- Rebuilding my CGE Pro
- COUNTING SUNSPOTS WITH A $10 OPTICAL TUBE ASSEMBLY
- Hubble Optics 14 inch Dobsonian - Part 2: The SiTech GoTo system
- iStar Optical’s Phantom FCL 140-6.5 review
- Who’s Afraid of a Phantom: Istar Phantom 140mm F/6.5, that is?
- SHARPSTAR 94EDPH APOCHROMATIC REFRACTOR
- My Losmandy G11T review
- FIELD TEST: THE NOH CT-20 ALT-AZ MOUNT
- SkyTee-2 Alt/Az Mount Review
- SharpStar Askar ACL200 200-mm f/4 astrographic telephoto lens
- A review of the Unistellar EVscope
- Astrotrac 360 tracking platform – first impression
- FIELD TEST: CARL ZEISS APOCHROMATIC & SHARPEST (CZAS) BINOVIEWER
- Omegon 32mm 70º SWA eyepiece review
- Review of iPolar hardware and software for polar alignment
CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.
Digital Astrophotography Beginner's Guide
Twenty-five years ago I abandoned astronomy to pursue
other interests. Today, I return to the hobby and discover a booming field
ripe with technological
innovation. The bad news is that everything I knew has to be learned again,
the good news is that the sky is full of rewards that were unthinkable
a quarter of century ago. This beginner's guide is not intended as a dazzling
of good astrophotography but rather as a mundane account of my new adventures
in astro-photography, complete with mistakes and growing pains. Amateur
is one of the few areas where the internet has lived up to its hype: a
staggering amount of free information is available on-line, shared by an enthusiastic
and talented community: this article is my way to thank all those who have
contributed knowledge, hardware tips and great software.. Clear skies to
Analog or Digital? Decision Time
The amateur on a budget has
three astrophotography options: digital cameras, web cams or , conventional
photography. These choices are not mutually
exclusive and, if you already own a digital camera, it is quite possible
both digicam and web cam photography for less than $500. In fact, this
is the path
I chose for my own investigations. Let's first examine the possible choices:
Having worked with conventional photography in my early days, I advise you to ignore that route today. While great images can still be obtained on film, if you have no experience, conventional astrophotography could be frustrating: focusing is hard, exposure somewhat random and, even in the days of 30 minutes mini labs, the delay between the shoots and the results can be agonizing. Believe it or not, digital images produced by today's amateurs are often better than what major observatories published in the 70s. Also the experience gained in conventional astrophotography will be lost if you choose to push your investigations further in the future. If that wasn't enough, conventional photography is also the most expensive choice in the long run.
Digicams take acceptable single shots (Olympus 2100 UZ, Scopetronix Adapter, no processing) but the setup can be extremely bulky
Digital Cameras are hot! One can find excellent second hand digicams such as the members of the Nikon 99x family (or even the more recent 4500) for a little less than $500 and still coerce an afocal adapter in the budget. The strong points of digital cameras are their high resolution, good quality sensors, different zoom levels, good color sensitivity, good single shot abilities and, in some models, good long exposure (bulb) abilities. However, digital cameras are not perfect: paired with an eyepiece, they often suffer from vignetting. Their weight and bulk can stress your mount. They will definitely affect its pointing accuracy. Vibrations will force you to use delayed or remotely controlled shots.
You don't want to drop that gun on your toes!
Unless you shoot in RAW or TIFF mode, the jpeg compression algorithm will actually work against further digital processing. Working in RAW or TIFF mode carries an additional price: it is relatively slow and its storage requirements will increase the cost of your imaging system. Don't forget that image quality may, in theory, suffer a bit since the diagram lenses are often made of several lenses: when one adds the two or three optical interfaces of the scope to the several lenses of the eyepiece and of the camera, one can quickly reach a dozen of interfaces. Finally, digital cameras do not defy Newton Laws: gravity drags them to the ground, action-reaction breaks them apart! While this may sound like a joke, this isn't. Look at the size and bulk of a typical setup and imagine that thing hooked to your focuser, in the dark...
This being said, Digital Photography is an extremely fast moving area and a very credible imaging choice: if you have a relatively large budget, a Canon D60 on rock solid mount is an outstanding tool. It's likely to get better and cheaper tomorrow, why not wait a bit?
web cams are small and light
At first, with their low resolution (typically 640 by 480) and low image quality, web cams seem an unlikely candidate for astrophotography. Looks can be deceiving: amateurs with 10 to 12 inches telescopes routinely use web cams to obtain planetary images that surpass what analog photography and very large telescopes could achieve in the 70s. Because of their acquisition speed, web cams are able to take advantage of the fleeting instants of good seeing that conventional photographers will miss. Since seeing is the limiting factor as far as planetary imaging is concerned and since planetary imaging is a very important part of what the amateur on a budget can do, a web cam is the ideal partner for the budget minded astrophotographer. They are also the cheapest ($120 gets you a conventional/deep sky dual mode camera and an adapter), do no suffer from vignetting and do not stress the telescope mounts or focuser. Since they are software controlled and have no shutter or curtain, vibrations are a non issue (it helps not to trip on the USB cable). They feed streams of uncompressed BMPs into AVI movies upon which image processing can work its magic. They introduce the novice to the basic techniques of image acquisition and image processing that have become the backbone of today's digital astronomy: the knowledge gained from their use will not be lost should you one day move to "real" astronomical CCDs.
Web cams are not perfect though. The right web cam can be hard to find in the US, unassisted focusing can be tedious, their field of view is small, their color accuracy isn't great. Last but not least, they easily get dirty. However, since these shortcomings are either unimportant for the amateur on a budget or since work-around do exist, I believe web cam astrophotography is the way to go.
Here is a summary of the different imaging opportunities open to the amateur on a budget
|Web cams||Digital Cameras||Classical Photography|
|resolution||poor||good to excellent||excellent|
|focus||somewhat hard||easy||extremely hard, delayed results|
|field of view||small (can be a +)||varied||varied|
|vignetting||none||some, eventually a lot||some|
|weight||light||heavy (lever effect)||from medium to extremely heavy|
|vibrations||none||some, requires remote or timer||a lot, requires masking|
|image processing||extremely easy and rewarding||possible, harder||almost impossible|
|results||instantaneous||quick (save & download time)||delayed|
|tools and techniques||many available free, lots of information||somewhat proprietary||lab material harder and harder to find|
|startup cost||$150 - less than $500 gets you cooled results||from $600 up to several thousands||from $50 to several thousands|
|operating cost||nil||media||approx $15 per film roll, batteries.|
|additional material||cheap plastic adapter works well||wide field eyepieces, strong connections||standard adapters for most cameras|
|remote use||easy with cheap cable up to six meters||hard||impossible|
|risks||none||fall, cheap mount's brake could slip or break||same as digital|
|optics||one, the scope||up to ten lens interfaces||from none to many|
Choosing a web cam
The Philips ToUCam Pro (or Vesta Pro) also known as the 740K or the "egg" has become the de facto standard for amateur astronomy. It is based on a Sony ICX098BQ CCD chip that is both very sensitive to light and of relatively high quality. Be especially careful not to choose the non pro version, also known as 730K as it uses a CMOS sensor that is roughly ten times less sensitive to light and therefore less suited to astronomical use. The objective of the camera has to be removed and an adapter is required to align the camera into the focuser. While this is an area where you can save and recycle old film cans to build a home made adapter, I suggest that you do not: alignment is too important. A hardware guru called Steve Chambers found a way to modify the design of a typical web cam to allow longer exposures, suited to deep sky imaging. Should you get such a Deep Sky capable web cam? That mostly depends on the precision of your telescope mount, but I would recommend that you go for the modified web cam since the price difference is relatively small. Besides, you will only know how your mount tracks after you have tried it. If you can not find a ToUCam, check David Molyneaux's site: it offers valuable information about web cams and their suitability to astro-photography.
The Toucam and its adapter (left) The eye of the "egg" - yes, it's dirty (right)
Don't forget to install the camera software drivers before going out! It is also a good idea to become familiar with its control panel as it has some quirks that could disorient you during your first session. You will also need software to acquire the images: there is a truckload of them, most of them freeware. I personally have settled on a program called K3CCD Tools by Peter Katreniak, but this is simply a matter of preference since most programs use the same method to acquire images.
K3 CCD Tools is one of the many freeware available to acquire astronomical images from a web cam It even does support the long exposure modification. Your pre-flight checklist should looks like
- is the camera connected and recognized?
- is the resolution chosen?
- is a default directory defined for the night's session?
- is there enough disk space?
- is the default naming convention chosen?
- is the default acquisition speed selected?
- Quiz: what can you say about the picture on the left? (see answer below)
Getting started - the acquisition phase.
If you have an equatorial mount, your telescope
will be polar aligned. It will also have cooled down. With astrophotography,
more than ever, having
properly polar aligned makes your life easier. Having the scope cooled
down makes sure you'll get the best images possible: if you are using a
MAK design, allow 2-3 hours of cooling before beginning your session.
Our first target will be the Moon. Use a low power eyepiece to center the moon, then a high power eyepiece to select a zone of interest. Once you are settled, remove the eyepiece, insert the web cam and focus. Bingo, that's all there is to it: on the Moon, the web cam is usually able to use auto-exposure to obtain decent images. Your first images of the Moon could look like my first image. Not too bad, for a first try.
first light on the moon
The Planets: Jupiter, Saturn
Imaging the planets is trickier though. The first problem is to put the planet in the limited field of the web cam: use a low power eyepiece, center the planet, then use a high power eyepiece for fine positioning. When the target is perfectly centered, replace the eyepiece with the web cam The second problem is to focus the planet while keeping it in the field of view: if you have a telescope whose focus shifts the mirror a bit and whose drive backlash is capricious, this simple operation can be frustrating: be patient, we'll soon revisit the focus topic.
The third problem is to keep the planet on the CCD: if you have a motorized mount, make sure it is tracking accurately; if you have a manual equatorial mount, make sure fine movement in Right Ascension is enough to keep the planet centered. If you have a manual azimutal mount, imaging is still possible but so tricky and time consuming that I would probably despair. If you can do it, you have my admiration. The fourth and last problem is setting the correct exposure. If you are coming from the Moon to Jupiter, chances are that the Jovian monster will be grossly over-exposed - here is my very first Jupiter picture. Hardly satisfactory, however, the satellites are there!
Fine tuning the exposure and focus
In order to improve the image, we'll have to put the camera out of full automatic exposure mode image controls to access the camera control panel and set the gain to its lowest value. Gain acts like a booster on the sensitivity of the camera at the expense of increased noise levels.: The lower you can keep the gain, the better the image will be
On Jupiter, setting the gain to its lowest value, the frame rate to 10, the white balance to outdoor usually yields acceptable results. The frame rate of ten images per second is a good compromise between image quality and atmospheric turbulence. Don't be afraid to experiment a bit! Saturn may require a bit of boosting. Whatever you do, do not overexpose your pictures. Overexposed areas are saturated at 100% level and there is no technology in the world that can extract information from such a signal, regardless of the number of frames available. Under-exposing pictures is not nearly as bad since stacking and eventual complex analysis can extract meaningful data from multiple exposures.
single avi frames, unprocessed.
Depending on your telescope, you'll easily get results like the ones above. Now, this is much better! Can we improve on that?
The impact of the seeing
One critical factor in astronomy is called the seeing.Consider, for example, this single frame of Jupiter: even though it is slightly underexposed and grainy it is much better than the one above: rather than two dull bands it show three bands and a faint GRS. The telescope and operator haven't changed, that atmosphere has! In fact, the above pictures of Saturn and Jupiter were taken on different days and the seeing was much better for Saturn. The seeing does matter a lot! If you think you have done your best and you are not getting good results, do not despair and try again another night.
single avi frame, unprocessed, good seeing.
Focusing, gamble or science?
Reaching the perfect focus can be hard, especially when one knows the enormous impact a slightly missed focus can have on the resolution of the final image. While it is possible to reach decent focus on Jupiter thanks to its satellites (push the gain and focus until the satellites are as small as possible) or on Saturn (if Cassini is visible, then the focus is about right) it can be impossible on stars. A very cheap tool can help you focus better: the Hartmann mask. The mask is basically an objective cap with holes (2, 3, 4) that you put on the scope. Select a reasonably bright star, center it and try to focus: as long as you are not in focus, you'll see 2 (2 or 3 or 4) images converging. Merge them and you have reached the Holy Grail of perfect focus. (Hartmann masks were invented in the early 17th century by a certain Scheiner, a two holes Hartmann mask should be called a Scheiner mask)
a standard web cam, a home made Scheiner mask and a modified "deep sky capable" web cam
How much power is enough?
In astronomy, greed is not rewarded. Those who jump for the highest magnification will not necessarily get more details. While this is obvious to any visual observer who tries to push his scope beyond its ability, the matter is less obvious with astro-photography where one feels that something can always be improved. Unsatisfied with the images I was getting with my Celestron Ultima 2x Barlow lens, I made the mistake of buying a TeleVue PowerMate 3x. Of course, the images became even worse! The reason is simple: both with the Celestron and with the TeleVue, I was pushing my equipment far beyond its resolving power! The table on the right shows a "back of the envelope" estimate of the ideal barlow given a defined web cam and optical instrument. It should be taken with a grain of salt since the Rayleigh criterion (one of several method to estimate the resolving power of an instrument of a certain diameter) can be used to demonstrate that the Cassini division should not be visible in instruments where it definitely is visible.. Nonetheless gives a useful ballpark figure - a 3x Barlow on an ETX-125 is definitely too much.
Where diameter, focal length and F/D are obvious, Rayleigh is the theoretical resolution of the optic, ideal is the ideal sampling rate according to Nyquist (basically half the Rayleigh criterion), pixel is the number of arc seconds per pixel on a ToUCAM at 5.6 micron per pixel with the default F/D and Barlow is the hypothetical Barlow lens needed to achieve ideal sampling. (Note: the Sparrow criterion seems to be better than the Rayleigh criterion)
In the good old days, cloudy nights allowed the amateur astronomer to maintain a plausible social life.. then came image processing! Today, it is not uncommon to see amateurs devoting more time to their image processing than to their observing sessions! What is the hype all about?
raw Saturn frame.
Digital image processing is a set of extremely powerful enhancement techniques that can be used to improve the appearance and actual resolution of your pictures, but before examining in in more details, let's remember that.
- image processing does not beat fine focus.
- image processing does not beat proper exposure.
- image processing, just like statistics, can tell any lie.
This being said, image processing is what makes web cam astronomy so rewarding! Since it is as much an art as it is a science, it is also a lot of fun to explore and to invent ones own cooking recipes. Again, many freeware programs are available. In my opinion, Registax (by Cor Berrevoets) offers the best lunar and planetary results in a very easy to use package. Registax combines three enhancement methods;
several 5 seconds AVIs of Saturn, stacked and processed in Registax
Image selection if you watch your avis frame by frame, you will notice that while some of the frames are sharp, others are fuzzy: this mostly reflects atmospheric turbulence. Registax offers both an automated and manual way of selecting the best frames.
stacking of 100 frames and wavelet transform
Stacking is the least controversial image enhancement technique. Its goal is to average the information contained in a large number of frames to increase the signal to noise ratio. While the details can be subtle (sum stacking vs median stacking for example) the principle of operation is easy to understand: noise , random in nature, cancels itself while the signal is proportionally reinforced.
This Jupiter shows a fair amount of actual detail, but is also clearly over processed. The red circle artifact is not a processing artifact but an operator error at capture time.
Wavelet transforms detect edges in images. Adding those edges to the summed image literally extracts and emphasizes the details it contains. One of the big advantages of the wavelet transforms is that they allow the user to select the size of the features he wants to extract: Registax supports several processing scales: level 1, 2 and possibly 3 are the most useful for planetary imaging.
NOTE: Image processing raises some ethical questions: is an image resized, sharpened and re-colored in Photoshop closer to reality than it is to a plain drawing? Must Jupiter Great Red Spot absolutely be red when it is not in reality? The "correct" answer probably does not exist.
The Cassini-Eudoxus region, processed in Registax.
And what about the deep sky?
You may have noticed that even though a "deep sky capable" web cam was used in this tutorial, very little has been said about Deep Sky imaging. The reason is very simple: it is extremely hard to get good deep sky results with cheap telescope mounts. While I am sure I can turbo-charge and fine tune my ETX-125's drives to obtain some results, I also know that I am ultimately doomed by its staggering periodic error. One possible strategy to work around a poorly driven equatorial mount is to blindly and automatically take a set of exposures and select the ones that are not blurred because they were shot during windows of precise tracking.
Nice stars in Praesepe pictured when testing the Hartman mask. The odd shape is due to tracking errors. I liked the color difference!
Wrapping it up..
That's all folks... now you probably have enough information
to answer the Quiz: the picture of Saturn in K3CCD Tools can't be a single
frame. It is
way too sharp and well defined. It has simply been pasted in the K3CCD
for aesthetic purposes.. I hope this article will help you define your
path in the rich universe of astro-photography.