4
$\begingroup$

I'm a fairly experienced photographer and have enjoyed taking lots of photos of the night sky with my existing gear - wide angle starscapes over pretty hills, mostly.

I'd like to move on, having seen what's possible with telescopes and mounted cameras. I've read around, and am now confused....

The plan sees me getting a telescope, tripod and mount of some sort (equatorial I believe I need?), and periodically putting that carefully into my car, driving into a rural area with dark skies and clear horizons, and pointing it at interesting things - my urban gardenless flat really isn't a good location for this.

I have Nikon full and crop frame and Olympus m4/3 cameras available - I believe I'd need a T mount adaptor to fit them? Is one likely to be hugely better than the other for this in a way I've not foreseen? I know the Nikon has much better low light performance but the Olympus puts more pixels on the subject with its small chip and has the very useful live time mode. I'm used to raw processing, but I'm expecting there'll be more specialist software I'll need to learn.

  • Am I anywhere near right or is this all mad?
    • What sort of things do I need to watch for on telescopes?
    • Am I going to trash a telescope by moving it around like this? I'm used to moving big lenses which I'm hoping is tolerably similar (accepting the very different optical designs)?
    • What have I completely missed?
$\endgroup$
  • $\begingroup$ For astrophotography, you'll need patience, knowledge, good technique and a lot of money. My advice would be to start with visual astronomy first. It's much easier, it's way cheaper, you'll learn a lot and there are many beautiful things to see, even from a light polluted city. A good 6 or 8 inch dobsonian costs less than many parts needed for astro photography. $\endgroup$ – Eric Duminil Aug 25 '18 at 8:33
6
$\begingroup$

Ah, welcome to the world of AP... check your sanity at the door, and that loud sucking sound you hear is your bank account being emptied.

My first and best piece of advice for you is to NOT buy anything at all right off. Find your local astronomy club(s) and join. Most clubs have at least a few people doing imaging, and you would be well-advised to meet these people and learn from them.

AP is not like conventional photography at all, for the most part. It's closer to digital signal processing. It is a blend of science, engineering, and art. What you know from conventional photography is not useless information, but you have a whole slew of new things to learn.

When we talk about telescopes and equipment for AP, we don't talk about "tripods" all that often. A tripod is one kind of support structure for a mount head, which does the actual work. The mount's role is to carry the telescope, camera, and associated equipment and point it at the object you wish to capture. While the concept is quite simple, the execution is extremely complex.

There are two fundamental problems that we must overcome to capture good images of astronomical bodies. The first is the fact that the majority of these objects are very faint. While people often think the purpose of a telescope is to magnify, this is a misnomer: an astronomical telescope's primary purpose is to collect more light and concentrate it into a smaller area (the eye or the image sensor). Magnification, while not irrelevant, is a secondary concern (and comes with its own issues). Because these objects are so dim, we must resort to long-exposure photography. While typical daylight photography uses exposure times measured in tenths and hundredths (even thousandths) of a second, exposure times for astrophotography often are measured in whole minutes. Actually, total exposure is often measured in hours, with individual sub-exposures done in minutes.

And this is where the second key problem emerges. While our shutter is open to capture an exposure of 1 or more minutes, we have to contend with the fact that the objects we are capturing are moving relative to our instruments. Due to the rotation of the earth, a given star, planet, galaxy, nebula, etc... in the sky above moves across the sky from east to west in an arc with a radius depending on the object's position relative to the celestial north or south pole. In a full day, an object moves 365 degrees around the pole (well, we rotate, but for practical purposes, the object moves). In a single hour, it moves 15 degrees. In one minute, 15 arcminutes. And in one second, 15 arcseconds. This may seem like a small amount of motion, and when you look up at the stars in the sky, they do seem to be fixed in place. But, when you start to reduce your field of view by using a telescope or camera lens, the motion starts to become more and more obvious.

As a conventional photographer, you are probably familiar with the standard alt-az style mount motion. Alt-az is short for altitude-azimuth, and a typical camera tripod works this way. The mount angles up/down (altitude) and left/right (azimuth) to point at an object. In order to follow an object across the night sky, then, you have to move the mount in both axes. The rates of motion between the axes will vary from one another as well as depending on the location of the object in the sky above. To automate this process requires a great deal of computerization, and even then it is difficult to do so accurately enough to handle the demands of long exposure imaging. And even if you do overcome those issues, you will find that the object you are following appears to rotate in the field of view as it crosses the night sky, a phenomenon we call frame-rotation. This is because the frame of reference of an alt-az mount is the horizon, not the sky itself.

The solution to this issue is the Equatorial mount. An equatorial mount has four axes. The first two, altitude and azimuth, are only used for alignment purposes, not for motion. They are used to align the third axis, known as Right Ascension (RA), with the celestial pole. Once that is accomplished, motion in the RA axis shifts the telescope's aiming point east and west. The fourth axis, Declination (or Dec), handles north and south. When properly aligned, the mount then needs only to move in the RA axis and at a single rate of speed to follow an object across the sky. This doesn't even require a computer, and thousands of mounts have been manufactured with clock-drive mechanisms to handle this motion. Computerization only enhances the process and makes it more accurate.

But not all equatorial mounts are created equally. Many just do not have the accuracy needed to handle long exposure imaging, and the level of accuracy needed increases as you increase focal length, focal ratio, and exposure time. A mount that could handle a 300mm camera lens or 80 mm short-tube refractor and produce 60 to 120 second exposures might not be able to handle an 8" f/4 Newtonian telescope.

Weight is a common issue. Mounts like this are all rated for a specific payload weight amount. Payload here includes not only the telescope, but also any cameras, adapters, finders, guide scopes/cameras, and mounting hardware - everything attached to the top of the mount. But these payload ratings are typically given for total payload maximums. When doing AP, the general wisdom is that you should keep your imaging payload to 1/2 or less the mount's rated maximum. So if your mount can handle, say, 40 lbs. of weight, for imaging you shouldn't load it with more than about 20 lbs. The rule isn't a hard limit, and not all manufacturers rate their mounts the same way, but it is definitely a strong suggestion.

As I said, the tolerances become tighter as you increase focal length and ratio. First, let me discuss ratio. The focal ratio of a telescope is calculated by dividing the focal length by the aperture. Thus, if you have a telescope with a 200 mm (roughly 8") aperture and a focal length of 1,000 mm, you have a focal ratio of f/5. If you have a focal length of 2,000 mm with the same aperture, it's f/10. In general, the lower the focal ratio, the better for imaging. The reason is the concentration of light. Imagine you have the two telescopes I just mentioned, both 200 mm apertures, one f/10 the other f/5. Now imagine you have identical cameras in each and are tandem-mounted on a GEM that can handle the payload without issue. Now, let's say you are imaging an object that appears roughly circular and evenly illuminated. Now, let's say in the shorter scope, f/5, the object you are imaging covers a spot on your camera's image sensor about 100 pixels across. The total area of pixels covered is then around 7,850 pixels. Let's say each pixel receives about 20 photons of light per second and you run a 1 minute exposure. Thus each pixel gets a total of 1,200 photons of light in the exposure, for a total of about 9.4 million photons collected. Now, if we look at the f/10 telescope, we find the image is larger. A longer focal length increases magnification. Double the focal length, you double the magnification. So now the object we are imaging has a diameter of about 200 pixels on the image sensor. This means that it has an area of about 31,400 pixels - double the diameter means 4 times the area. Now, because the aperture is the same, we do NOT get more light, we get the same number of photons, they are just spread out over the larger area. Instead of 20 photons per second, each pixel gets only 5. In one minute, each of them gets only 300 photons. The total number is the same, about 9.4 million, but they are less concentrated. This means the image is only 1/4 as bright as that in the f/5 scope. To get the same level of exposure, you would need to increase your exposure time by a factor of 4 to 240 seconds. Yes, you get a larger, more detailed image, but it takes you four times as long, and this can be a problem.

No mount is perfect and all will show drift over time, even with autoguiding technology. Let's say that the mount you are using has an error that is equal to about 1 pixel per minute in the f/5 telescope. In that 1 minute exposure, everything is stretched by 1 pixel. This is hardly noticeable and not likely to be a problem. Let's say you had a star near the center of the image that has a diameter of 5 pixels. In the exposure, you would see it as 5 pixels wide by 6 long. Hardly noticeable. But now let's look at the f/10 telescope. Because we double magnification, we now have 2 pixels of stretching. Now the star is stretched 2 pixels. Still, this isn't a huge problem, but we also need to increase exposure, so we quadruple the stretching to 8 pixels. Now we have significant stretching in our image and blurring of our object (and that's assuming the mount is THAT accurate, which is not a given). This is why astrophotographers refer to shorter focal ratios as "faster" telescopes.

Aperture is typically the critical measure for a telescope. But for AP, this isn't always the case. A lot of very high quality imaging is done using short-tube refractors with fairly small apertures. This also gives you a much wider field of view. And this, in turn, reduces the exacting tolerances necessary for the mount.

However, this isn't ideal for all targets. And this is another key problem with AP: there is no one-size-fits-all solution. What is ideal for planetary imaging is not ideal for imaging large nebulae, which is not ideal for imaging most galaxies... you need to consider multiple factors in equipment selection. Personally, I use an 8" f/10 SCT (with a Barlow lens giving me f/20) for planetary imaging, an 8" f/3.9 Newtonian for globular clusters, medium-sized nebulae, and most galaxies, and an 80mm f/3.75 refractor for wide-field nebulae and the Andromeda Galaxy (which is 3° across, so very large).

But the mount is the key to all of it. The conventional wisdom in AP is to focus on the mount first. You can have a million dollar top-of-the-line camera and a million dollar top-of-the-line telescope, but if you don't have them on a mount that can provide accurate motion, you won't get good images. Alternatively, you could put a mediocre camera and telescope on a good mount and get pretty decent images - the mount is THAT critical to the process.

Prices on these kinds of mounts varies, and you can easily spend into the tens of thousands. The cheapest mounts in this range are designed to handle DSLR Cameras and lenses, usually not more than 300 mm in length (perhaps a little shorter). The price on a mount like this is around \$350 to \$600 USD, depending on quality and capability. And, again, these are limited in what they can handle.

Stepping up from that are basic German Equatorial mounts with powered tracking ability. While computerization (What we call GoTo) is not required, it is usually found on these kinds of mounts and allows for the addition of autoguiding technology. Autoguiding is the use of a second camera and, usually, telescope that is harnessed to software to lock onto a guide star and make minor adjustments to the mount's motion to keep it on-target. Autoguiding will not make a bad mount into a good one, it will simply allow an already good mount to capture longer exposures. The lowest-price mounts in this range usually have a payload weight maximum of 30 lbs. or less and a price tag between \$800 and \$1,200. However, I wouldn't actually recommend ANY of them. The two most popular here are the Celestron AVX, running about \$900, and the SkyWatcher HEQ5, at about \$1,100. The SkyWatcher might actually be usable, but the 30lb limit means only 15lbs for imaging, and that's very low indeed. You can still probably fit a short-tube refractor on here with a DSLR and lightweight guiding package. The Celestron AVX promised to provide entry-level AP capability, but the reviews of the mount show that it hasn't lived up to its promises. This mostly seems to be due to poor quality control in the production line, resulting in some that work just fine, and a lot that do not provide the accuracy needed for imaging. They are still usually fine for visual observing, just not for imaging.

The next step up finds mounts capable of handling between 40 and 45 lbs. of payload with prices between \$1,500 and \$2,000 USD or so. This is where I STRONGLY recommend people start. While they are not cheap (though you can often find them used for about 30% or so less), they often have the quality and accuracy necessary. There are four models in this range I recommend: The Celestron CGEM II, Orion Atlas EQ-G, SkyWatcher EQ6-R, and the iOptron iEQ45 Pro. My personal preference is the iOptron (I have the early version of this mount and love it), but all should be competent. My mount is capable of 45lbs and I run it at around 26 (yes, over the recommendation, but hot hugely so, and it handles it fairly well) with an 8" f/3.9 Imaging Newtonian and CCD Imaging camera.

You CAN go with a cheaper mount. But you rapidly lose capability and accuracy. The lower mounts aren't really intended for imaging use, the heavier ones are generally engineered to stricter tolerances needed for imaging.

One nice thing about these four mounts, and most others in this range, is that they all use a common interchangeable modular mounting system for attaching the payload. I personally have three sets of payloads and can switch them out in a matter of minutes. It also means that if you buy the mount now with a low-end telescope option, you can later upgrade the telescope itself without having to get a totally different mount. It also means you can have different payloads for imaging and for visual observing. I've never heard of anyone having too much mount for imaging.

Now, as for cameras....

DSLRs are not the best tool for the job, but they are widely used all the same. There's a few considerations here.

First, DSLRs tend to have a great deal of signal noise - noise which must be dealt with. The preference for AP is generally thermoelectrically cooled CCD imaging cameras. These use Pelletier cooling systems, heat sinks, and fans to cool the image sensor and dramatically reduce signal noise. Remember, we're dealing with faint objects, and the noise can often drown out the image signal in the exposure. These kinds of systems also typically better isolate noisy components and reduce their effect on the image. Also, the nature of CCD technology makes them inherently less noisy than CMOS sensor technology prevalent in DSLR cameras. However, CMOS cameras are improving and many manufacturers are now offering cooled CMOS cameras.

With or without a cooled system, you still will need to capture calibration frames. There are three common types: Dark frames, Bias frames, and Flat frames. Dark frames are images taken with the shutter closed. They are done with the same exposure time/gain/ISO settings and, preferably, the same temperature range. These are then digitally subtracted from the "light" frame image you capture. Some cameras actually have built-in technology for this (my Nikon D5000 does). But for AP, it is usually done through capturing separate dark frames and processing them in bulk after the fact.

The Bias frame is also done with a closed shutter, but for the shortest exposure time possible. This maps the noise created when the image data is read off the image sensor itself. It is also digitally subtracted in processing. I don't believe any DSLRs have technology to do this automatically, though I could be wrong.

The flat frames are a different beast altogether. They are taken with the camera still attached to the telescope or camera lens. To produce flat frames, you want even illumination entering the optical train (e.g. the front of the lens/scope) and an exposure length that produces an average of 50% illumination. These images are used to map the optical system's light and dark spots and even out the illumination (compensating for problems like vignetting).

As I said, noise is a major problem in DSLR cameras. In well-lit images such as daylight, well lit interior, or flash photography, the noise is nearly irrelevant. But in low-light long-exposure imaging, it can be dramatic. This is especially true if you use higher ISO levels to increase sensitivity. CCD imaging systems are not immune, but they are designed specifically to handle it.

CCD and CMOS systems designed for AP are also typically more sensitive. The term Quantum Efficiency (Qe) describes the sensitivity level of an image sensor to various wavelengths of light. Essentially, it indicates what percentage of photons that hit the image sensor are detected. A typical DSLR has a peak Qe around 30-40%, with some newer ones moving up closer to 50% and a handful of newer ones in the 70% range. CCD and CMOS systems for AP typically have peak Qe starting around 50% and some have peak Qe around 95%. This means much more efficient imaging and shorter exposure times. The actual Qe varies across the spectrum, with the peak being the highest sensitivity.

These issues, so far, are fairly simple to deal with, though often require longer exposure and/or more frames in the stack (I'll come back to stacking). There's one problem with DSLR cameras, however, that isn't as easy to deal with. With the exception of the Canon 20Da, Canon 60Da, and Nikon D810A, all DSLRs, and, in fact, nearly all digital cameras in general including cell phone cams, webcams, etc..., have a built-in filter for color-balance improvement. In most cases, it's mounted directly over the imaging sensor "chip" and also provides some degree of protection from dust and damage. This filter is designed to reduce the amount of deep red and near IR energy that reaches the sensor. Human eyesight is not very sensitive to these longer wavelengths, and the filter helps produce a more "life-like" color balance to match human vision. Without the filter, images tend to be shifted toward the red. Software color balance profiles can reduce this significantly, but for conventional imaging, the filter is the best option.

However, for astronomy, this is a problem. There is a lot of astronomically interesting light in the deeper red part of the spectrum. When an electron in a Hydrogen atom falls from its second-highest to third-highest energy state, it emits a photon at a wavelength of 656.28 nanometers. We call this Hydrogen Alpha, and it's a major component of emission nebulae and star-forming regions in galaxies. When you see red in a photo of a nebula, it's often H-Alpha. Some nebulae, such as the Rosette Nebula, are extremely heavy in H-alpha. Unfortunately, the filter in a DSLR blocks the majority of this light from reaching the sensor - something like 90% if I remember correctly.

The solution for this problem is to not have the filter in place. To date, only three factory-issued DSLR models have been released with this filter removed (actually replaced with one that is transparent to these wavelengths). For other DSLR cameras, you would need to have the camera modified. There are several services you can find on the internet that will do this for you, but there's a cost. It is not something I recommend people do themselves unless they have experience with camera repair and/or can afford to run the (very large) risk of ruining the camera. Doing so also makes the camera ill-suited for continued conventional imaging, unless you get a clip-in or screw-on replacement filter (more cost). You can do AP without a modified camera, but you will be significantly handicapped.

Of course, for best results, a monochrome CCD and filter wheel is the best option of all, but this significantly increases the budget.

You mentioned one of your cameras has more/smaller pixels. In most cases, this is not actually desirable. Larger pixels tend to be more sensitive (more collection area to gather photons), and more pixels may provide higher resolution, but also means larger file sizes and MUCH slower processing. The Hubble Space Telescope's primary cameras have a resolution around 2048 x 2048 pixels (about 4 megapixels). The issue is not resolution as much as sensitivity. I've worked with some very high end astronomical imaging CCD's with resolutions of only around 1 megapixel: but they produce FAR better images than any DSLR on the market.

When it comes to brands, Canon and Nikon are far and away the leaders, with Sony gaining some ground. But overall, Canon is really the brand of preference among astrophotographers using DSLRs. Canon has been very welcoming and accommodating to the AP community. They produced the first two pre-modified cameras well before Nikon produced their first. Canon has made a great deal of information available, including detailed specifications and even firmware source code. Nikon is slowly coming around, but has been a LOT slower to do so and still has not offered anywhere near as much support as Canon. Because of this, you will find a lot more resources and, especially, software for camera control of Canon cameras than Nikon or anyone else. For people just getting started in AP and on a budget, I generally recommend looking for a used pre-modified Canon DSLR as an imaging camera. Sites like CloudyNights and Astromart frequently have these in their classified ad sections.

And now... software...

A major difference between conventional photography and AP is the processing. I already mentioned calibration images. But the big deal here is image stacking.

Most Digital AP, whether DSLR or CCD based, uses stacking processes to produce better images. Stacking is, essentially, a signal processing method using statistical analysis. You capture a set of images of the same subject and exposure time. You then run each through a calibration routine which removes the dark and bias signals and then applies the flat frame to even out the illumination. The images are then aligned (nearly always through lining up stars in each frame). At the end of the alignment phase, any pixel all the way through the stack represents the same point in space. Then, each pixel in the stack is statistically analyzed. For example, if you have 100 images in a a stack, and a particular pixel in the stack has brightness values between, say, 100 and 105, you average them out to come to what the most likely "true" value is, say 103. In many cases, a weighted average is used in which pixels outside of 1 or 2 standard deviations above or below the average are discarded entirely, and only values in a consistent range are finally averaged together to produce the most likely "true" value. This has the effect of reducing noise and, using drizzle techniques, can actually increase resolution. Even without drizzle, you end up with a much better overall image, with greater clarity and less noise. The larger the sample size, the better. My most recent image of M51, the Whirlpool Galaxy, uses 3.5 hours of 120 second (2 minute) exposures (so over 100 individual exposures). The more exposures I use, the better the final result and the more faint detail I can gather.

Of course, this processing takes time. The larger your image files, the longer it takes. My main imaging camera produces image files in FITS format (the most common format for AP) that are about 50mb in size each. 100 of those comes to about 5 gigabytes, not including the flat, dark, and bias files (usually 10-20 of each). This ends up taking a lot of storage space and a lot of processing time. There are other file format options, but it is critical that you use a lossless format (like FITS or TIFF), and the only way to reduce those files is through compression, which works against you in processing (you have to decompress each frame in processing, so you save space, but dramatically increase processing time).

There is good processing software available, some for free. I do most of my calibrating and stacking currently using DeepSky Stacker, which is a free product. I do most of my image capture in Nebulosity, which is not free, but works well for what I need and I got before the last price increase. While live view on a camera is helpful, it's not necessary with a program like Nebulosity which will loop exposures and provide metrics to help achieve accurate focus. You can get MUCH better focus this way than through the live-view on a camera. It will then also allow you to automate the capture, setting the number of exposures, length of exposures, what to name the files, where to store them on your hard drive (typically these store directly to hard drive, not on the camera), and in what format (usually FITS or TIFF, not the camera's proprietary RAW/NEF/etc... or JPG - and you really do NOT want to use JPG, it's FAR too lossy).

After you've captured, calibrated, and stacked, you will then want to do post-processing. Some people use GIMP, which is an open-source option. I mostly use Photoshop, which has a lot of capability and there's a wealth of information and tutorial advice. I am also trying to learn to use PixInsight, which is probably the best astrophotography processing software on the market, but is neither cheap nor easy to use.

At this point, you probably are starting to see how very difficult Astrophotography can be. Doing "pretty decent" AP is neither cheap nor easy. I'm about 10 years into it and have spent about \$10,000 total (that includes several bad choices in equipment and also includes a backyard roll-off-roof observatory), and I'm just now feeling relatively competent. My main issue now is time and weather.

For someone just starting, my recommended starting budget is at least \$2,000 and preferably closer to \$4,000 or more. At \$2,000 you can get a competent mount and low-end telescope to get you started (and, depending on deals and what used equipment is available, a pre-modified DSLR or lower-end CCD camera). You CAN get started cheaper than that... but you rapidly start losing capability and flexibility. A \$1,500 or so mount can handle a camera/lens combo for very wide field imaging, a smaller scope for normal wide field, a medium sized scope for narrower field deep-sky (e.g. galaxies and smaller nebulae) and longer focal length scope like an SCT for planetary imaging. If you go with a lesser mount, you're not likely to have as much flexibility there. And, again, most of these mounts allow for interchangeable payloads. So you can start with something simple like a short-tube refractor, and add to your equipment options as you have the ability to do so (my annual tax refunds have often ended up going to the hobby).

In the end, I return to my original advice: joint a club. Most clubs have people doing this you can learn from. A lot of clubs have people regularly selling well-maintained used equipment to finance their next upgrade. Some clubs have loaner equipment to borrow. Most either have a dedicated observing site or have scouted out good locations away from light pollution. And many also have their own small observatories with higher-end equipment, and, sometimes, even imaging equipment. Most clubs in the US have annual dues less than \$50 per year. It becomes a very cost-effective way to pursue the hobby.

I know I've given you a lot here. And I've honestly barely scratched the surface. AP is not for the weak!

Good luck and clear skies!

$\endgroup$
1
$\begingroup$

Everything you've said is pretty much on the money. You're definitely going to be wanting an equatorial mount (mount and tripod are pretty much interchangeable here), because equatorial mounts don't have the issues with field rotation and zenith deadzones that alt-az mounts use. You're also going to want a mount with the ability to add automatic tracking, otherwise your objects will streak across the field during your longer exposures. Don't worry about damaging the telescope just by moving it around and transporting it; just treat it gently and make sure it doesn't roll around and hit things that can damage it. Store it appropriately, too, in a dry place with dust covers on.

For your cameras, it can end up being completely preferential whichever one is better. The Nikon's low light performance could make it so you can take fewer or shorter exposures without worrying about noise; however, you might decide that you prefer the live time mode. Whatever the case, consider investing in a remote switch for either so that you can trigger the shutter without physically touching the camera in order to avoid vibrating or moving the camera/telescope.

Software wise, you're going to be wanting to find something that can do image stacking, bias subtraction, flat fielding, and other special processings. Deep Sky Stacker and PixInsight are two popular options, and there are even more alternatives around. You'll also need to take those special exposures: biases, darks, and flat fields in order to reduce noise and other distortions even further.

Finally, you're just gonna need to practice. There's a lot that goes into just figuring out how your personal workflow is best, and you can only get better at it with time.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.