FFTPLOT

Section: User Commands (1)
Updated: July 2012
Index Return to Main Contents

NAME

fftplot - Fast Fourier Transform and plot.

SYNOPSIS

This reads in a data-file containing one or more columns of int/float data. The FFT is calculated.
Data is plotted: raw,amplitude,power-spectrum. The strongest frequency is identified.

DESCRIPTION

fftplot:  plot a data-file, and its frequency spectrum by FFT. Show summary information.

USAGE:    fftplot.py  [OPTS]  infile.dat

OPTIONS:  -f, --samplefreq  FREQ        define the sampling frequency in Hz. [default: 1.0].
           -c, --column      N,sum,all   use Nth column's data (N is 0-based). [default: all].

           -a, --amplitude               fft y-axis: amplitude. [default: power]
           -p, --power                   fft y-axis: power spectrum (redundant). [default: power]
           -d, --decibels                fft y-axis: power in dB. [default: linear power]
           -l, --logfreq                 fft x-axis: frequency on log-scale. [default: linear]

           -s, --squelch_lf  FREQ        exclude frequencies <= FREQ from summary and plot. 0 hides DC.
           -z, --zoom                    zoom x-axis to "area of interest" (10x around modal point).

           -t, --title       TITLE       override plot title
           -x, --xlabel      LABEL       override x-axis label
           -y, --xlabel      LABEL       override y-axis label
           -u, --ulabel      L,A,BE,LS   override legend labels.

           -r, --plotraw                 also plot the raw input file. [default: just plot FFT].
           -o, --outplot     FILE.png    save main plot: FILE.png/.jpg, even if -n, "auto" names for infile.

           -n, --noplot                  don't draw the plot window on screen (useful for testing/scripting).
           -h, --help                    show help.

INPUT:    Input file, infile.dat ('-' for stdin) should be columns of (int/float) data, separated by
           whitespace, with optional #comments, i.e. the standard form for numpy.loadfile(). cf. genffttest.
           (compressed files (.gz/.bz2) can also be read transparently, no need to uncompress first.)

NOTES:    * The FFT is normalised, such that a signal x = A.cos(2.PI.f.t) + B.sin(2.PI.f.t),
           the power-spectrum has a point at (A^2 + B^2), and frequency f. But if f is unaligned,
           i.e.  f != n*(samplerate/num_samples), then its power will be spread over several bins.
           * This does no apodisation: the window is effectively rectangular.
           * Use squelch_lf when the chart is dominated by dc/low-freq. (Can still pan to see LF).
           * When zooming automatically (-z), the DC level (f=0) is never considered "interesting".

AUTHOR

The author of fftplot and this manual page is Richard Neill, <maths#AT(spamblock)#richardneill.org>
Copyright July 2012; this is Free Software (GPL v3+), see the source for copying conditions.