Astrophotography Skills - Selecting Targets

From McWiki
Jump to: navigation, search

There is one task that is so fundamental and so obvious that it may escape your attention as a skill that you need to develop at all: picking something to photograph.

You can't just point your telescope to a random place in the sky and expect to capture a beautiful image. You will need to pick a suitable target object (and find it). An ideal astrophotography subject has a balance of three characteristics:

  1. It is located at a position in the sky that you can see from your observing site at a time of night that works for you, and at an altitude that minimizes distortion;
  2. It is a suitable size for your gear and your skill level.
  3. It is a suitable brightness for your gear, available time, and skill level.

Let's go through those criteria in a little more detail now.

Position in the sky

Your selected target object should be located at a good position in the sky on the night, and at the time, that you will be doing your imaging. "Good position in the sky" means that it is a good balance of three things:

  1. Freedom from obstructions;
  2. High altitude;
  3. Position relative to the Meridian.

Freedom from obstructions

This, of course, is the most important factor for choosing a target object, even though it is not very technically sophisticated. The object has to be visible from your location at the time of night you are available to observe. It can't be below the tree line, behind a building, or in the glare of streetlights. There is no point in considering all the various more technical aspects of target selection, described below, if your target isn't in your line of sight.

You need to consider more than just physical obstructions, too. Also consider whether anything in your line of sight may produce bad atmospheric disturbances that should be avoided. For example, an object located directly over a chimney is in a bad location. Even though it may not be obstructed by being behind the chimney, warm air currents rising from the chimney may cause turbulence and ruin the stability of the air between you and the target.

TheSkyX planetarium program, with custom local horizon photo installed
A planetarium program in which you can customize the displayed horizon to represent your actual location is very helpful for making such plans. In some planetarium programs you can "block out" various altitudes around your location, helping you understand what is obscured by your local horizon. On some programs, such as TheSkyX, you can actually insert a 360° panoramic photo taken from your observing location, so the computer display represents the actual surroundings you will see. This is extremely valuable for planning what objects will be visible, and when.

If you frequently image from the same location, and especially if you have a permanent observing site, it is well worth taking the time to do an accurate survey of the horizon all around your site, and to enter this information in your planetarium program of choice so you can plan your observing against the horizon and obstacles that you will face.

High altitude

The next most important factor in an ideal object location is its altitude above the horizon. The higher an object's altitude, the less atmosphere you have to look through to see it. And, the less atmosphere you are looking through, the clearer your view. Aside from allowing you to breath and protecting you from radiation, the atmosphere is not your friend. It is filled with dust and moisture, and is a roiling mass of turbulent air currents, which is what causes stars to "twinkle". That's why professional observatories are on mountain tops, and the Hubble telescope is in outer space; it's to minimize the amount of atmosphere they look through.

Let's look at some simple diagrams to understand this.

Observing two stars, one overhead and one near the horizon
In this not-to-scale diagram we are standing on the surface of the earth, and the earth is surrounded by a ring of atmosphere. Let's consider two stars, one directly overhead, and one near the horizon.
Showing the distance, through the atmosphere, to a star directly overhead
When we are looking at the star directly overhead, the red line in this diagram shows the amount of atmosphere we have to look through.

Demonstrating that the amount of atmosphere traversed to observe a star near the horizon is greater than the amount for the star directly overhead
Now let's move that red line, without changing its length, down to point toward the star near the horizon. As you can see, there's quite a bit more atmosphere to look through - beyond the end of the red line - than there was for the overhead star.

So, all else being the same, objects that are higher in altitude will involve looking through less atmosphere, and you'll get a clearer view of them. And, for a given object, observing it at the time of year and the time of night when it is the highest above the horizon will give you the best view.

Position relative to the Meridian

All else being the same, choose an object that lets you stay on one side of the Meridian during your imaging session.

What on earth does that mean? What is the Meridian? And why should it matter which side I'm on, or whether I'm on both sides?

Showing the Meridian - a line from NCP to due South
The Meridian is an imaginary vertical line in the sky running from the North Celestial Pole (approximately the star Polaris), up past the Zenith (the spot directly overhead), and down to the horizon at due South. It divides the sky into "the East part" and "the West part".
Meridian Flip on an equatorial mount

This is important because an equatorial mount must be positioned quite differently depending on whether the mounted telescope is pointed to the east part or the west part of the sky. To point toward the west, the telescope must be on the east side of the mount with the counterweight on the west side of the mount. To point toward the east, the telescope must be on the west side of the mount, with the counterweight on the east side of the mount.

If you are following a target across the sky, sometime around when the target crosses the Meridian, the mount must be flipped to the other side. This is called a "meridian flip".

Try to do all of your imaging of a given target on one side or the other of the Meridian, to avoid having to do a meridian flip.


There are several reasons why you should avoid doing a meridian flip during your imaging session.

First, if you are stacking multiple images together, images taken before and after a meridian flip will be upside down relative to one another. This complicates your stacking operation, although most stacking software can handle it.

Second, after the meridian flip, you will have to find your target again, consuming precious time you could be using for imaging. Since the object will now be upside-down, reproducing the exact framing will be an additional challenge.

More important, some mounts and some telescopes exhibit undesirable mechanical motion when you do a meridian flip.

On an equatorial mount, a meridian flip causes the mount to change from lifting the counterweight with its right ascension motor to lowering the counterweight with its right ascension motor. The other side of the teeth in your drive gears will now be pressing on each other. Backlash in the right ascension drive gears will almost certainly result in the mount's periodic error changing when this occurs.

Worse, some optical designs have play in parts of the optical train, and after a meridian flip these parts will shift to "the other end" of their range of movement. Some examples of things that might shift after a meridian flip include:

  • Mirror flop: telescopes with movable main mirrors, especially SCTs that focus by moving the main mirror, will have the mirror shift slightly after a meridian flip. This will throw the image out of focus.
  • Camera flop: the mechanical linkages between your telescope and your camera, including the focuser, extension tubes, and the camera itself, may shift slightly as gravity pulls down on the camera from the other side after a meridian flip. This may throw the image out of focus, or out of collimation.
  • Guide scope flop: if you are using a separate guide scope, the guide scope or the guide camera may shift slightly after a meridian flip, throwing your guiding out of calibration. You may hear the term "differential flexure" used by serious imagers. This refers to attachments - usually guide scopes - changing their relationship with the main scope as components flex and flop.

Positioning summary

Ideally, then, you would like to begin imaging an object when it is either several hours East of the Meridian, or just after it has passed to the West side of it, so that you can image for several hours without having to deal with a meridian flip. At the same time, you would like your target to be at as high an altitude as possible during your imaging session. And, most importantly, of course, your line of sight needs to be free of obstructions and local air disturbances.

Apparent size

The next critical factor in selecting a object for imaging is its apparent size - that is, the height and width that it appears to occupy in the sky. (The actual size is not important - a large object that is far away and a smaller object that is relatively close could both appear to be the same size.)

Example beginner problems

Let's illustrate the importance of matching the size of your imaging target to your gear with two unfortunate stories of disappointing experiences in astrophotography. (The confused beginner in both of these true stories is me, although some equipment brands have been changed to protect the other innocents.)

Let's first make what is probably the most common error. We have purchased a mid-sized (130mm) reflector telescope on an equatorial mount, and, since we are already experienced at normal photography, we already own a Nikon D800, a modern DSLR camera. This camera has 36 megapixels, a large number by today's standards, and this must be a good thing, since camera manufacturers always post their high number of megapixels as a measure of quality.

With a suitable connector, the Nikon is mounted on the reflector and pointed at Saturn with the expectation of capturing a beautiful image, similar to what we've seen in magazines.

A simulated image of Saturn taken with a short-focal-length telescope and a DSLR.  Where is Saturn?
That's not what happens. At first, we think Saturn isn't in the image at all, and that we have simply made some kind of an error in pointing the telescope.
A simulated image of Saturn taken with a short-focal-length telescope and a DSLR.  Saturn is there - but unacceptably small.
However, on much closer inspection, we realize Saturn is there - it's just ridiculously small in the image. That's frustrating. What's the point of having 36 megapixels when only the handful of pixels in the very centre of the image are actually covering our target?

Now let's switch equipment and make the opposite error. We have purchased a fairly large SCT - a 235mm Celestron - on an equatorial mount. Since we were buying Celestron gear, we thought it would make sense to also buy Celestron's astrophotography camera, a Nightscape CCD. We have seen many beautiful astrophotographs of the famous Andromeda Galaxy, M31, so we point at that and take a 5-minute exposure.

Simulated image of M31 with a long telescope and a small camera chip
The result looks like there is something wrong. The field is just awash with light, as though it was badly out of focus. The famous spiral arms and dust lanes aren't visible at all. What's going on?
The chip is capturing only the centre part of this large object
What's going on is that the galaxy is far larger than the field covered by this camera with this telescope. All we have done is image a small rectangle inside the bright core. The galaxy is too big to be imaged with this setup.

What size is the right size?

When considering the size of an object to photograph, we should start by considering how big we would like that object to appear in our final image. Do we want the target to dominate the frame, or to be relatively small against a large background field of stars or other objects? Both are valid compositions, and it depends what we are trying to accomplish with our image. Also, that's not a black-and-white distinction; and we can always take an image where the object is slightly smaller than the whole frame, then crop the frame down so the object fills the resulting rectangle.

For the rest of this discussion, let's assume that we would like our target to dominate the frame.

Warning, Math!: I'm going to use some very basic math, below, to evaluate the suitability of certain target objects. Although I like to "work the math" this way, it is by no means necessary. Just take a quick test image of a potential target object and, if you find it to be too large or too small, try another, smaller or larger as appropriate. Hint: Messier objects tend to be large and bright (that's why Messier was able to see them with his primitive telescope), while NGC objects tend to be smaller and dimmer (since that catalog was made 100 years after the Messier list, and optics had improved.)

We should also consider how big a final image, in pixels, we actually need. Do we want an image to display on a computer monitor, or something that will be printed? Computer monitors usually require about 100 pixels per inch. (Did you just say "Wrong! 72 dots per inch!"? Actually, that standard hasn't been the case for quite some time. The monitor I'm using as I type this, for example, is 110 dots per inch.) Printers generally require about 300 pixels per inch (although, in my experience, with astrophotographs you can usually get away with less resolution, say 200 dots per inch).

So, think about how large an image you hope to produce, and on what medium, to get an idea of what pixel dimensions you require. Then, you can use your calculated image scale to determine the range of apparent sizes that suitable target objects would have.

Some examples

Let's work through some examples.

Suppose I want to produce an image that will look nice in a web browser window on a computer screen. I want the object to dominate the frame with no more than, say, 10% extra space around each of the edges. Computer monitors come in many sizes, with 1024 x 768 pixels being a common resolution. So, I might try to have an object that covers about 800x600 pixels to display in a window on that sort of screen, with a bit of margin on the sides.

My camera has 3348 x 2574 pixels on its chip, so if I can fill one quarter of my camera frame or more I should be able to crop to a size that will work. My camera's image scale when used with my AT8RC telescope is 0.7 arc seconds per pixel. (In fact I will probably use 2 x 2 binning for my image since the atmosphere would never produce 0.7 arc seconds per pixel accuracy. However we can use 1 x 1 binning and 0.7 arc seconds per pixel to do this calculation because if we increase the binning we effectively reduce the number of pixels by the same factor.)

The largest object I can image would be one that completely fills my camera chip. At 0.7 arc seconds per pixel, the maximum width would be 3348×0.7 = 2343 arc seconds, and the maximum height would be 2574×0.7 = 1801 arc seconds. Or, approximately 39 arc minutes by 30 arc minutes.

I said that the smallest object I would want would fill roughly 1/4 of my camera chip. That would be 585 x 450 arc seconds, or 10 x 7 arc minutes.

I can look up object sizes in a variety of databases or online sources to help me select objects that fall within that range of sizes.

Good choices

Here are some examples of good choices for this chip and telescope. In the following images, the dark rectangle exactly reproduces the total image that would be produced by the camera and telescope in this example.

Simulation of M51 on camera frame
Messier 51, the Whirlpool galaxy, is 11 x 7 arc minutes.
Simulation of M101 on camera frame
Messier 101, the pinwheel galaxy, is 28 x 26 arc minutes.
Simulation of M81 on camera frame
Messier 81, Bode's galaxy, is 29 x 14 arc minutes.
Simulation of NGC6888, the Crescent Nebula on camera frame
NGC6888, the Crescent nebula, is 18 x 12 arc minutes.

Bad choices

The above are examples of objects that are a good fit to the size of my camera chip with my chosen telescope. Now, here are some bad choices.

Too small

First, some famous objects that I might like to photograph are too small for this chip and telescope.

Simulation of NGC4038/4039 in camera frame - a bit too small
The antennae galaxies, NGC4038 and NGC4039, are too small at about 5 x 3 arc minutes.
Simulation of NGC2392 in camera frame - much too small
NGC2392, the Eskimo nebula, is worse, at only 48 arc seconds square. It is far too small to be visible with much resolution in this setup.
Simulation of Saturn in camera frame - almost invisible
Worst of all are planets. Although they vary in size with their distance from Earth, Jupiter, Saturn, and Mars are generally around 30, 15, and 10 arc seconds in diameter.
Too large

There are also objects that are too large for this setup. The images below are much larger than the image produced by my example equipment, and the image area of the equipment is shown as a yellow rectangle.

Simulation of Veil Nebula in camera frame - far larger than the piece of sky this equipment covers
The veil nebula is a great example. This system of several NGC objects is almost 3° – 180 arc minutes – in size. I could, however, image the small components one at a time.
Simulation of M31 in camera frame - too large to image the spiral arms
And, as we saw in the example above, M31, the Andromeda galaxy, is also too large, being about 190 x 60 arc minutes in size.

To image these large objects, I would need to either use a shorter focal-length telescope, or a focal reducer, or both. Or, I could take a mosaic of adjacent rectangles of sky, building up a larger image in small pieces. (This, however, is considerably harder and beyond what I would recommend for a beginner.)


Finally, the brightness of an object can determine whether it is a relatively easy or relatively difficult target. What we especially care about is not the simple magnitude of the object, but the surface brightness, which is the brightness divided by the surface area.

Objects with a high surface brightness (such as Messier 51, and Messier 42) can be imaged with relatively short exposures. For example, Messier 42 requires only a handful of exposures of a few minutes duration each, and Messier 51 can be nicely captured with a dozen or two dozen five-minute exposures.

On the other hand, objects with a low surface brightness, such as Messier 101, may need longer exposures than your gear can tolerate, or may require stacking more exposures than you have time to collect. For example, you might need to take 10-minute exposures to capture the subtle detail in the arms of Messier 101. If your mount and autoguiding cannot produce good 10-minute exposures, you would have difficulty imaging this object unless you could stack several dozen five-minute exposures together; and you may have trouble collecting several dozen five-minute exposures in your available observing time.

Plan in advance

As you can see, you can put quite a bit of time, and even some mathematical analysis, into the selection of your imaging targets. Whatever amount of pre-planning you decide to do, do it in the daytime; don't waste your precious time outdoors, under the dark sky, trying to figure out what to image. Make at least a rough list of suitable targets in advance.

Software can help you with this. Some integrated control programs, such as TheSkyX contain features to make observing lists. Or, simple dedicated planning programs such as AstroPlanner are available for this function. With such software, you can say "show me all objects of such-and-such type, visible on this date and time, in this part of the sky, and which have the following range of sizes and brightnesses". There's your target list for the evening.

If you are the type that likes to make lists, you can make a long-term plan for imaging targets. Make a list of all the objects you might like to image that are a good match for your scope and camera (primarily by their size) and that are visible from your location. Determine the range of dates during the year when each is favourably positioned in the sky, and sort your list by these dates.

Now you have a target list for the entire year. In fact, probably for multiple years, since you will not likely be able to produce images of satisfying quality of all the targets on your list during a single season.

Update your list periodically: when your equipment changes, when your surroundings change, or when your skills or interests change.

Finally, you may end up with more than one telescope, or more than one camera. Organize different lists for all combinations of your gear. You may even find that certain seasons are best-suited to one combination of 'scope and camera, while a different combination is best for the objects available at another time of year.

For example, for me, there are more targets in winter that require a large field of view and more that require a narrow higher-magnification field in summer. So I tend to mount a short-focal-length scope as winter approaches, then change to a long-focal-length scope in spring.

Now that we've selected a good target, our next step is to find it.

Simulated sky images on this page were produced with TheSkyX Professional Edition.

Back to Astronomy Writings Up to Richard's Astronomy Section Back to Astrophotography Essential Skills


You need JavaScript enabled for viewing comments