Kvantiseringsstøyens

Kvantiseringsstøy, often translated as quantization noise, is an artifact that arises in digital signal processing when an analog signal is converted into a digital one through a process called quantization. Analog signals are continuous, meaning they can take on any value within a given range. However, digital systems represent signals using discrete values, which are limited to a finite set of possible levels.

Quantization involves mapping each continuous analog value to the nearest available discrete digital level. This process inherently introduces an error, as the original analog value is rarely exactly equal to one of the allowed digital levels. This difference between the original analog value and its quantized digital representation is the quantization error. When viewed as a signal, this error is referred to as quantization noise.

The magnitude of quantization noise depends on the number of discrete levels used in the quantization process, which is determined by the bit depth of the analog-to-digital converter (ADC). A higher bit depth allows for more quantization levels, resulting in smaller step sizes between levels and consequently, lower quantization noise. Conversely, a lower bit depth leads to larger step sizes and more significant quantization noise.

Quantization noise is an unavoidable consequence of digitizing analog signals. While it cannot be eliminated, its impact can be minimized by using ADCs with sufficient resolution (bit depth) for the intended application. Dithering, a technique that involves adding a small amount of random noise to the analog signal before quantization, can also be employed to distribute the quantization error more evenly and make it less perceptible.