Menu
Aliasing is a digital technology effect that occurs in motion pictures, audio, or computer graphics. It occurs when analog signals are digitized if the sampling is not correct. The frequency components are interpreted incorrectly – for example in image scenes or in the audio range when overtones are read out as low frequencies. Instead of being in the correct form, these are aliased and result in incorrect waveforms compared to the original signal.
Aliasing is a type of distortion found in software synthesizing, computer graphics, motion pictures, audio signal transmission, digital synthesizers, or signal processors. The frequency signal is exceeded within the signal processing or measured at an insufficient sampling rate. This causes disruptions in the transmission.
Such an effect occurs in film when very fast scenes take place, for example a rotating wheel, which then appears to run backwards in the perception. There is an illusion for the eyes because the rotation frequency of the camera is not sufficient to capture the rapid movement of the rotation. A similar thing happens with a very elaborate pattern on an actor’s clothing. A striped shirt looks like a psychedelic trip because the camera doesn’t capture the actually higher frequencies due to a limited pixel count.
The aliasing effect can be avoided with modern editing methods. This is possible by multiplying the sampling frequency by an integer number. The associated procedure is called oversampling. It provides a stratified sample that has the same ratio in each variable as the overall intake when tested. Certain groups are represented by oversampling. The higher sampling rate of the signal compensates for the aliasing effect.
The disruptive effect at the edges of objects can be seen using computer graphics as an example. The graphics consist of limited pixel grids and a certain image resolution. If two- and three-dimensional surfaces are created, the aliasing effect occurs due to missing edge smoothing. It is therefore necessary to scan the image content and render in real time so that figures do not appear angular or pixelated. When creating the grid areas in the graphic, each individual pixel is assigned the appropriate color.
This is comparable to signal processing in audio or film frequencies. Each abstract image description must be algorithmically calculated and evaluated in individual pixels. If this is not done, the aliasing occurs as a pixel image with distorted edges. Whether graphics or audio: the Nyquist frequency is decisive. It indicates whether the components are correctly sampled at the correct signal frequency to be digitized. If they are above the output frequency, interference occurs.
For digital audio, aliasing is a major problem and an undesirable effect. If frequencies change the signal, interference occurs that can be avoided by using special components in the conversion process. The components are filters that allow frequencies above the Nyquist frequency to be converted. At the same time, they prevent signals from being changed in frequency at the sampling frequency and during conversion.
