What bit depth counts
Every color on a typical screen is built from three channels: red, green and blue. Bit depth is the number of binary digits used to describe the brightness of each channel, and so it sets how many distinct levels each channel can take. The relationship is a power of two:
- 6 bits per channel gives 2^6 = 64 levels.
- 8 bits per channel gives 2^8 = 256 levels.
- 10 bits per channel gives 2^10 = 1,024 levels.
Because a pixel combines three channels, the total number of colors is the per-channel count cubed. For 8 bits that is 256 × 256 × 256 = 16,777,216 colors, often rounded to “16.7 million.” For 10 bits it is 1,024 × 1,024 × 1,024 = 1,073,741,824, a little over a billion. Sometimes the same quantity is described as “24-bit” or “30-bit” color, which simply adds the three channels together (3 × 8 and 3 × 10).
More colors in total is the headline, but the practical effect is about the size of the steps between neighboring levels.
Why gradients show bands
A smooth gradient, such as a clear sky fading from light to deeper blue, asks the display to change brightness gradually across many pixels. If only a few levels exist between the two ends, the gradient has to be built from a few flat regions. Each region is a single level, and the jump to the next one can be visible as a stripe. That stripe is banding.
A simple hypothetical example makes the scale clear. Suppose a gray gradient runs from black to white across 1,920 pixels of width, and the display has 256 gray levels. If every level is used once, each level covers 1,920 ÷ 256 = 7.5 pixels. The steps are narrow, and the step in brightness between them is small, but on a large smooth area they can still be noticed. With 1,024 levels, each covers 1,920 ÷ 1,024 = 1.875 pixels, so the steps are both narrower and gentler.
Real images are rarely that tidy. A gradient that covers only a small part of the full range, such as a dim vignette that moves across only a few dozen levels, produces much wider bands, since there are fewer levels to share across the same distance.
Where banding shows up most
Banding is most visible in:
- Dark scenes, where the eye is sensitive to small brightness changes and the encoded steps are often coarser in relative terms.
- Large areas of slowly changing color: skies, studio backdrops, fog, smoke, soft lighting on a wall.
- Computer-generated images and graphics with smooth shading.
- Heavily compressed video, where the compression process itself can flatten subtle gradients before the display ever sees them.
Natural photographs with film grain or sensor noise often hide banding, because the noise breaks up the sharp edges between levels.
Dithering and frame rate control
Some panels natively support fewer levels than their spec sheet suggests. A common technique is to alternate rapidly between two neighboring levels over successive frames, so the eye averages them to something in between. This is often called frame rate control, and a panel that uses it is sometimes described as “8-bit plus FRC” or “10-bit (8-bit plus FRC).” A similar idea, called dithering, spreads the difference across neighboring pixels in a fine pattern rather than across time.
These techniques can work well. They add intermediate steps, soften banding and cost nothing visible in most content. They are not identical to native depth, though: at close inspection or in very smooth, dark gradients, a faint shimmer or texture may be seen. The honest reading of a spec sheet is that native 10-bit panels and 8-bit-plus-FRC panels differ in how they get there, and which matters depends on what you do. Where a listing does not say, it is worth looking for the detail rather than assuming.
The chain matters, not just the panel
Smooth gradients need every stage to support enough levels. A 10-bit panel shows 8-bit content with 8-bit steps unless the content, the operating system, the graphics source, the application and the cable all carry more. A few things to check:
- The content. Many video files and images are 8-bit. Higher depth content, often used with high dynamic range, spreads the brightness range over more steps to avoid visible bands.
- The connection. Display connections and video connectors differ in how much bandwidth they have. Higher resolutions and refresh rates may force a lower bit depth or reduced color sampling, depending on the link.
- Software settings. The output depth in the graphics settings may default to 8 bits.
- Processing. Color adjustments applied in 8-bit can create gaps in the histogram, causing banding that was not in the original. This is one reason editing displays and professional workflows often use higher-depth pipelines.
Calibration can also matter. Adjusting a display’s tone curve in a limited-precision pipeline can drop levels, a point that comes up in discussions of gamma curve adjustments and professional calibration.
How to check for banding yourself
Test patterns containing smooth gray ramps and color ramps are a simple way to look. View a black-to-white ramp at normal viewing distance in a dim room and see whether you can count distinct stripes or whether it looks continuous. Using test patterns this way is a general-purpose check, not a measurement. Keep in mind that a visible step might come from the file, the graphics output or the panel itself, so changing one thing at a time can show where it originates.
Does more bit depth always look better?
Not always. For ordinary office work, web pages and most everyday video, 8 bits per channel is usually adequate, and many people never notice banding. More depth tends to matter for color-critical editing, for scenes with large smooth gradients, and for high dynamic range material that stretches the tone range more widely. Even then, the benefit depends on the whole chain, and a higher number on a spec sheet is not a guarantee of smoother gradients on its own.