Binning is combining the signal from a small block of neighboring pixels on an image sensor into one larger pixel. In 2×2 binning, four pixels become one, so the image has a quarter of the pixel count but each new pixel holds four times the light, which means less visible noise. Binning trades resolution for a cleaner signal per pixel. It is used in astrophotography and scientific cameras, in many phone cameras (where it is called pixel binning), and inside some cameras when they record video.
How Binning Works
Each pixel on a sensor collects light during the exposure and turns it into an electrical signal. Binning adds the signals of a square group of pixels together and reports the total as a single value. The binning level names the size of the group:
- 1×1: no binning. Every pixel is read on its own.
- 2×2: four pixels combined into one. Width and height each halve.
- 3×3: nine pixels combined into one. Width and height each drop to a third.
- 4×4: sixteen pixels combined into one. Width and height each drop to a quarter.
The total amount of light reaching the sensor does not change. What changes is how it is divided up. Fewer, bigger pixels each hold more signal, so random variation between them (noise) is smaller compared with the signal. That ratio is the signal-to-noise ratio, and raising it is the whole point of binning.
What Binning Does to Resolution and Noise
The table uses a 24 megapixel sensor (6000 × 4000 pixels) as the example. Noise in a well-lit image is mostly shot noise, the natural randomness in how photons arrive, and it grows with the square root of the signal. Combining N pixels multiplies the signal by N and the shot noise by the square root of N, so the signal-to-noise ratio improves by the square root of N.
| Binning | Pixels combined | Output size | Megapixels | Signal-to-noise gain (shot noise) | Best case with on-chip charge binning (read noise) |
|---|---|---|---|---|---|
| 1×1 | 1 | 6000 × 4000 | 24 | 1× | 1× |
| 2×2 | 4 | 3000 × 2000 | 6 | 2× | up to 4× |
| 3×3 | 9 | 2000 × 1333 | 2.7 | 3× | up to 9× |
| 4×4 | 16 | 1500 × 1000 | 1.5 | 4× | up to 16× |
The last column applies only to very faint signals, where the noise added by reading out the sensor (read noise) dominates. If the charge from the group is summed on the chip before it is read, the group pays the read noise once instead of N times, and the gain can approach N. That is why binning was so valuable on the scientific and astronomy cameras that first used it. In practice the real gain falls somewhere between the two columns.
Hardware Binning and Software Binning
There are two ways to bin, and the difference matters.
- Hardware (charge) binning sums the electrical charge of the pixels on the sensor itself, before readout. CCD sensors do this naturally because charge is shifted across the chip to a single output, and grouped charge can be read as one packet. Read noise is incurred once per binned pixel, so this method gives the largest benefit in very low light. Some CMOS sensor designs can also combine charge or signal on the chip.
- Software (digital) binning reads every pixel separately and then adds or averages them, either in the camera or later on a computer. Each pixel still carries its own read noise, so the gain in very faint light is smaller. For ordinary exposures, where shot noise dominates, the result is close to hardware binning. Software binning is essentially the same as reducing an image to a smaller size by averaging, which is why a downsized photo looks cleaner than a full-size crop of the same file.
Most CMOS cameras bin digitally. Hardware binning on a color sensor is also limited by the Bayer filter: neighboring pixels sit under different color filters, so adding them blindly would mix red, green, and blue into one value and lose the color. Color binning has to combine pixels of the same color.
Not sure what read noise and shot noise look like in a picture? Noise explains the kinds of noise and where each one comes from.
Pixel Binning in Phone Cameras
Many phone sensors with very high megapixel counts use pixel binning by default. Their color filter is arranged so that each color covers a 2×2 group of pixels (or a 3×3 or 4×4 group on some sensors) instead of single pixels. Combining each same-color group produces a normal Bayer pattern at a quarter of the pixel count, so a 48 megapixel sensor delivers 12 megapixel photos with larger effective pixels.
The tiny pixels on a phone sensor collect very little light each, so binning is how these phones make cleaner images in dim rooms and at night. In bright light, many phones offer a full-resolution mode that reads every pixel. It can capture more detail of a still subject in good light, but it produces larger files and usually looks noisier in dim light. For most shooting, the binned default is the better choice.
Wondering when the full-resolution mode on a phone is worth turning on? Phone camera settings walks through the settings that matter most.
Binning in Astrophotography
Astronomy cameras usually let you choose a binning level, and there are three common reasons to use it.
- Matching pixels to the sky. The atmosphere blurs stars to a disk of a certain size, called the seeing. If your pixels are much smaller than that blur on the sky, extra resolution records only blur. Image scale in arcseconds per pixel is 206.265 × pixel size in microns ÷ focal length in millimeters. A 3.76 micron pixel at 1000 mm gives about 0.78 arcseconds per pixel; binned 2×2 it becomes about 1.55, which often suits ordinary seeing better.
- Faint targets. On a dim nebula, binned frames reach a usable signal in less time, which helps with short sessions or when framing and focusing.
- Color with monochrome cameras. A long-standing method shoots the detailed luminance frames at 1×1 and the red, green, and blue frames binned 2×2. The eye takes detail mostly from brightness, so the lower-resolution color still looks sharp once combined.
Binning also shrinks file sizes and speeds up downloads, which helps when you capture hundreds of frames for stacking.
Binning in Video
A video frame needs far fewer pixels than a high-resolution sensor provides, and the sensor has to be read many times per second. Cameras reduce the data in one of three ways. Reading every pixel and scaling down (oversampling) gives the sharpest, cleanest footage. Binning combines neighbors and is a reasonable middle ground. Line skipping simply ignores rows and columns, which is fastest but tends to produce aliasing and moire on fine patterns. If a camera’s video looks jagged on brick walls or fabric, the readout method is often the reason.
Pros and Cons of Binning
| Advantages | Disadvantages |
|---|---|
| Higher signal-to-noise ratio per pixel | Lower resolution: a quarter of the pixels at 2×2 |
| Shorter exposures reach a usable signal | Fine detail is lost and cannot be recovered later |
| Smaller files and faster readout | Summed signal can clip sooner on bright stars or highlights |
| Better match to atmospheric seeing at long focal lengths | Little real gain over resizing when the camera bins digitally |
How to Use Binning
- Check whether your camera offers it. Astronomy and scientific cameras list binning in their capture software. Phones bin automatically; the choice is usually a full-resolution or high megapixel mode you can switch on. Most interchangeable-lens cameras do not offer binning for stills.
- Start at 2×2. It gives the biggest improvement for the resolution you give up. Go to 3×3 or higher only for very faint targets or very long focal lengths.
- Recheck exposure. When the camera sums the group rather than averaging it, each binned pixel collects more signal, so the same exposure looks brighter and may clip highlights or bright stars.
- Keep calibration frames matched. In astrophotography, dark and flat frames must be taken at the same binning level as the light frames.
- If your camera has no binning, bin in software. Averaging the image down to a smaller size in an editor gives most of the benefit for ordinary photographs.
Common Mistakes
- Expecting free quality. Binning does not add light. It redistributes it, and you pay in detail.
- Binning when you need detail. For a big print of a still subject in good light, full resolution is the better choice.
- Confusing binning with raising ISO. ISO amplifies the signal you have. Binning collects more signal per output pixel. They solve different problems.
- Mixing binning levels in a stack. Frames binned differently do not line up and calibrate correctly together.
Try This
Take one photo of a dim room at a high ISO. Open it in an editor and view a section at 100 percent. Then make a copy reduced to half the width and half the height using an averaging resize, and view that at 100 percent. The smaller copy is a software 2×2 bin: it has a quarter of the pixels and visibly less noise. Compare the two and decide which you would rather print at a small size.
Binning Questions
What is sensor binning?
Sensor binning combines the signal of neighboring pixels into one output pixel, either on the sensor before readout or digitally afterward. It lowers resolution and raises the signal-to-noise ratio of each pixel.
What is pixel binning on a phone?
It is the same idea built into high megapixel phone sensors, which group same-color pixels in 2×2 or larger blocks. A 48 megapixel phone sensor typically delivers 12 megapixel photos this way, with less noise than reading each tiny pixel on its own.
Is image binning the same as resizing?
Software binning is very close to resizing an image down by averaging pixels. Hardware binning on the sensor can do better in very faint light because the read noise is added only once per group.
Does binning increase sensitivity?
Each binned pixel collects more light, so it reaches a given signal faster, and that is often described as higher sensitivity. The sensor as a whole collects the same light as before; binning changes how that light is shared out among pixels.
Want to know how pixel size and pixel count trade off in a camera? Pixel pitch explains why larger pixels gather more light.