Thoughts on High-Resolution Imaging : CCD cameras can produce fantastic lunar and planetary images, but only if attention is paid to the basics.
There is no such thing as a perfect all-purpose telescope. Each design has strengths and weaknesses. Some instruments excel for deep-sky observing, others for wide-field photography, and still others for high-resolution views of the Moon and planets. To determine which instrument is best for planetary observation, it is tempting to examine each telescope design in exacting detail. But first we must define what constitutes a good planetary image.
Contrast and Resolution
Most amateurs agree that a planetary image must possess two important qualities to be considered good: high contrast and high resolution. High contrast allows us to distinguish subtle differences in brightness, while high resolution shows small details. Although these two qualities are often confused, they are not necessarily related - one instrument can deliver images with excellent contrast but limited resolution, while another can have very good resolution but low contrast.
To produce the best contrast and resolution, a telescope must have fine optical quality regardless of its design. Even today's sophisticated computer processing cannot compensate for images obtained with poor optics; otherwise so much time and money would not have been spent to fix the Hubble Space Telescope. But if decent optical quality is necessary, perfection is not, because optics with a wavefront error less than about 11/410 of a wavelength of light do not in practice create noticeably better images.
Another important issue for high-resolution work is the size of the telescope's objective. The image that a telescope forms of a point source such as a star is called an Airy pattern. It is made up of a central bright spot - the Airy disk - surrounded by concentric rings of increasingly fainter light. The angular size of the Airy pattern becomes smaller as the diameter of the telescope objective increases. Since the smallest visible details in an image depend on the size of this Airy pattern, the conclusion is obvious: the resolution of a telescope increases along with its aperture. While this is true in theory, our turbulent atmosphere usually prevents a large telescope from achieving its best performance.
A telescope is more than just mirrors and lenses. It has mechanical elements to support the optics, and if these are incorrectly designed, the lens or the mirror can be distorted and deliver poor images. Moreover, if the mechanical system that keeps the optics aligned is unreliable or if the telescope tube bends, the optical performance of the telescope will quickly fall.
One mechanical parameter that is very important for planetary observing but is often neglected by amateurs is the reliability and precision of the focusing system. It must be smooth and allow very small focusing adjustments.
The precision and stability of the telescope mount are also important. It is impossible to make the best of good optics if the telescope does not track a planet smoothly or if there are vibrations from the motor drive.
Although good optical and mechanical qualities are necessary, we must also know how to properly use a telescope in order to obtain the best planetary images. Of fundamental importance is thermal equilibrium. Any difference in temperature, even slight, between the telescope and the surrounding air can be a source of trouble. Such differences will cause air currents inside a telescope tube. They can also distort optical elements and create noticeable image aberrations.
Planetary imaging demands great care when you are focusing, since even tiny errors degrade images in the same manner as serious optical defects such as spherical aberration. The focusing tolerance of a given telescope is inversely proportional to the square of its f/ratio with a longer- focal-ratio telescope having a greater focus tolerance.
But we must also consider how the focusing system works. For example, the tolerance of a Newtonian telescope with a rack-and-pinion focuser can be quite different from that of a Schmidt-Cassegrain telescope, even though both instruments have similar effective f/ratios. This is because a typical Schmidt-Cassegrain achieves focus by moving the primary mirror, which is often f/2 (the amplifying power of the secondary mirror produces the longer effective focal length). Thus, while the positioning tolerance of an f/8 Newtonian's rack-and-pinion is about 0.002 inch, the position of the Schmidt-Cassegrain's primary must be good to approximately 0.0001 inch, which represents about 11/4400 of a turn of the focus knob!
Optical collimation is also extremely critical, though it is often neglected or ignored. The images given by a poorly collimated telescope are as awful as the sound given by an out-of-tune violin. I believe that poor collimation has given Schmidt-Cassegrain telescopes a reputation for producing "average" images. Because these instruments have a fast f/ratio primary mirror, they are very sensitive to collimation errors and their performance can be ruined by a fraction of a turn of a collimation screw.
Refractor or Reflector?
It's no secret that refractors have an excellent reputation as planetary instruments. Certainly some of this is well deserved. Refractors are less sensitive to thermal changes since they have closed tubes and light travels through the tube only once, rather than twice for a Newtonian and three times for a Cassegrain system. Moreover, temperature changes have only one-quarter the effect on the optical performance of a lens that they do on a mirror. Even so, thick lenses, especially apochromatic triplets, do require time to come to thermal equilibrium.
Unlike reflectors, a refractor has no central obstruction caused by a secondary mirror. Although the light loss due to such obstructions is small, obstructions do have an effect on image contrast. This must not be overemphasized, however, since the contrast reduction is subtler than many amateurs think. Anyone who doubts this can perform a simple experiment: place central obstructions of different sizes in front of a refractor's objective. There will be little effect on the image until the obstruction grows to be more than 25 to 30 percent of the objective's diameter.
On the downside, refractors, especially conventional achromats, create problems for CCD cameras that "see" wavelengths (colors) well outside the range brought to a common focus by the objective. This is less of an issue with lenses made from ED and fluorite materials, but even some of these give the best images only when a filter is used to restrict the wavelengths reaching the camera. Such filters add significant light loss to an imaging system.
From an optical standpoint, a high-end refractor with very good color correction will outperform a reflector of equal aperture for planetary observing. But something interesting happens when we factor in the practical issue of cost. Refractors are expensive compared to reflectors. A refractor, no matter how well made, cannot escape the laws of physical optics. When it comes to high-resolution views, a refractor cannot compete with a reflector of the same cost, which can have an aperture two or three times larger. This additional aperture gives the reflector greater resolving power and more than compensates for the loss of contrast due to the central obstruction. In good seeing conditions, a reflector of 8- to 10-inch aperture will withstand magnifications of 500[yen]. This would be unrealistic for a 4-inch refractor, since this magnification would show the Airy disk with an apparent diameter of 23 arcminutes, more than two- thirds of the Moon's diameter seen with the naked eye!
It is not due to chance that, as Jean Dragesco mentions in High Resolution Astrophotography, the best lunar and planetary images made by amateurs during the past 20 years have all been taken with reflectors between 8- and 16-inch aperture, and some of these telescopes had very large central obstructions.
There are many variations on the reflector available commercially - Newtonians, Cassegrains (classical, Ritchey-Chretien, and Dall-Kirkham), Schmidt-Cassegrains, and Maksutov-Cassegrains, to name the most common. Factors such as the figure of the mirrors, the presence of a corrector, f/ratio, and the size of the central obstruction cause the various designs to have slight differences in sensitivity to optical alignment, contrast, chromatic aberration, and so forth.
As with the refractors, one can endlessly debate the advantages and drawbacks of each reflector design for planetary work. But as critical as these differences seem, they are really the trees that keep us from seeing the forest. Experience has proved that any of these instruments can deliver extraordinary high-resolution images. Once a telescope meets basic optical and mechanical requirements, it is not a slight variation in central obstruction nor a fraction of a wavelength better optical correction that makes a major difference. The most important factor is something that an amateur cannot buy - know-how. The huge variation in the quality of photographic and CCD results obtained by amateurs using similar instruments is the best proof of that.
So when it comes to the question of what instrument is best for high- resolution imaging, the answer is, simply an instrument that its owner has mastered!
Results speak for themselves - digital imaging techniques allow today's experienced amateurs with small telescopes to create images that surpass the quality of planetary photographs taken in the past with large professional telescopes. But what are the reasons for such a leap forward?
One is that CCDs record more detail than photography in equal conditions of atmospheric turbulence, or, said another way, CCDs record comparable information in less-favorable seeing conditions. Furthermore, thanks to their digital nature, CCD images are ideally suited to computer processing.
Because CCDs require much shorter exposures than conventional photography, they are better able to fight the worst enemy of the planetary imager - atmospheric turbulence. CCDs cannot, however, do this all the time. If the average size of atmospheric convection cells - patches of thermally stable air that give rise to atmospheric seeing - are small relative to the aperture of a telescope, a planet's image will be blurred for CCDs as well as for photography or visual observation. In such cases, the advantage of CCDs is small or even nonexistent.
On the other hand, if atmospheric convection cells and the aperture of the telescope are of comparable size, a planet's image remains sharp but undulates. In this case, the short exposures possible with CCDs offer a significant improvement in image sharpness.
Why do CCDs require shorter exposures? It is not, as many people falsely believe, just because CCDs have high sensitivity. The main reason exposures can be short is that for a given telescope a CCD can work at a shorter effective focal length (and hence faster f/ratio) than is needed for photography. A typical lunar or planetary image obtained with a CCD having 10-micron pixels will require only about half the effective focal length needed to record the same information on photographic emulsion, even a fine-grain film such as Kodak Technical Pan. This results in halving the effective f/ratio and a corresponding four-time decrease in the exposure time.
To produce a good planetary image requires more than just recording a large amount of information in the original exposure. Indeed, most raw planetary images displayed on the computer monitor appear soft and drab. Images must be processed to bring out the maximum detail locked in the digital data.
One processing technique - unsharp masking - is well known from the days of photography. But what used to require hours of painstaking darkroom work can now be accomplished in seconds with a few clicks of the computer's mouse.
This and other processing techniques, such as high-pass filtering, deconvolution, and image restoration, are so easy to implement with a computer that many amateurs push the process too far and produce image artifacts. Furthermore, these techniques have other drawbacks, particularly the amplification of image noise. This is most pronounced when the original image is underexposed and the processing is extreme. Obvious cures are to increase the exposure time (or better yet combine several short exposures that freeze the astronomical seeing) and restrict the amount of image processing.
No amount of processing can compensate for a poor original image. That is one reason professional astronomers are working so hard on adaptive optics. The aim is to improve the image before it reaches the CCD. Elaborate image-processing techniques are less valuable than a small improvement in the raw image. Getting the best results requires obtaining lots of images and selecting the best ones rather than overprocessing an average image. All the steps needed to produce a high-resolution view make up a chain, and if any link is defective - should it be the atmosphere, the instrument, or the processing - the result will be ruined. CCDs do not allow high-resolution planetary imagers to ignore bad seeing, be negligent with their telescopes, or have blind faith in image processing. CCDs do, however, give them the opportunity to tremendously improve their results and productivity when they pay attention to the details of the imaging process. And that is a lot!
A software engineer for Thomson-CSF, Thierry Legault lives in the suburbs of Paris. His Web site, www.club-internet.fr/perso/legault, contains several excellent articles on high-resolution CCD imaging.
|Printer friendly Cite/link Email Feedback|
|Publication:||Sky & Telescope|
|Date:||Jan 1, 2000|
|Previous Article:||A Sunrise Total Eclipse.|
|Next Article:||Personalizing the Planets.|