The Quest for a Better View: Non-Standard Visual Filters for Planetary Observing with Small Aperture Telescopes
By Darren Hennig, M.Sc. – September 15, 2004.
Introduction
I have been an astronomy enthusiast since I was old enough to learn what that golden orb in the sky was at night. Observing the moon, looking at its beautiful, alien landscapes from here on Earth seemed surreal. This was all amplified by the lunar landings of the time. What a time of wonder for a small boy of 5! Since then, I have always been intrigued by solar system bodies; alien worlds so far away, yet close enough to be glimpsed at, almost touched, through the eyepiece of a telescope.
The fun and excitement I felt then when observing planets has continued to this day. I just cannot seem to get enough of it! Over the past 20 years, I have employed the use of just about every filter I could get my hands on, and this article is an evaluation of some less common, even unknown, filters, or those filters I have determined to be the most effective for use in small refractors.
One of the biggest challenges in using astronomical filters for observing planets with a scope of small aperture, is to achieve a balance in emphasizing the details of interest without robbing the much-coveted light collected by the instrument. Quite often, observers use “standard” colored filters when observing planets. These not only horribly skew the tint of a planet’s image; they make a high magnification image rather dim as well. We will first look at what filters do, and how they function to be useful to us in visual observing of solar system bodies. We’ll then look at which filters may be the most beneficial for certain planetary objects.
A Filter’s function
So, you may ask… what is a filter, exactly? Well, filters are used in a variety of situations, from the one used to make one’s morning coffee, a graphic equalizer or tone control in a stereo, to – yes, those used in optical systems. The general function of all filters is the selective blocking or quantitative reduction of unwanted or unnecessary matter or energy through the device. Matter filters include the coffee example. The essence of the coffee beans and the water solvent are carried past the filter, which is used to prevent the ground beans themselves from passing on to the pot or carafe. A tone control on a stereo blocks certain audio frequencies, and passes others through the amplifying electronic circuits in a stereo, thus allowing the listener to adjust the sound to their taste.
Optical filters perform a similar function to the stereo example. In the case of optical filters, basically the filter devices reduce or block selected wavelengths of light, passing the remainder on through the system. Proper selection of the appropriate filter device allows the observer to “tune” or adjust the image tonal or color quality to their taste or needs.
Regardless of the filter system used, we can see that they always operate via subtraction; that is, that they remove a portion of the total throughput in order to perform their function. This will become a very important consideration when we choose filters for small apertures later on.
The Visual spectrum and color balance
Most experts generally agree that the average human observer is capable of detecting wavelengths between 380nm and 780nm. This range includes the deep violet all the way into the deep red. Human observers may vary with the degree that they can detect these wavelength extremes, but generally, this is considered the visual portion of the electromagnetic spectrum.
Below is a general diagram of the visual electromagnetic spectrum:
The respective colors are shown corresponding to their wavelengths. Higher frequencies, corresponding to shorter wavelengths, go progressively to the blue and violet ends of the visual spectrum. Lower frequency, longer wavelengths will thus progress towards the red end of the spectrum.
So… how do we see color? Well, the human eye perceives colors using three receptors in the retina known as cones. Each cone type responds to different portions of the visual spectrum. In short, these cones respond to their respective visual spectral ranges. The diagram below approximates the relative responses of these cone receptors in the human eye:
The blue-responding receptors are designated with the Greek Beta () symbol; Red and Green receptors are shown with the Rho () and Gamma () labels, respectively. As can be seen here, the receptors respond differently to various wavelengths, with a small amount of overlap. The combination of information from all three receptors gives us the perception of visual color information, which is then transmitted by the optic nerve to be processed by the brain.
Since our eye responses are non-linear, it is difficult to express color perception in a way that is very intuitive. However, a chart was developed to yield a two-dimensional way of expressing color, which can be very useful in describing the action of filters to the perceived color. The CIE chart describes color in the way that our eyes perceive color, shown below:
The X and Y coordinates correspond to a rather complicated addition of the cone response functions above. Colors are shown with their corresponding wavelengths. The functions and their derivations will not be described here. Suffice it to say that with the above chart, color information can be expressed using these coordinates. Note the black line on the chart. This is referred to as the color temperature line. It shows the color of an object at a given temperature, using the black body radiation law. All sources tend to follow this emission behavior, especially astronomical ones. For example, a red, cool star with a surface temperature of 2,000 °K would look red-orange. A hot star at 20,000 °K would appear bluish white. Our sun is close to 5,800 °K. The solar photosphere color can be seen on the chart. We can find this as the 0.333, 0.333 [X, Y] coordinates, very close to the center of the white area. This area is what we perceive as “white light”. Our cones have evolved to perceive bright daylight as white - in other words, a combination of all colors of the visual spectrum, as radiated by the sun as an illumination source.
So… how is all this useful? Well, imagine a line, going from 470nm on the outer (blue) edge of the chart, through the (0.333,0.333) center, and on to about 565nm at the edge near the yellow. This line is like a lever, having a fulcrum, or balance point (also known as a locus) in the white. One can imagine virtually an infinite number of these, all going through the center, crossing to opposite edges of the color graph. These are lines of complimentary colors. If one were to move from the graph’s central white area, towards the green region of the spectrum, we have effectively added green, and removed (subtracted) some purple/violet from the overall color. I use this example to illustrate why a minus violet filter, for example, tends to have a green-yellow hue; violet and green are complimentary colors. Remove all or part of one, and you shift the color “balance” towards the complimentary color. By reducing or removing violet and some blue, we have effectively “forced” the image color balance towards the yellow-green. The more that is removed, the greater the overall depth of color towards the compliment. One can do this same thing with polygons as well, for considering three or four colors in balance, and so on.
Further examination of the above chart also shows something else: Primary and secondary colors. The primary colors we know from elementary school – red, green and blue. The secondary colors are derived from the addition of two out of three primaries: yellow, cyan (aqua or blue-green), and magenta (purple). Each secondary is a compliment to one primary. The Green and violet (magenta) couple, is one example; yellow and blue are another complimentary couple. Thus, yellow filters act by passing yellow and absorbing blue, and red filters absorb cyan, or some blue and green. The degree to which the filter removes certain wavelengths will determine the color of the image when the remaining light is passed on to the observer’s eye.
A point on the outer edge of this chart corresponds to a single color, or a monochromatic wavelength. So, an edge point at the hydrogen alpha wavelength of 656nm, for example, would correspond to that specific shade of red, only. This is why a hydrogen alpha filter image looks red. As one widens a filter’s response from only one wavelength, the color point shifts away from the edge. As it moves towards the center locus (0.333,0.333), incorporating additional color information, it passes more wavelengths. Eventually, a “filter” that passes the entire spectrum would sit right on the locus, or white light center point.
Optical Density and Filter Figures of Merit
I would be remiss in my general treatment of filters if I also did not cover optical density. This property is another way of expressing the relative throughput of a filter. Usually, optical density, sometimes referred to also as diffuse density, is most important for neutral tone filters, such as full aperture solar devices. Neutral density filters also often are rated using their optical density figure of merit. Some manufacturers use optical density (known from now on as OD) to express their filters in general.
Optical density is simply the logarithm (base 10) of the inverse of the optical transmission, or the negative logarithm of the transmittance:
OD = log10 (1/T) = -log10 (T);
Where T is the optical transmission, and OD is the optical density.
So, a filter having a 10% transmission, or 0.1T, has an OD of 1.0. Visual solar filters, often having 1/1000th of a %T, or 0.00001T total light throughput, have an OD of 5.0. More on optical density later on.
Many colored filters have a varying or complicated transmission behavior throughout the visual spectrum. It can be difficult to express, with a single number, exactly how much total light is actually being transmitted by the filter. The common rule of thumb used is an average percent transmittance (%T), at least for astronomical filters. This figure is obtained by taking a graph of transmittance versus wavelength, and taking the area under the transmittance curve. This value, divided by what would be ideal transmittance, i.e.: 100% throughout the visual spectrum, yields the average %T of the filter. This is the value reported for a colored filter when the %T rating is given. It does not describe in any way the maximum transmission of the filter in its passband, only the fraction of total light it passes compared to an idealized, flat-response optical filter.
Filter responses to visual wavelengths are often measured using a device known as a spectrophotometer. This device measures the transmission of light through a filter sample, relative to the baseline response of the instrument with a reference, usually air. The difference is the absorbance of the filter, and this value then is converted to the transmittance value. Below is an actual filter run performed on a real spectrophotometer:
This particular example is from an actual Celestron LPR filter, made by Baader Planetarium. Note that the filter response is moderately complex. The actual %T value from 400nm to 700nm is about 38%, yet the maximum transmission of the filter in its various pass bands is much higher than this, being ~95% near 500nm, and over 97% at 650nm.
Another figure of merit, usually used with narrow band filters for deep sky or line filters for astrophotography, is full width at half maximum, or FWHM. This is the width of the pass band curve in nanometers at one half the maximum value of the curve’s peak. With the 490nm peak seen in the above graph, the FWHM for this filter is about 45nm or so. The smaller this value, the narrower the pass band, and hence the more selective the filter is for passing specific wavelengths. This figure says nothing of the shape of the peak, however, and the peak can be even rather far from symmetric, as seen in the above example. The FWHM also does not say anything about what the maximum value of the peak is, either. Color filters do not use this parameter to describe their performance, but I include it here to complete the discussion of figures used to describe filters in general.
Useful Planetary filters for Small Apertures
Now that we understand a bit better how filters work to adjust the color of an image, and how they are often rated, we can now use this information to our benefit. We will now look at the various planetary targets and what filters can be useful to obtain significant detail without skewing the color balance excessively. The list included here are for filters I deem the most useful in general for obtaining the best image color balance and maximizing detail for various planetary bodies.
When observing planetary bodies, the idea in employing visual filters either on the diagonal or eyepiece directly is to enhance certain details. This is often accomplished by using complimentary colors to darken certain features of interest. For example, a red filter tends to darken the blue and green features of a planet, making them more easily seen. Quite common for Mars, a #25 Wratten (red) filter tends to be the favorite amongst many amateur astronomers for observing the maria and highland regions of this planet. I will counter this, saying that this filter certainly does work, but it also robs one of significant portions of the spectrum! A #25 passes wavelengths beyond about 610nm, so any color information available in the spectral regions below 610nm is effectively… invisible. The more tinted contrasting filters tend to waste color information as well, and in doing so also makes the planet significantly dim and reducing total detail. Very high power observations with small aperture scopes become more difficult than need be. I will get into why this is when I discuss which filters I prefer for Mars later on. One other function that these non-standard filters perform is color augmentation. For years, photographers have employed many of these filters to “bring out” or accentuate, certain colors of interest in their subjects. When using these filters, subtle tones that are not as prominent in the unfiltered image seem more vibrant, and thus the observer’s attention may be drawn to them more easily. This is the most important aspect of using these types of filters for visual astronomy. I cannot emphasize this important point enough.
Now let’s look at these non-standard astronomical filters I have found to maximize my success at discerning planetary detail when using them. The list below has some noticeable overlaps with some common filters that are used in visual astronomy. What makes those filters non-standard is the size format. Generally, budget astronomical filters typically used by many amateurs have very poor quality glass in them. There are exceptions, such as Baader colored filters, and the Vernonscope line. Larger format filters, often procured from sources such as camera stores, and specialty photographic equipment outlets, use much better glasses that are more consistent in both quality and surface finish. Many of these are also available in a multicoated variety, and this further enhances their use in visual astronomy, because ghosting and scatter is minimized.
Non-Standard Planetary Filters - overview
46mm and non-standard formats; require adapters
Cooling filters, 82A and 82B
Warming filters, 81A and 81B
Warming filters, 85 series
Fluorescent Color compensating filters, FL-D*
Skylight filters, 1A and 2A
Polarizing and Neutral density (ND) filters
Unusual format filter sizes
The formats I generally use for visual astronomy are either 48mm or 46mm. The former format is very common in astronomical circles, but is becoming rare in photographic size formats. However, the 46mm format is much more readily available, being a common format on many of the newer digital and video camera equipment. I strongly urge those interested in obtaining some of these filters described below to obtain them in 46mm size format. There are two benefits in using the 46mm format. The first benefit with the 46mm format is that it may be attached to a 2” [48mm thread] diagonal, eyepiece, or other accessory, by the use of a 48mm-46mm step down ring. These rings run less than US$8 each, and are very handy items. The ring adapts the filter thread size to the 46mm format.
The second benefit is that the 46mm filters can still be used with 1.25” diagonals, or in 2” diagonals using 1.25” eyepieces, by use of a special adapter. There is no simple adapter available currently for converting the threads of a 1.25” [28.5mm thread] eyepiece to the 46mm format. One can be fabricated by taking a 1.25” filter cell, and removing the glass. I usually have to trash the glass upon removal, because the threading is so tight on these smaller filters that it is easier to break the glass using a nail and hammer, with the filter on a block of wood. Once the shards are carefully removed with pliers, the cell is washed in soap and water, and dried. The retaining ring is removed, as it is not needed. To complete the adapter unit, a 27mm-37mm step up ring is required, and also a 37mm to 46mm adapter ring [step up type]. The first smaller ring fits loosely in the 1.25” filter cell. It is secured using two part, 5-minute epoxy resin, and some graphite. These are commonly available at most hardware stores; the graphite is available as a lock lubricant. Alternatively, some mechanical pencil leads can be crushed finely. Take 1 part of each, mixing thoroughly. The graphite makes the resin much stronger, and it becomes black, rather than clear in color. Apply a small amount to the outside of the outward, “male” threads on the 27-37mm step up ring, where it fits into the 1.25” cell. Gently place the filter cell onto the ring thread, and make sure it seats squarely onto it. Rotate the filter cell with the ring under it (doing this with the step up ring on a table is a good way to do this), to ensure an even coating of resin is applied to the threads of the filter cell. Before allowing the final 20 minutes needed to cure the resin, take the time to ensure that the filter cell is centered with the step up ring. Do not fret too much, but try to get it as close as coaxial as possible. I found the best way to do this is by rotating the step up ring, and watching the relative position of the filter cell above it from a side on perspective. If centered, there should be no detectable lateral movement in the filter cell upon rotation. If there is, adjust it slightly, and repeat. Check it after 5 minutes to make sure it did not move. Then, allow curing time of about 20 minutes.