Heads up: this post contains links to gear we think is genuinely useful. Some are affiliate links, which means we may earn a small commission if you buy through them, at no extra cost to you. Here’s our full affiliate disclosure.
Point a camera at a galaxy for thirty seconds and you’ll get a smudge of gray noise with maybe a hint of something in the middle. That’s not a failure on your part — a single exposure, no matter how good your gear is, just doesn’t collect enough light to show a faint deep-sky object cleanly. The photos you’ve seen online, the ones with visible spiral arms or the wispy tendrils of a nebula, are almost never a single shot. They’re the result of stacking: taking dozens or hundreds of individual frames of the same target and combining them into one image where the faint signal stands out and the random noise fades into the background.
If you’ve already got a camera and a star tracker and you’re capturing frames but your final images still look grainy or flat, stacking is very likely the missing piece. It’s not glamorous, and it involves sitting at a laptop instead of looking through an eyepiece, but it’s the step that actually determines whether your night under the stars turns into a photo worth keeping.
What Stacking Actually Does
Every photo you take of a faint object is really two things layered on top of each other: the signal (the actual light from the galaxy, nebula, or star cluster) and the noise (random variation from your sensor’s electronics, heat, and the sky background). In any single frame, the noise is often nearly as strong as the signal, which is why deep-sky targets look so washed out straight off the camera.
Noise is random from frame to frame, but signal is not — the same photons keep landing in roughly the same pixels every time you expose the same patch of sky. Stacking software takes advantage of that difference. It aligns your frames on the stars (correcting for any drift or field rotation), then averages them together. Because the noise doesn’t line up between frames, it partially cancels out, while the real signal reinforces itself and gets stronger. The result is a dramatic jump in what’s called signal-to-noise ratio, which is really just a technical way of saying “how much of the picture is actual data versus static.” A stack of 40 frames doesn’t just look four times cleaner than one frame — the improvement follows the square root of the number of frames, so more really does help, though with diminishing returns.
The Four Frame Types You Need
Most beginners assume stacking just means combining your photos of the target. That’s part of it, but a proper stack also uses a few calibration frames that clean up sensor quirks before the real combining happens.
Light frames are the actual exposures of your target — the data you’re there to capture. Dark frames are shots taken with the lens cap on, at the same exposure length, ISO, and sensor temperature as your lights. They capture the thermal noise and hot pixels unique to your sensor so the software can subtract that pattern out. Flat frames are images of an evenly lit, featureless surface — twilight sky or a light panel works — taken at the same focus and aperture as your lights. They map out vignetting (the darkening toward the corners of the frame) and any dust motes sitting on your sensor or filter, so those imperfections can be corrected rather than baked into your final image. Bias frames are the fastest possible exposures with the lens capped, used to isolate the read noise that’s baked into the camera’s electronics regardless of exposure time.
You don’t need all four to get started — a stack of lights and darks alone will already look noticeably better than an unstacked single exposure. But once you add flats, the difference in a wide-field image especially around the edges is hard to miss.
Choosing Your Stacking Software
The two names you’ll run into constantly are DeepSkyStacker and Siril, and both are free. DeepSkyStacker has been the default starting point for a lot of beginners for years: you load your lights, darks, flats, and bias frames, hit a button, and it handles the alignment and stacking automatically with very little decision-making required. It’s a good place to learn the concepts without getting overwhelmed by settings.
Siril is the more capable option and has kept up active development, with a real community and detailed workflow guides, including a solid explainer from BBC Sky at Night Magazine that walks through the calibration and stacking sequence in more depth. It gives you more control over calibration, offers scripting for repeatable workflows, and runs natively on Windows, Mac, and Linux. If DeepSkyStacker feels too automatic and you want to understand every step of the pipeline, Siril is worth the slightly steeper learning curve. There’s also a helpful step-by-step DeepSkyStacker walkthrough if you want to start there instead.
A Basic Stacking Workflow, Start to Finish
The order matters less than making sure each frame type actually matches your lights in the settings that count. Capture your light frames on the target first, keeping exposure length, ISO or gain, and focus consistent across the set. Immediately after (or before) your imaging session, while the camera is still at roughly the same sensor temperature, shoot a matching set of dark frames with the lens cap on. Flats are best captured at the end of the session or the next evening at twilight, before you touch focus or rotate the camera at all, since even a tiny nudge will throw off the vignetting pattern they’re meant to correct. Bias frames can be shot any time, since they’re independent of temperature and duration.
Once everything is loaded into your stacking software, it will typically calibrate the lights using the darks, flats, and bias frames, register (align) the calibrated lights against each other using detected stars, and then integrate them using an averaging method that also rejects outlier pixels — useful for throwing out the odd frame ruined by a passing plane or a gust of wind. What comes out the other end is a single, flat-looking image with a much better signal-to-noise ratio than any of your individual frames, but it will usually look dim and gray. That’s normal — the stretching and color balancing that makes it look like the images you see online happens afterward, in a separate processing step.
Quick tip: always shoot in RAW, never JPEG, if your camera supports it. JPEG compression throws away exactly the faint tonal information that stacking is trying to recover, which defeats the purpose before you even get to the software.
Common Stacking Mistakes That Ruin an Otherwise Good Session
The most frequent mistake is mismatched calibration frames — darks shot at a different ISO or exposure length than the lights, or flats captured after the camera’s been refocused or rotated. Calibration frames only work if they match the conditions of the lights closely; a “close enough” dark or flat can actually introduce new artifacts rather than removing old ones.
Another common issue is simply not capturing enough light frames. A single 60-second exposure and a stack of five 60-second exposures will look similar; a stack of sixty will not. Beginners are often surprised by how much total integration time — the combined length of all your light frames — separates a mediocre result from a genuinely good one on the same target with the same gear.
Finally, skipping frame rejection settings during stacking is an easy way to let one ruined frame (satellite trail, sudden cloud, tracking hiccup) drag down an otherwise clean stack. Most software defaults to some form of outlier rejection, but it’s worth checking that it’s actually enabled rather than assuming.
Frequently Asked Questions
Do I need calibration frames every single session?
Darks can be reused across sessions as long as your exposure length, ISO, and roughly the ambient temperature stay the same — many astrophotographers keep a “dark library” for exactly this reason. Flats, on the other hand, should be recaptured whenever you change focus, rotate the camera, or adjust the optical train, since they’re mapping the exact state of your sensor and optics at that moment.
How many light frames should a beginner aim to stack?
There’s no fixed number, but a reasonable starting goal is 20 to 40 frames of a few minutes each on a bright target like a large nebula or globular cluster, giving you roughly an hour or more of total integration time. Fainter or smaller targets reward even more. The improvement scales with the square root of your frame count, so going from 10 to 40 frames helps a lot more than going from 40 to 60.
Can I stack photos taken on a tripod with no tracker at all?
Yes, within limits. Short, untracked exposures of a few seconds (to avoid star trailing) can still be stacked to improve wide-field shots of the Milky Way or star fields, and it’s a great way to learn the software before investing in a tracker. Faint, small deep-sky targets are much harder this way since your individual exposures are so short, but it’s a legitimate starting point covered in our astrophotography for beginners guide.
Stacking is one of those skills that sounds intimidating until you’ve done it once, and then it becomes just another part of the routine — capture, calibrate, stack, process. If you’re still deciding what to shoot with, our guide to astrophotography cameras and our broader astrophotography hub are good next stops, and if dew, cables, or other small gear is what’s holding your sessions back, check our accessories buying guide for the fixes that actually matter.