by Don Selle
Auto-Guiding Part 1
To Autoguide or not to Autoguide – it’s all About Noise
One of the questions that beginning astrophotographers ask these days is whether they need to acquire the equipment and learn the skills associated with autoguiding. With current technology, this is a valid question. This article will help you answer it. A little background info should help start us out.
Astrophotography, unlike most other types of photography, is all about capturing very dim objects many of which include elements that are barely brighter than the sky around them. Because of this, astrophotographers post process their images to increase the contrast between the object and the background sky thus creating images that look attractive to other people.
Due to the properties of light, dim low contrast objects mean a higher noise content (lower signal to noise ratio or SNR) than the subjects typical of daylight photography. If you’ve ever tried to take photos of a nighttime candle lit event, you will have noticed that noise lives in the dim parts of the image. It shows itself as “grain” and off-color blotches in the dimmer areas that are hard to ignore when you brighten the image.
This is the central dilemma of astrophotography. Contrast enhancement to improve the look of astroimages also enhances the noise which always hides in the dimmer areas. Camera manufacturers and image processing and software companies have spent major money developing technology to fight this problem for daylight images. Much of the daylight photography anti-noise technology can be and is used in astrophotography. For astrophotography this technology is not enough by itself since the objects we image are so much dimmer that a candle lit soiree. We must also attack the problem at its source. This means we must increase the SNR of the images we capture.
One of the ways to do this is to take very long exposures. This increases the amount of light captured for the image, and as a result improves the SNR of the image. In the days of film astrophotography, this meant single image times measured in hours.
These days electronic cameras allow us to capture many shorter duration sub exposures and combine them to achieve the same improvement in SNR that such a long single image brings. Stacking shorter sub exposures also brings other advantages. As an example, a white light violation or a satellite track is less likely to photo bomb your image, as these subframes can be discarded.
Longer images mean that the scope must very accurately track the movement of the night sky. Since no mechanical device is totally accurate, this requires that very small tracking adjustments need to be made to the mount in real time. That’s where autoguiding comes in.
From the very beginnings of astrophotography, guiding the mount on a star near to but separate from the camera field of view was used to improve the telescope tracking. Guiding, when done right, ensured that the exposure would be sharp, without trailing stars and smeared galaxies. Guiding was done manually by the telescope operator who made very small changes to the telescope pointing to keep the guidestar at the center of the cross hairs in a guiding eyepiece.
In the days of glass photographic plates, even the largest professional telescopes were guided manually. Early in his career at the Mount Wilson Observatory, Edwin Hubble would guide the 100-inch Hooker telescope manually which was a very tedious task. Milton Humason later took over those duties and took many of the spectrographs that lead to Hubble’s discovery that the universe is expanding.
Autoguiding uses a second camera to take very short exposures, each of which is analyzed by a computer to determine if the guidestar has changed position in the frame. If it has, either the mount tracking is off, or the field is drifting due to an imperfect polar alignment. Small adjustment commands are sent to the mount so that it “catches up” and keeps the stars in position on the imaging chip, and this keeps the stars in the main image from trailing, and from being blurred.
The introduction of CCD cameras to professional astronomy changed everything. Since the cameras are electronic, and the images are digital, this meant that small CCD cameras and computer software could replace the astronomer making tracking adjustments, and autoguiding was born. By the late 1990’s, CCD cameras had replaced most film cameras amateurs were using for astrophotography. This meant that like professional telescopes, amateurs could autoguide making astrophotography much more accessible.
These days, mounts track better. The newer CCD and CMOS based cameras are much more sensitive, and have inherently lower “shot noise”. Faster, lower f/ratio telescopes are also available in bigger apertures and catch more photons than older slower designs. Image capture software is also better, and most have a feature called live stacking, where individual subframes are integrated (stacked) on the fly as they are downloaded. This begs the question, do you really need to autoguide?
If you can do without autoguiding, the payoff can be significant. Eliminating the guide camera, guide scope or off axis guider and the software and routine associated with autoguiding can save you some cash and significantly simplify your imaging.
The capability of today’s imaging equipment pushes the answer to a qualified – maybe. In fact, these days many are imaging without autoguiding. If your equipment and imaging goals allow unguided imaging, you should go for it.
You can answer the question yourself in a two-step process. First measure the capability of your equipment.
- Set up your equipment for an imaging session. If you have an equatorial mount, carefully polar align your telescope. This will minimize image drift that is due to polar misalignment.
- If your mount has periodic error correction (PEC) take the time to go through the routine necessary to calibrate it.
- Rotate your camera so that the principal axes of the imaging chip align north/south and east/west.
- Pick a 4th to 6th magnitude star near the celestial equator (near declination 0 degrees) as this will be where the declination drift due to polar misalignment will be greatest. Errors in your mount tracking will be in the RA axis.
- Take a series of images of increasing duration and inspect them. There will be an image of a given duration in which the stars will be slightly oval. Keep your sub images under this duration.
- If you are using LRGB filters, use the L filter, but keep in mind that your color filters will be less exposed as they pass less light for the same length exposure as your luminance filter.
The second step is to find an optimum sub exposure calculator and use it to determine the optimal exposure for your camera. If the optimal exposure is less than or equal to the maximum time your mount can track without oblong stars, you should try imaging without autoguiding and see how it works for you.
These calculators consider all sources of noise in the image and determine the exposure time such that those that can be subtracted from the image dominate those which are more random in nature. It turns out that sky brightness is one source of noise which can be significant. If you are imaging in fairly bright skies, your optimum sub exposure time will be quite short.
This is good if you want to avoid autoguiding, but the downside means you must take many sub exposures to equal the SNR of a long overall exposure. It also means that your total exposure time will be longer than it would be if you were taking longer sub exposures under dark skies. Still this trade off means that it is possible to image the brighter objects without autoguiding, and also makes electronic assisted astronomy very accessible.
You can read the theoretical background of all this including the math here at the link below. While there are more comprehensive approaches to this subject, this paper outlines a simplified approach which gives good approximations, and is the basis from several sub exposure calculators.
https://www.ccdware.com/Files/SubExposures.pdf
With a manual calculator, you will need to use manufacturer specs or measure several characteristics of your camera like gain and read noise and plug them into the camera. You will then need to take an image of a dark part of the sky and determine the sky background brightness. Make sure they are all in the correct units, plug them into the calculator and out comes a recommendation for your optimal sub exposure duration. You will need to do this once for and OSC camera and once for each filter if you are using a monochrome camera.
There are some image capture programs like SharpCap Pro that have this functionality built into them. Its Smart Histogram function automates, and this makes the process easier. If you are interested in how this works and some of the theory behind it, check out the SharpCap documentation here:
https://docs.sharpcap.co.uk/3.2/#The%20Smart%20Histogram%20Brain%20Window
What are some of the limitations you might face if you do choose to forego autoguiding? Here are a few that might help you decide which way to go:
- Lower focal length optics are more likely to allow you not to autoguide, as they are much more forgiving to small changes in star position on the imaging chip. This is due to the fact that your image scale (arcsec per pixel of your camera) will be much larger. More drift due to polar misalignment can be absorbed and very small mechanical flaws may not be as noticeable.
- This approach also favors those using one shot color cameras, as you are getting all of the color data in every sub frame, keeping the overall imaging time lower.
- You will be able to create images that you are happy to show to friends and family. Publication quality images may not be possible, especially using longer focal length optics.
- It is very unlikely that you will be able to take narrow band images without autoguiding due to the fact that these filters reject all but a small percentage of the total light collected by the telescope.
- While SNR theoretically increases proportionally to the square root of the total number of frames, you will not achieve in practice. This means that you will take many more short sub exposures to achieve an equivalent SNR if you autoguide and take longer sub exposures. To achieve the same results of 50 minutes of total exposure time using autoguided 10 min sub exposures (5 ea.) will take you 68 min of sub exposure time at 2 min unguided (34 ea.) to achieve a similar SNR. On top of this your total time imaging time will include almost 7 times the amount of dead time due to downloading all the extra sub exposures.
- In order to process your images, your workflow will need to accommodate a large number of sub exposures. This can take a considerably longer time to accomplish and may be tedious as well.
Still, if you are interested in creating decent images of brighter targets, today’s equipment and software does make unguided imaging feasible. For many imagers, the savings in cost and lower complexity involved are very welcome advantages. It may be worthwhile to give unguided imaging a spin.
Next month we dive into some of the techniques and details of autoguiding.
