resample {signal}

Change the sampling rate of a signal

Description

Resample using bandlimited interpolation.

Usage

resample(x, p, q = 1, d = 5)

Arguments

x

signal to be resampled.

p, q

p/q specifies the factor to resample by.

d

distance.

Details

Note that p and q do not need to be integers since this routine does not use a polyphase rate change algorithm, but instead uses bandlimited interpolation, wherein the continuous time signal is estimated by summing the sinc functions of the nearest neighbouring points up to distance d.

Note that resample computes all samples up to but not including time n+1. If you are increasing the sample rate, this means that it will generate samples beyond the end of the time range of the original signal. That is why xf must go all the way to 10.95 in the example below.

Nowadays, the signal version in Matlab and Octave contain more modern code for resample that has not been ported to the signal R package (yet).

Value

The resampled signal, an array of length ceiling(length(x) * p / q).

Author(s)

Original Octave version by Paul Kienzle [e-mail]. Conversion to R by Tom Short.

References

J. O. Smith and P. Gossett (1984). A flexible sampling-rate conversion method. In ICASSP-84, Volume II, pp. 19.4.1-19.4.2. New York: IEEE Press.

doi:10.1109/ICASSP.1984.1172555

Octave Forge https://octave.sourceforge.io/

See Also

filter, decimate, interp

Examples

xf <- seq(0, 10.95, by=0.05)
yf <- sin(2*pi*xf/5)
xp <- 0:10
yp <- sin(2*pi*xp/5)
r <- resample(yp, xp[2], xf[2])
title("confirm that the resampled function matches the original")
plot(xf, yf, type = "l", col = "blue")
lines(xf, r[1:length(xf)], col = "red")
points(xp,yp, pch = 19, col = "blue")
legend("bottomleft", c("Original", "Resample", "Data"),
       col = c("blue", "red", "blue"),
       pch = c(NA, NA, 19),
       lty = c(1, 1, NA), bty = "n")