immediately perceived as being a godsend: wonderful accessories that could reveal hitherto hidden details that were obscured to the human eye by 'competing' light radiation that swamped fainter nebulosity.

But years later, visual astronomy has become more nuanced. Observers are trying for fainter and fainter objects; some are even difficult to photograph! In particular, planetary nebula seekers are sometimes concerned about large angular diameter, faint objects with incredibly tenuous and dim outer gas shells: and if the filters used to observe them are CREATING the illusion of nebulosity. The subject seems to have come to a boil after a very few "nonexistent" planetaries in the Abell catalogue, measured and plotted years ago from the Palomar Sky Survey plates but mistakenly identifying emulsion defects as nebulae, were "sighted" by one or two advanced observers, using filters. This has given rise to much commentary on the effects of "illusion", relating to faint wide nebulae, particularly in the presence of significant stellar glow.

In the author's opinion, this "problem" is not very significant. We discuss illusion effects created by using low power on star fields in another essay, at this link. Such illusions have been known by many former generations of observers, wary of light scatter in their telescopes and oculars. But, to test the situation in his OWN equipment, the author has begun a series of experiments. The first we've achieved was done on 8 January 2007, using a 10" aperture Orion SkyQuest Dobsonian telescope and some Orion and Celestron eyepieces of varying price ranges. The observing site used was the author's regular viewing locale, at 3,400 feet above sea level in the mountains south of San Jose, not far from the Pacific ocean and benefiting from very stable, laminar airflow and generally boasting excellent seeing. The telescope's tendency to create light scatter was evaluated on a night with very clear sky and atmospheric stillness, using stars in the range of 2 to 11 magnitude (in fact, our logbook recorded that on this night conditions had "Superb seeing: 9 out of 10!")

It was found that some scatter was evident with any eyepiece, around stars brighter than 5th magnitude; it diminished below that, and even before 10th magnitude was invisible. No significant scatter was seen against the sky background at high magnification, in the vicinity of isolated 10th to 11th magnitude stars that were elevated above the heavy air near the local horizon.

Then, all of the author's filters were tried out, at exit pupils related to the proper ranges defined by Marling. When there WAS scatter visible, adding a filter tended to "concentrate" it into a smaller area, while not making it any brighter. The filter just caused the light scatter to fall off quicker, the further away one looked from the center of the star image. No "extra" nebulosity was observed, and what scatter was visible was not significantly "enhanced" or increased.

All real problems observed, however, seemed to be related to one particular ocular: the next- to- lowest power eyepiece used in the test, an Orion "Sirius" Plössl of 32 mm focal length (a very inexpensive model), but only with the brightest stars. As our sketch below illustrates, with this eyepiece -- which yields an exit pupil of 6.8mm (slightly larger than necessary for this aging writer's dark-adapted eye) and a magnification of 37.5x -- a distinct artifact was observed when filters were added. It was, in fact, worst with the OIII filter (Orion brand): visible as a bright ring, colored green, around the vivid star. An Orion model UltraBlock filter caused only a very slightly perceptible sheen; both a Lumicon model H-BETA, and an Orion brand SkyGlow LPR filter seemed to add only a small "red squiggle" effect right around the disk of the star used for the test: bright Sirius, which is a blinding -1.6 magnitude, the brightest in the night sky from the author's locale. It was elevated nearly as high as possible, less than an hour before transit.

Tests done with fainter stars from 5th magnitude and lower, even with this ocular, showed no perceptible "nebular illusions". And, with an even lower- power 40 mm Celestron E-Lux, that irritating bright green ring that was visible with the Orion 32 mm ocular was not present. Please note also that in the sketch below, the secondary vane reflection artifacts -- the crossed lines at 90 degrees -- were drawn; these had nothing to do with the filters and were visible without them.

This test tended to put our mind to rest about this particular telescope. Other observers (such as Jaakko Saloranta) warn that filter illusions may be more noticeable with small aperture optics, particularly at low power. Saloranta has found, for instance, that the OIII filter, used at very low magnification in an 80 mm scope, can -- under some conditions -- create 'a field full of fuzzy planetary nebulae', a totally spurious effect that makes it very difficult to isolate the real nebulous object amongst the stars.

In fact, the author may have noticed an unusual situation a few times, using the Orion "SkyGlow" LPR filter while looking for an extremely faint, low surface brightness Abell planetary nebula. Such objects are barely brighter than the background seen between stars, caused perhaps by airglow, gas and dust in the plane of the galaxy, and light pollution scatter. But telescope optics also cause light scatter, as can local atmospheric moisture. In certain instruments -- such as the author's own 10" f/4.7 Newtonian telescope -- the brighter stars are almost always accompanied by very slight haloes of light. The dimmer the star, the dimmer the halo: until the light scatter is below human detection. If you add a filter, the haloes are 'diminished' in diameter, fading out over a narrower radius, to a very much darker sky background. And, all starlight is dimmed slightly by any filter, which will have an overall loss even at the passband frequencies. It could be that under certain circumstances -- exit pupil, brightness and number of the stars in the eyepiece field, and reduction of maximum light-to-dark amplitude by the filter losses -- that the viewer now notices MORE of the small 'scatter haloes' around each star, even the dimmer ones. In the author's case, this made it difficult to pinpoint precisely which star was at the periphery of the planetary nebula Abell 19 which, itself, was almost invisible with direct vision during a scan of all the stars in the eyepiece field. Most of the stars had a 'halo' -- which halo was the planetary nebula? The quandary was resolved by taking out the filter and using very high power.

In the future we shall repeat these experiments with our larger Schmidt-Cassegrain, and our smaller refractors. One may always eliminate the cause of alleged filter illusions by choosing other oculars and filters, and looking closely without a filter to see if any vague perception of nebulosity is still visible (in many cases, it should be.)

One dealer of nebular filters -- Celestron -- has in fact suggested in their advertising that their manufacturing process eliminates such illusional effects: in the online web page for their O-III filter, Celestron's copy states: "Each filter has an ultra hard, vacuum-deposited coating carefully designed to block all of the visual spectrum ranging from 400 to 700 nm This eliminates the un-natural colored halos surrounding bright stars common with O III filters of less sophisticated coating technology." I haven't had an opportunity of trying that filter and, indeed, have not used any Celestron nebular filters for more than a dozen years, so I can't verify this claim; but it's a thought-provoking assertion.

EMISSION NEBULA TEST with VARIOUS FILTERS

One very discerning and skilled amateur astronomer -- David Knisely -- has put together several versions, over the years, of a test of nebulae using the four primary types of filters discussed in this article. It was not a blind or double-blind test, and he states that there was a certain amount of subjectivity in his judgments. Most of the versions on the Net did not specify the telescope used, but a newer revision of his Filter Performance Comparisons, from September 2007, finally includes the specific instruments and magnifications for each filter, making it much more useful than before, in my opinion; so now I feel comfortably enthusiastic about linking to it in this article as a useful continuation of the topic.

The results of our own double-blind test procedure -- and many conflicting experiences that I have had with specific objects and filters, plus somewhat differing opinions of numerous trustworthy observers -- have convinced me that individual tests by one observer may be arguably anecdotal, and perhaps even unrepeatable; and that a larger, truly scientific study, utilizing controls and proper statistical analysis, would be valuable. Until that materializes (if ever), there very well may remain difficulties in reliably generalizing from one individual's experience under one unique set of circumstances in order to predict valid group expectations. The lesson our group test taught us is that each observer should (a) learn how to use filters correctly; and (b) make his or her own observations and conclude what works best, without worrying too much about the advice of other observers (since there are indeed so many unpredictable differences between the skill levels, knowledge, taste, and gear of individual amateur astronomers.)

However, if you have read the very interesting and informative Knisely article cited above, perhaps you might be interested in an alternative personal test I recently conducted with an 80 mm telescope and the nebula NGC-281. In fact, I came to some very different conclusions in comparison to Mr. Knisely's opinions, using a different telescope: evidence of how varying the circumstances and observer can change the perceptions. But I'm certain that his opinions are equally valid and practical, and make sense in the context he describes.

Please note that when I did the observations of nebulae reported here, I did not know what telescope Mr. Knisely had used, because I had used as reference the widely-read 2006 version of his filter tests, which does not give the instruments and magnifications for each object. Thus, my own trials were not conceived to be a scientific falsification test of his, or anybody else's, observations and conclusions. Perhaps this could be attempted, roughly; but it would require even more information about his test than I have at present -- and then would not control for the differences of individual human perception and judgment.

NGC-281 in Cassiopeia Tested Using 80 mm f/5 Scope

Here is an example of rather large deviation of opinion between myself and Mr. Knisely, with respect to the nebula+cluster NGC-281 (links to Deep Sky Browser pages for object: nebula; open cluster.) I shall preface this examination with a report on the known characteristics of NGC-281 so that we may understand, from the work of professional astronomers as well as an extremely skilled amateur researcher (Steve Gottlieb), how the nebula might best be isolated from the stars in the region.

This NGC designation signifies both an emission nebula and an open cluster: the former is nebula NGC-281, and probably also IC-11: 35x30' diameter, no reliable vis. mag. rating, but as Lynds nebula LBN 123.17-06.28, or LBN 616 (see this page), is rated as having 'Color 3, Brightness 1', meaning that it is brighter on the red Palomar plate and given highest brightness rating of 1 [in the range from 1 to 6]. The latter characteristic is as an open cluster [alternately, IC-1590 or Collinder 8: 10' diameter, 7.4 vis. mag.]. The object was discovered by Edward Barnard in 1881, using a 6-inch refractor during the early part of his amateur career as a comet-hunter (no filters were used, obviously!) He later contributed a second "discovery" of the same object, listed as IC-11 in the 1895 Index Catalogue; detective work, by Corwin and others, has revealed that its report had a position error and is almost certainly Barnard's earlier nebula, NGC-281, from the description.

The skilled amateur visual observer and deep-sky historian Steve Gottlieb has written very elaborate descriptions of the object, which may be found in this text file, or brought up via entering the object's designation in the NGC/IC Project Public Database page. Gottlieb has stated that, among other details, using his 17.5 inch telescope he has had a "spectacular view of this detailed HII region at 100x using an OIII filter. This 15' nebulous complex has a mushroom appearance and is separated into three main lobes apparently by dust... There appears to be a much fainter detached piece off the south end... The section to the north is faintest and separated from the eastern lobe by a curving dark lane. A dark intrusion is visible south of the triple star which appears to be due to obscuring dust."

The late Donald Osterbrock (one of my astronomical mentors) and Ralph Stockhausen published a study in the Astrophysical Journal in 1961 that did carefully calibrated photometric measurements of the Hβ and oxygen emission lines, compared to other bright reference nebulae (specifically, planetaries), and calculated differences in radio frequency fluxes. Of significance to an amateur visual observer is that the relative measured NGC-281 logarithmic flux values of Hβ (486.1 nm) and OIII (N1) (500.7 nm) are only very slightly different; these are compared to Messier 27/NGC-6853 ("the brightest known planetary in Hβ") and IC-481 (a very prominently visible small diameter planetary), as well as NGC 7027 and 7662: the "Dumbbell" nebula shows less H-beta than OIII, while the "Spirograph" nebula shows almost equal flux values for these two nebular lines. Thus, ideally the visual nebular filters such as the UHC and OIII type should work reliably on NGC-281, given a sufficient exit pupil.

With this in mind, I viewed NGC-281 during a recent observing session (on Monday 2 June 2008) at the northern California site Coyote Lake County Park, much favored by south bay area observers, as the eastern sky is quite dark, avoiding the San Jose metroplex lights. The elevation at the boat ramp parking lot (used by astronomers) is 970 feet, and I set up my telescopes about 100 yards from the edge of the lake. The NELM at the zenith, around 3 am, was better than 6.1 (as far as I could tell; my eyeglasses are not especially good and my prescription is at least 4 years old, and I cannot see by naked eye the faintest stars that can be discerned with slightly higher magnification using diopters. I do not use glasses to view through telescopes.) The NELM in the region of NGC-281 was not significantly worse, and according to calculations using TheSky VI, the nebula was elevated from about 29° to 38° above the horizon (the transit time being 09:15); this was not the ideal culmination, but I chose the nebula because it is hard to appreciate from my regular high-altitude observing site on the other side of the Santa Clara valley, toward the southwest (where instead of looking into a dark sky, I have to see NGC-281 through the light domes of Morgan Hill and south San Jose.)

I recorded that the seeing was "very good" with excellent transparency, almost no visible twinkling of stars at 150x; and to the naked eye, the 'glow' and 'grain' of the Milky Way were beautifully seen, coursing up from the NE horizon through Cassiopeia, into Cygnus, and all the way to Scorpius: I consider this set of conditions to be very satisfactory (for a locale not much more than twenty miles south of a large city.) The telescope used was my Orion ST-80 f/5 80 millimeter aperture achromat, which has been modified with the replacement of its focuser with a larger one accomodating 2" barrel oculars. I was thus able to use my 1.25 and 2.0 inch barrel eyepieces, and my Orion filters: 1.25" SkyGlow ("LPR" type), UltraBlock ("UHC" type), and OIII; Lumicon H-Beta 1.25"; and 2" UltraBlock and OIII, at varying magnifications/exit pupils.

Finally, I have studied the history, early photographs, astrophysics, and observing of this object for many years, having logged it numerous times since at least the early 1980s (at Lick Observatory with a 4-inch; with my 80 mm scope in Nevada; at Lake San Antonio; in the Ventana Wilderness; and at my regular mountain observing site north of Santa Cruz, in scopes from 5.5 to 17.5 inch aperture.) My friend the late Dr. Donald Osterbrock suggested in 1989 that I study the 1922 paper by Edwin Hubble that includes NGC-281, and I received his advice about understanding Hubble's research when I obtained it in fall of 1989 from the Lick archives. I also was privileged to find the original published reports by Barnard of the nebula's discovery, in Lick's fragile old copies of The Sidereal Messenger, and the earliest widely published photographs that I could locate (which were, if memory serves, printed in the British magazine Knowledge around the turn of the 20th century.) I am sure that Mr. Knisely is also a very experienced observer of this nebula; his excellent observing reports of a diverse and large variety of celestial objects have been frequently posted for many years on many Internet forums, discussion groups, and websites.

Now, with this preliminary to give you a perspective, here is a comparison of my results, with those reported by Mr. Knisely, using his same general format (bearing in mind that he used exclusively the Lumicon brand equivalent filters, and offered only one numerical "grade" per filter, while the filters I used were only one Lumicon model, and three Orion brand types (sometimes both 1.25" and 2.0" versions), and gave a range of numerical grades, depending on the magnification/exit pupils employed):

I consider my opinions to deviate significantly from Mr. Knisely's observations, and speculate that the different telescope aperture size may be a major factor. We likely also have individual tastes and perceptions. I can assert, from years of experience with this object, that it looks different with varying telescope apertures, and in varying fields of view and a variety of exit pupil diameters. A big scope allows you to get a larger image scale at the same exit pupil of a smaller telescope. Wider fields permit the nebula to stand out, compared to narrow fields; but if the magnification is very low, the nebula might not be seen at all unless you use a narrow-band filter. My perception of Mr. Knisely's description of how he evaluated the results may indicate that I had a slightly different goal, in part: specifically, I made note here of NGC-281's odd shape, darker and lighter regions, and the effect of the field stars and the cluster; I believe Mr. Knisely focused, in testing an enormous number of objects, on trying to judge principally the "overall surface brightness, area of nebulosity observed, and contrast of detail."

My test, above, was made without any immediate prior reference to Mr. Knisely's results, which I hadn't recalled. But, after preparing the above information and organizing it in a manner similar to his presentation, I did the observations again at the exact same location, very soon afterwards on two days, at about the same local time. Bearing in mind Mr. Knisely's recommendations, I tested the nebula once more, focusing on critical comparisons of the H-BETA, OIII and UHC types. Below are my findings, when I remembered clearly and knew what he had suggested. This time, I used only 1.25" barrel eyepieces, and include the magnification steps and filter changes that made significant differences; and I am now recording the results not by filter only, but also by the steps of magnification employed:

SRW Test 2:

80 mm f/5 & 1.25" Eyepieces: -- 32 mm: 12.5x, 6.3 mm exit pupil; 25 mm: 16x, 5 mm exit pupil; 15 mm: 27x, 3 mm exit pupil; 11 mm: 36x, 2.2 mm exit pupil; 7.5 mm: 53x, 1.5 mm exit pupil.

12.5x with H-BETA: Image scale is so small that it's not very effective, and the nebula is virtually invisible without a filter; but with H-BETA it is very distinctly seen, a small puff of haze around the central point. The use of the filter at this power also seems to cause a very slight, generally uneven, lumpy, or "curdled" variable background brightness difference in various places in the general area, well outside NGC-281's diameter, particularly to the N (perceptible though noticeably dimmer than NGC-281 itself.) Some of this background character might be "filter artefact effect"; or perhaps it is genuine, the filter assisting one in discerning what is known as 'high latitude galactic cirrus' that has been detected in various celestial regions.

12.5x + OIII: Nebula perhaps paler, but very slightly more distinctive than with H-BETA. Now, slight variations in brightness are visible, all around central point, with hint of the dark region.

Both, above: very important to attempt to get the bright stars Eta Cas (3.5 mag) and Alpha Cas (2.2) out of FOV, if possible!

16x + H-BETA: Filter helps a bit but while glow is just a vague shapeless patch, which tends to look equal across its extent, it was almost invisible without it: therefore, very useful for detecting the presence of the nebula (which, with the H-BETA filter, is much brighter than many of the Sharpless nebulae that I've observed.)

16x + OIII: Nebula shape is more distinct than with H-BETA, and curdling effect is gone. Nebula looks asymmetrical, so it improves detail somewhat over H-BETA, though the nebula itself is not quite as high in contrast against the background.

16x + UHC: Much better: nebula's character is more distinct, even with a less-dark background.

27x + LPR: Very good, with asymmetry registered now.

36x + LPR: Excellent -- best result, with distinctive shape, asymmetry, dark region, and variable density areas; with averted vision, like a picture!

53x, no filter: Neb'y shows up about as well as it did with H-BETA filter at 12.5x, though the FOV is much narrower. It might be difficult to find it initially at this magnification, with no filter.

NO. OF SESSIONS INVOLVED: two, from 3:20 am to 3:50 am, 5 June 2008, and 2:30 to 3:00 am, 6 June 2008 (same location as earlier SRW test, above.)

In conclusion:

• Unlike DK, I found that the H-BETA filter, at the lower powers, is effective in helping to bring out the glow, and to reveal exactly where the nebula is, even though it was invisible without the filter.

• Furthermore, DK says that the H-BETA gives "no more detail than is seen with the Deep-Sky filter" (LPR type), which I take to mean that neither was especially useful compared to the UHC and OIII, in his opinion. I disagree strongly, for at critical magnifications with small exit pupils, the LPR type bested the H-BETA, revealing the shape, asymmetry, and brightness variations; but at the largest exit pupil employed, the LPR type was ineffective -- while the H-BETA did work significantly better than the LPR filter, and could show exactly where the nebula was. The nebular glow disappeared almost completely when the H-BETA filter was removed at that low magnification. Thus, the H-BETA was quite helpful with large exit pupils, showing the outer edges of the nebula and making it stand out significantly; whereas the LPR type was barely able to reveal any trace of the nebula at the lowest powers, not even suggesting its true extent or shape.

• In addition, the UHC type gave me a better sense of the nebula's character than the OIII (the opposite of DK's assertion.)

• In general, the smaller the exit pupil, the less effective were the extremely narrowband filters, which worked according to Marling's recommendations; the LPR filter -- often much derided by many amateurs -- does seem to work well at small exit pupils, where it provides a subtle but very distinctive enhancement: at a critical magnification, it yielded an excellent result.

• Despite the fact that I had checked DK's test and suggestions, I had no difficulty in arriving at some very different personal conclusions and finding that they were satisfactory to me; but I indeed did agree with some of his preferences. In fact, I might agree with more of them had I used a larger telescope closer to the 10-inch aperture size of DK's optics.

• From now on, I intend to try to use all four types of filters in viewing NGC-281 with any scope, and to test the full practical range of magnifications, up to the point where (a) the nebula becomes too dim, (b) the field is too narrow, or (c) contrast is diminished. I was surprised to see all of the differences I perceived; pleased to learn that the H-BETA could help at lowest powers; and fascinated by the alternating perceptions of this or that detail. Despite having viewed NGC-281 many times with larger scopes, I feel that the test has given me the best appreciation I have had so far with an aperture as small as 80 mm.

• I do not consider these preferences of mine that deviate from DK's findings to be merely the "small differences" that author he once predicted in one of the online versions of his test survey; in fact, some of my observations are the reverse of what DK posits.

• SRW's RECOMMENDATIONS: Find the nebula at lowest powers with the H-BETA filter, if it's invisible; use the UHC filter to see details at somewhat higher magnification. The LPR type filter is useful at smaller exit pupils, and you may find a critical magnification where it shows a remarkable "photo-like" quality, with overall detail and character that are superior to other filters. Finally: don't rely on these suggestions: do your OWN tests!

EMISSION & REFLECTION NEBULA TEST with VARIOUS FILTERS

Mr. Knisely says in his Filter Performance Comparisons study that he tested only "some diffuse emission and planetary nebulae"; it is apparent that he did not evaluate the effects of filters on reflection nebulae. These last types of objects are defined here: they are "clouds of dust which are simply reflecting the light of a nearby star or stars...[which are] hot enough to cause ionization in the gas of the nebula like in emission nebulae but are bright enough to give sufficient scattering to make the dust visible... They are usually blue because the scattering is more efficient for blue light than red (this is the same scattering process that gives us blue skies and red sunsets)... Reflection nebulae and emission nebulae are often seen together and are sometimes both referred to as diffuse nebulae. An example of this is the Orion Nebula."

A significant number of emission nebulae sought by amateur telescope observers are not simply one-wavelength or monochromatic objects. Many have light- frequency components that fall into various regions of the visual spectrum, from the colors blue to red; but the dark-adapted eye is generally sensitive mostly to the blue-green region. However, the exact visual sensitivity has been shown to vary considerably from observer to observer; and the larger the scope aperture, or bigger the exit pupil, the more the eye will register increasing traces of light in wavelengths that can be detected.

An object like Messier 20, the "Trifid" nebula, has at least two, or possibly three, predominant light wavelengths that can be detected by visual observers, and at least three -- possibly four -- that may be registered on film or CCD sensors (plus, in both visual and imaging circumstances, a defined region of broadband light glow that tends to look bluish due to Rayleigh scattering: see the paper "The Trifid reflection nebula" by Beverly Lynds and Earl Oneil Jr. for more information on the blue nebula associated with the supergiant star HD 165614.) While the night eye can't reliably detect the ruddy hydrogen-alpha light emitted by the Trifid, it can see the greenish hydrogen-beta line in the "flower petals" of the central lobed nebula, around the double star; and the bluish reflection nebulosity to the NNE side is also detected as a faint, powdery glow: see these various broadband and multiband images to appreciate the enormous complexity of the object. You can, in fact, 'turn the reflection glow on and off' by inserting a narrowband filter, such as an OIII.

Mr. Knisely's study focuses, he says, on the comparision of emission nebulae, and in fact he does not discuss the visible effects of filters to optimize the reflection nebula associated with M-20. I would like to present his findings, and then my alternative opinion, considering how to improve the definition of the regional reflection- nebula's glow, and therefore to get a better view of both types of nebulosity.

DK:

M20 TRIFID NEBULA ("diffuse emission/reflection nebula in Sagittarius")

TELESCOPE: 10" f/5.6 Newtonian Reflector

DEEP-SKY LPR: "(2) Small difference between filtered and unfiltered views with a slight gain in contrast with the filter, but with any light pollution, the filter may be of greater use."

UHC: "(4) Nebula is slightly fainter than with DEEP-SKY filter, with a slight gain in contrast over the DEEP-SKY and more contrast gain over unfiltered views."

OIII: "(3) Nebula is fainter than with UHC or DEEP-SKY, and main trifid section appears slightly smaller (hurts the northern reflection nebulosity), but dark detail in the inner "lanes" shows up slightly better."

H-BETA: "(4) Nebula is somewhat fainter than in UHC, but trifid section shows a bit larger area of nebulosity than the UHC does. It kills the reflection nebula and reduces the brightness of the detail right around the central star."

RECOMMENDATION FOR M20: "UHC/H-BETA."

I'd say that he does describe quite well the effects of changing the UHC and OIII filters. However, M-20's reflection nebulosity (one of my own tests for the transparency of the southern sky) is hurt by the insertion of the OIII filter, and killed by the H-Beta one. That does not make the view of a significant aspect of the object 'better' at all; the reflection nebula is now 'worse'. But, the reflection component of M-20 is quite strongly enhanced by the standard broadband light-pollution rejection filter (such as the Orion "SkyGlow" or Lumicon "Deep Sky"); also, I've found, the dark lanes are enhanced too by this filter.

If your goal is to see "all" of M-20 at highest efficiency, you can benefit from having both the LPR and the UHC type filter, alternating them with an unfiltered view, in appreciating the totality of the Trifid nebula. In considering all the information that may be gathered about M-20, by changing from Deep Sky to UHC, or UHC to OIII, you are improving the emission components but degrading the reflection component; I contend that one of the narrowband filter types does not necessarily make the view of the whole region "better". With respect to emission nebulosity only, the author of that study gives a recommendation: to use a filter that unfortunately "kills" the wideband scattered reflection glow: the UHC/H-BETA duo. In my opinion, if you want to see the reflection component, that final recommendation is, therefore, weighted against trying to see a very significant aspect of the nebula.

My interest in the reflection nebulosity in the M-20 region has been long-standing: I was fascinated in the 1980s when I located at Lick Observatory one of the historic first pictures of Messier 20, taken there in 1899 with the Crossley telescope (unfortunately, badly reproduced here in the online version of James Keeler's paper), which then revealed the subtle high frequency energy that is often overshadowed by the prominent "trifidus" emission region (so described by John Herschel.) The blue-sensitive photographic emulsion of the time registered the reflection glow effectively, better than it could be seen by eye: acting, as it were, like a filter.

Thus when I wrote up my own account of the nebula a few years later, I focused on the means of enhancing both the emission and reflection nebulosity. Contrast DK's analysis with what I've written, as long ago as 1992, in this article: it was originally included in one of my early versions of the software program "Eyepiece" as my attempt to simulate some of the enhancement effects of filters. In 2007 I revised the section of the old program's help file and turned it into a webpage in my "eyepiece simulation" series. I reprint the picture panel below, in "compressed" size which will acquire some distortion artefacts; to see it better, go to the link cited above. In addition to Ron Wood's color image (at right) I tried to simulate the views one might obtain with an 8-10 inch aperture scope, with an LPR filter (left) and UHC type (middle).

During one of my last viewings of M-20 prior to the date of writing these comments, I was using merely a 4-inch telescope. I personally concluded that the LPR type filter (I was employing an Orion SkyGlow at moderate levels of magnification, between 100x and 150x) gave me a very great enhancement of the object: it slightly improved the emission nebula contrast (the Trifid's dark lanes -- catalogued as Barnard 85 -- were almost totally invisible without it), and significantly improved the reflection glow. Other filters hurt some and helped some; at very low power the OIII and UHC type were excellent (except for the reflection component), and of course they work very well at a larger range of magnifications in a bigger scope aperture. And don't forget that there is also a conspicuous star cluster associated with Messier 20, which won't be hurt very much by the LPR type filter but may be greatly dimmed by the UHC, OIII, and H-beta: use those filters, and you lose appreciation for the stellar aspect of this marvelous object. My own recommendation would have been, for this set of conditions, to use the SkyGlow (augmented by the UHC and OIII under slightly different conditions): now, it is up to the reader to wonder WHO'S RIGHT? Are we both right? Is only one of us right? How do you go about gathering evidence with which to make a judgment? Is a "final judgment" in this case even necessary?

I'd argue that the answer to the last question is a resounding NO! There is no "final" or "correct" judgment by any one individual, speaking for all observers, asserting a universal rule for getting the "best" view of such a nebula. If that is so, it illustrates my opinion that filter tests by individual observers are anecdotal, affected by circumstance, and understood in context of one person's entire gestalt as an observer. In this instance, I'd agree that Mr. Knisely's suggestions for enhancing various details in the emission part of M-20 are reasonable and helpful; and I'd suggest that there are many other subtle characteristics of the nebula and immediate field that are enhanced by filtering techniques he did not advocate. You might not agree with either one of us!

In a thought experiment, I have conceived an almost perfect analogy: imagine two music critics with different newpapers, who studied at different conservatories and who had entirely independent journalistic backgrounds. They were both assigned to review the same production of an opera, and each went to one performance but on alternate nights (so, while the soloists were singing according to their rehearsed intentions, and following the same conductor each time, not every nuance was rendered exactly the same way.) It is in fact inconceivable that the two critics would turn out identical -- or possibly even rather similar -- reviews!

CONCLUSION:

Eyepiece filters for nebular contrast enhancement will provide a boon for observers who view in regions where there is any visual light pollution; some models will help identify specific objects (such as small planetary nebulae or dark nebulae) and will reduce upper atmospheric skyglow. Colored eyepiece filters will be of occasional help for viewing enhanced planetary detail, but may not be as universally useful -- as some advertisers would insist -- if optics and seeing are superb . All in all, eyepiece filters are indispensable tools in the observer's quest to detect more information than is available in a casual glance. What brands should you buy? We advocate testing filters during star- party or observing sessions until you are sure a specific filter will meet your needs. Reputable dealers and manufacturers will have a reasonable return policy.

Consumer Tip: I've never purchased eyepiece filters by mail order. I prefer to buy them in person, at a reputable telescope and astronomy products dealer, and here's why. I just purchased a 2 inch narrowband nebular filter at a local retailer. They had several in stock, and I was permitted to look at the printed response graph that was packed with each filter. The response peak of units in stock ranged from 90% (worst case) to 95% (best case.) I bought the filter with the highest response transmission peak! Some companies sell filters that have the best specifications as "premium filters" for a few dollars extra. I doubt that I could ever discern, say, a 2% difference in transmission; but a 5% difference is probably right at the threshold of detectibility. -- srw

Nebular filters for viewing and astrophotography: available in a large variety of sizes and models for eyepieces, guiders, and camera lenses from Lumicon (first to make them for amateur astronomy, in the 1980s), and for eyepieces and camera lenses from the Orion Telescopes Company (which introduced their LPR and nebular filters in the late 1980s, and OIII and Hydrogen-beta filters after 2005), and others, such as Thousand Oaks Optical (who introduced them at the end of 1993), Astronomik, Baader, Celestron, Meade and their many dealers.

INTERNET:

• Steve Gottlieb, deep-sky maven par excellence, who has observed more than 7,000 of the NGC and IC objects, has written a very interesting and useful post to the Observers forum mailing list called "Observing with the H-Beta filter," which includes a number of interesting objects that respond well to it, with improved contrast: well beyond the usual few items generally discussed.

•  Barbara Wilson has posted on the Net her interesting article about The Magic Horse Head Eyepiece Filter; and there are numerous reviews on the "Cloudy Nights" website.

• The newsgroup sci.astro.amateur is an interesting venue to dip into for information about the personal experiences of amateur observers: but with reservations! Often, specific recommendations -- or dismissals -- of the usefulness of a filter are "self-negating" as the rebuttals sway back and forth, leading the beginner to suffer from extreme disorientation and confusion! Which advice is "right"? Which is even somewhat practical? One learns to judge the posts' tone, and thoroughness, in order to evaluate the seriousness of intent of contributors. But, even so, experienced observers can have opposite viewpoints (as in, for instance, these exchanges about filters that began with a query about the OIII type but sidelined into disagreements about almost every type.) Nevertheless, the adventuresome reader with a high tolerance for debate might enjoy looking through this directed search for opinions about nebular filters.

• This webpage by Maurice Gavin has some 'pretty pictures' made by means of a spectroscope, showing the response characteristics of a number of broadband, UHC-type, and OIII filters, and the predominant spectral lines of sodium vapor light pollution. A different approach to analysis was taken in the presentation by Rob Brown, using an alternative spectroscopic system that, unfortunately, is not very well documented nor shown in a photo. His measurements seem to me to be comparatively useful, though uncalibrated and not at all like the maximum response transmissions of Orion and Lumicon filters that I own, and have measured: so his peak response figures may (as he admits) be off, and are likely to be slightly under-rated. The 'problem' with both approaches, however, is that the dark adapted human eye is not ideological: so no matter what the NUMBERS may be in a given (arbitrary) test situation, the results perceived in YOUR telescope can be entirely different, and unpredictable. So, my best advice is to take the measurements with a grain of salt and rely on comparative experiences you have at each particular observing session to inform you about which filter works optimally with any given object, scope, and sky condition.

• A long article has been written that presents one individual amateur observer's opinions about a series of alternated filtered views of many objects: David Knisely's Filter Performance Comparisons For Some Common Nebulae. I admire the 9/12/2007 version of his report, but there are several versions of it on the net, including ones that do not specify the telescope(s) used for any of the individual nebulae; and the tests -- which I have no doubt were carefully done and are sincerely expressed and accurately presented -- are only by one person. They will, however, help to introduce unusual and very interesting nebulae to many readers; will increase the understanding of observers about diverse objects that are enhanced by the very narrowest-bandwidth filters; and the suggestions are expressed with lucidity in a very helpful format.

• Sue French, the "Deep Sky Wonders" columnist of Sky & Telescope, suggests that if you wish to determine the prominent nebular lines of planetary nebulae in order to plan effectively for using a specific filter, you may employ the "Gallery of Planetary Nebulae Spectra" of Karen Kwitter and Dick Henry of Williams College. Over 100 nebulae are catalogued; enter the professional reference ID number of a planetary (DO NOT use Messier numbers; employ NGC or other substitutes) and then obtain a page with astronomical data about that object, with a link to "Line ID Templates." Clicking that link brings up several PDF files with images of spectra. For instance: here is the page for NGC 7009, the "Saturn" planetary nebula in Aquarius (which I've discussed in one of my astro-blog posts.)

PUBLICATIONS:

"Visual Astronomy of the Deep-Sky" by Dr. Roger N. Clark, 1990, Sky Publishing; chapters 2 & 3;

"Night Sky Pollution: Measurement, Evaluation and Reduction by Filter" by Dr. Jack Marling of Lumicon Company; Livermore, California; Proceeding of Riverside Telescope Makers Conference, RTMC80, pp. 56-81; excerpted by permission of the author in our DOS freeware programs "Redscope" and "Eyepiece", and posted here for your convenience.

Reviews of nebular filters may be found in the February, 1991 issue, and recently in the article "Secret weapons - Nebula filters cut skyglow and let you see more with less", by Phil Harrington, in the August, 2005 issue of "ASTRONOMY" Magazine, and the July, 1995 issue of "SKY & TELESCOPE". We did not always replicate the findings of their reviewers, but the reports are quite thought- provoking, and are generally well-done.

The Planet Observer's Handbook by Fred Price, Cambridge University Press.

SOFTWARE:

"Eyepiece" (DOS freeware) Telescope parameter program for determining the exit pupil of eyepieces with respect to appropriate nebular filters, and prime -focus exposure times for a large variety of NGC, IC, Messier objects. The program is now so old, written for obsolete DOS, that it is available free.