Deepace KC908 Manual page 42

This menu includes resolution bandwidth, sweeping time, detection
method, trigger method, and other settings.
Resolution bandwidth, also known as analytical bandwidth, is originally
designed to map two signals on the same spectrum and make them
distinguishable no matter how far they are from each other. Moreover, it's
equivalent to the bandwidth of an intermediate-frequency filter on a scanning
spectrometer. In a FFT spectrometer, two adjacent points in fourier series can
naturally distinguish two different frequency, thus RBW(Resolution Bandwidth)
is equivalent to FFT resolution to some extent. However, pure FFT is not
enough. Assume that a signal lands exactly in the middile of that two adjacent
points. This situation makes the reading of both points smaller than actual,
inducing a measurement error. To solve this problem, the data must be put
through a window process. Figuratively speaking, a window process means
summing the weighted total power of several adjacent points and displaying
the data of the central point as a result. This price of this process is the
deterioration of the resolution. The performance of different window types
varies. Taken the fact that it's a custom to adopt gaussian filter into account,
KC908 adopts a gaussian window. In this situation, RBW represents an
equivalent IF filter's 3dB bandwidth.
A simplified model would be that the spectrum line for measuring a
simple-frequency signal should be infinitely narrow. Yet in the real world that