Chapter 2:
Signal-to-Noise Ratio
Author(s): Vidi Saptari
Published: 2003
DOI: 10.1117/3.523499.ch2
To quantify the instrument's performance, the spectral signal-to-noise ratio (SNR) is used as the main measure throughout this book. The term has been used somewhat inconsistently; in some cases, it is used to quantify spectral repeatability, and in others, it is used to quantify spectral accuracy. Thus, it is appropriate to start with a clear definition of the term as it applies throughout this book. In this book, the SNR measures the instrument's ability to reproduce the spectrum from the same sample, the same conditions, and the same instrumental configurations over a certain amount of time. This, in fact, is a measure of spectral repeatability, which measures the ability of the instrument to detect certain changes in the spectrum such as those caused by changes in the sample's spectral characteristics. Therefore, noise is the measure of the spectral deviations between measurements, regardless of the output spectrum's proximity to the "€œtrue"€ value. Accuracy, on the other hand, is a measure of the discrepancy between the actual measured value and the “true” standard or calibrated value. Wavelength accuracy of a spectrometer is crucial. It is important for the instrument to be able to conform to the "€œcalibrated standard" in producing wavelength information within the instrument's intended resolution. For example, if the true absorption peak of molecule x is 4000.2 nm, the instrument should be able to give the correct peak wavelength information within its designed resolution. Thus, following the example, if the instrument has a designed resolution of 1 nm, then it should register the peak of molecule x at 4000 nm. This is required for proper information transfer between instruments.
Online access to SPIE eBooks is limited to subscribing institutions.

Signal to noise ratio





Fourier transforms


Back to Top