Oversampling

Part of speech: noun

Definitions

  1. The process of increasing the sampling rate of a signal beyond its original rate to improve resolution or aid in digital processing ; collecting more data points or samples in a survey or dataset than the minimum necessary to enhance accuracy or representation ; a technique in statistics and signal processing where additional data is obtained or measurements are made at a higher frequency to improve quality and reduce errors
  2. An approach of capturing data or signal points at a rate higher than the standard or required rate for better detail and less distortion ; gathering extra examples in data analysis or experimental design to improve reliability or balance ; digitally increasing the number of samples in a signal to facilitate filtering or better reconstruction
  3. The method of obtaining more frequent or numerous observations of a dataset or signal than conventionally needed to refine analysis outcomes ; collecting data points at a rate exceeding normal frequency in fields like audio or image processing to enhance fidelity ; a technique involving recording input signals at higher than Nyquist rate to prevent aliasing and improve signal quality

Etymology: Oversampling is a technical coinage built from two existing English parts: the prefix "over-", meaning excessive or above, and "sample". The "sample" half has a longer history than its modern signal-processing sense suggests. It came into English through Old French "essample" and Middle English "saumple", ultimately descending from Latin "exemplum", an example, which is itself linked to "eximere", to take out. A sample was originally a single example taken from a larger whole, a sense that survives in "a sample of cloth" and in "an example of good conduct". Oversampling takes that idea of a selected instance and applies it at a higher rate than strictly required. In digital signal work, a signal is measured more often than the minimum needed to capture it faithfully, and the extra measurements are then used to improve precision or to simplify later filtering. The meaning therefore moved from taking a representative piece to taking more pieces than the situation demands.

Synonyms: overcollection, overselection, excessive sampling

Antonyms: undersampling, insufficient sampling