Deep sky objects in the night sky are faint, very faint, extremely faint, or unbelievably faint. For visual observations we use telescopes with large aperture (diameter) to collect as much light as possible. When doing astrophotography we may also use large telescopes, but we have another option – we can extend exposure time, and this way collect more light. Somewhere between visual and photographic observations lies a technique called live stacking. The idea behind live stacking is to capture relatively short exposures (a few seconds) using the sensitive camera and stack them into one final image immediately as they come. This way you (and all friends around) may observe the target on the screen and see how it becomes more and more revealed while the new captures come and are stacked into it. Let’s dig a bit deeper into the live stacking process.
The image above was captured using live stacking with an 8-inch SCT telescope and PlayerOne Saturn-C color uncooled camera.
So what do we need to have and to know? The idea was already described – we point the telescope with the camera to our target and start to do exposures, let’s say 5 seconds each. There will be not much visible on the one exposure, but the software we run aligns each incoming image with the previous one, and then stack (add) them together continuously. This way every 5 seconds the image we see is getting better and better. At some point, we may stop the process and save the image.
Let’s now take a closer look at some aspects of the live stacking technique.
Exposure
Exposure time in live stacking determines how often the displayed image will be extended with captured data. We do not need to select very short times (1s or below). Such short times are reserved for planetary and lucky imaging techniques when we want to “freeze” the seeing and reveal much detail. On the other hand exposures longer than 10 seconds would force us to wait quite long for visible updates. From my experience times in the range of 2-5 seconds are fine, while up to 10-15 seconds may be useful for very faint objects, or when imaging with narrowband or multiband filters. I do live stacking with a color camera or a mono camera with wide-band filters using 4-second exposures.
Camera parameters
Almost all modern CMOS cameras will be fine for live stacking. When doing short exposures the most important factor is the low read noise of the camera. CMOS cameras have low read noise in general, but to get into the very low noise parameter range we need to raise the camera gain. For the live stacking process with exposure time in the range of 1-10 seconds, we want to have a gain set somewhere in the middle between the camera unity gain and maximum gain. We need to get over the unity gain to escape the quantization noise. We also do not want to set the gain to maximum to keep the pixel depth and dynamics at a reasonable level. Unity and maximum gain values for the camera in use need to be checked in the manufacturer’s page. For example ASI 183 MM Pro camera I use the gain value of 300 (30dB) at 4s exposures.
CCD cameras can also be used for lucky imaging, but due to their larger read noise, single exposure time needs to be increased.
Telescope and mount
Any telescope will work. A large aperture will collect more light, and a long focal length will give a larger scale. But a small telescope will also give a lot of fun with live stacking – see how happy are the Seestar users. For small cameras, no field flattener is required. Small tilt or small field curvature does not trouble so much for live stacking, because by definition the outcome of this process is not perfect. So neither we nor other viewers need to be concerned about small flaws in the picture we see live on the screen.
The mount we use needs to be able to track the sky. It can be either equatorial or alt-az mount – both will work fine unless we would like to do live stacking for hours – the field rotation in alt-az mount will become too large to align all the frames. Tracking quality does not have to be perfect – we are doing only a few seconds-long exposures. It is good to have a mount well polar aligned (eq mounts), or star aligned (alt-az mounts), so the target will not drift away from the field of view during the live stacking session. Another option is to use guiding, but that adds another component to our setup, that we would rather keep simple.
Example setups
For travel live stacking – SW Adventurer GTi mount, Askar FMA230 telescope, and PlayerOne Saturn-C camera. It gives almost 3×3 degrees FOV with simplicity and is lightweight.
For backyard live stacking – Celestron AVX, SCT 8″ with 0.63x reductor, ASI 183MM Pro. Nice and sensitive setup for galaxies, globular clusters, compact open clusters, and nebulae.
Software
There is quite a lot of software either dedicated to or capable of live stacking. The one I use is SharpCap – which works very well and has a big advantage, in that I already know this software 🙂
The Live Stacking process is started from the Tools menu – it opens the Live Stacking panel in the bottom part of the screen. I will not go into much detail on how to use live stacking in the SharpCap – there is extensive documentation available, and also plenty of tutorials. I will list a few bullet points with general remarks on the process, that may be useful and applicable also for other software.
Capture format and resolution. For a few seconds exposures the 8-bit color space is usually enough, especially when the gain is increased. Resolution usually will be set to maximum available. Binning can be set to 2 if the pixel scale is too high (like a small pixel with a long focal length telescope).
Exposure and gain were already described before, so set it to some reasonable values.
The raw stacked image will be dim, so we need to adjust its stretch and white balance (when using a color camera) to reveal the image. Usually, there is an automatic stretch option that can be later adjusted to our taste.
During the live stacking process, there are few things happening in the background:
Image alignment. Each next image is aligned to the already existing stack, so small imperfections of mount tracking will be eliminated. There are usually some options for the alignment process – like the minimum number of stars to align with, star detection sensitivity, or some noise reduction to apply before star detection.
Stacking. The simplest algorithm is to average all the frames – but then satellite trails and other single-frame flaws may persist in the stack. A better option is to use the Sigma Clipping algorithm, which eventually will remove that kind of problem.
Calibration. Live stacking usually depends on the high stretching of the final image. All the camera defects, bad columns, amp glow, and dust on the optics will be perfectly visible in these conditions. That’s why we should make calibration frames and use them during live stacking. Making calibration frames with SharpCap is very easy. 20-30 darks and flats will be enough.
There are of course other options to set or adjust – worth mentioning is Colour noise reduction – an excellent tool to remove color pixels from an uncooled OSC camera, or brightness and FWHM filter, that allows to automatically exclude frames of poor quality. Probably even more options will be introduced in the future versions of SharpCap.
It may take a bit of time to read the documentation and get the knowledge required to perform optimal live stacking and be happy with the results. But the learning curve for this process is really quick because you can see live the changes that each adjustment and option introduces.
All the images in this post were captured using SCT 8 inches telescope with a 0.63x reductor riding on the Celestron AVX mount. The cameras in use were the PlayerOne Saturn-C color camera with an IMX533 sensor (uncooled) and the ASI 183MM Pro camera with a cooled monochromatic sensor. Both are quite sensitive and low noise cameras, however, 183 suffers from amp glow and some bad columns, that’s why for ASI 183 proper calibration frames set was crucial to achieve a good outcome. Saturn-C camera on the other hand required a Color noise reduction option to be set to remove hot pixels from the uncooled sensor.
All images were captured in my backyard under the suburban sky at about 19.50 mag/arcsec2. Despite the visible light pollution galaxies fainter than 17mag were already visible in the Coma cluster after only 10 minutes of stacking.
When you take a closer look at the images from the ASI 183 camera you will spot an uneven background and some dust donuts. This is due to the fact, that flat calibration frames were not applied – I forgot about it. Lesson learned.
The image above shows single frames at exposure times 2, 4, and 8 seconds. The faintest stars recorded in these frames are 13.4, 14.4, and 15.3 mag respectively.
Clear skies!
