FILTUPDESIGN Function

Designs a filter to be used with FILTUP.

Usage

result = FILTUPDESIGN(m, l[, alpha])

Input Parameters

m—(scalar) The interpolation factor. This should be the same value as that being passed into FILTUP.

l—An integer used to determine the order of the filter. The filter order is given by n = 2lm.

alpha—(optional) A scalar value used to scale the filter cutoff frequency. The normalized cutoff frequency is given by alpha divided by m. (Default: alpha = 1)

Returned Value

result—A filter structure to be used with FILTUP.

Keywords

None.

Discussion

FILTUPDESIGN generates an optimal linear phase interpolator filter. The filter generated has the property that every m-th point of the interpolated signal is equal to a corresponding point in the original sequence.

The filter design technique implemented in FILTUPDESIGN is based on a paper by Oetken, Parks, and Schussler, 1975.

Example

In this example, a signal is interpolated to get a signal four times the length of the original.

m = 4

nx = 100

freq = 0.9

; Create a signal.

x = SIN(!Pi*(freq)*FINDGEN(nx))

alpha = 1.0

; This will cause the resulting filter to be of order 32.

l = 4

; Create interpolation filter and interpolate and filter.

h = FILTUPDESIGN(m, l, alpha)

; To check interpolation filter, decimate interpolated, filtered

; data. Original data set is returned, to within a close epsilon.

result = FILTUP(h, x, m)

newx = FILTDOWN(result, m)

; Difference between interpolation filter data.

PRINT, TOTAL(ABS(newx - x))

; PV-WAVE prints: 2.5938099e-14

See Also

FILTUP

For Additional Information

Oetken, Parks, and Schussler, 1975.