Welcome to GFFT
------- -- ----
This is the second public release of GFFT, an FFT-based
spectrum analysis program with many features. Improvements
include:
-- an AmigaGuide discussion of Spectrum Analysis with
many pictures!
-- Support for different screens and (system default)
fonts
-- progress indicator with stop button
-- support for GFFT commands in the icon tooltypes
-- bugfixes and other small improvements
If you would like to get the binary optimized for machines
with FPU, please send $25. For best results, please specify the
exact CPU and FPU (e.g. 25 MHz 68030 with 25 MHz 68882) you have.
Address: Charles P. Peterson
4007 Enchanted Sun
San Antonio, TX 78244-1254 USA
E-mail: Charles_P_Peterson@fcircus.sat.tx.us
cpp@crl.com
charlesp@darwin.sfbr.org
Please help me support this program by sending a donation of
any amount. I will acknowledge all donations of $10 US or more
with a short thank-you note and instructions as to how to disable
the donation reminders. For a donation of $25 US, I will also send
you a binary of the current release optimized for your system (with
donation reminders disabled), AND advise you of the next release,
which will be sent for the cost of shipping (which may be prepaid;
allow for the weight of one floppy disk and packing materials).
The ideas of contributors will be given extra weight in planning
future releases.
Even if you cannot afford to make a donation now, please
let me know if you find the program useful, or how I might make
it useful to you.
Unique among programs available to the personal computer user,
the design of GFFT emphasizes the ability to produce spectrum
analyses of very high precision and accuracy. The tradeoff is
speed; don't expect anything like real-time performance.* GFFT was
never intended to be a real-time analyzer. My experience with most
consumer real-time analyzers is that they are little better than
toys, like a light show to watch while you are listening to music,
and to really appreciate music, it's best not to watch such things
anyway. By operating from stored sample data files, GFFT offers
higher resolution and/or accuracy than most real-time analyzers.
There is no limit to the length of samples that GFFT can analyze
and average, and the longer the sample that is averaged, the more
accurate the results can be. Without averaging over a significant
period of time, a spectrum analysis is not much better than a list
of random numbers.
(*However, with suitably chosen parameters, GFFT can also give
more approximate results in a few seconds. But why can't you wait
a little longerfor something a little less random? See the
SpectrumAnalysis.guide for a full discussion.)
GFFT accepts 8, 12, or 16-bit data in in IFF 8SVX, AIFF, and
AVR formats, or in unformatted files, and can accept data points
entered through a console in floating point. GFFT can be operated
from either a Workbench GUI or from a CLI in interactive or batch
modes. It has built-in help facilities for all operating modes.
Before running GFFT on your Amiga system, read the file named
INSTALLATION. If you experience problems while running GFFT, see
the file named Bugs. If you would like to read about how to use
GFFT, you can read the file named GFFT.HELP, which is also used by
GFFT to display help messages. After installing GFFT according to
the instructions, you can get help while running GFFT either from
the Workbench or from the CLI.
If you want further information about the GNU General Public
License which applies to GFFT and the specific rights it gives you
to modify and distribute GFFT, or the no-warranty disclaimer which
applies to it, refer to the file named COPYING.
Though GFFT is copyrighted under a General Public License and
the author respects and agrees with the ideas of the Free Software
Foundation, GFFT is not currently part of project GNU. So please
direct any inquiries about the program itself to the author, and
not the Free Software Foundation.
The most important files in this archive are provided with
icons so you can use them from the Workbench. Two files not
provided with icons are the the Product-Info and History files.
GFFT README TABLE OF CONTENTS
Preface (the long part you just read)
1. What can I do with GFFT?
2. What features does GFFT have?
3. What are my rights to use, modify, and distribute GFFT?
4. The trademark and copyright statements
1. What can I do with GFFT?
With GFFT you can explore the world of sound in a new domain--
the frequency domain.
With GFFT you can "see" what frequencies are present in a
particular sound, and how they vary over time. This might be
useful if you are involved in some kind of research, or if you want
to understand a sound better so that you can synthesize a
simulation of it.
With GFFT, an audio sampler, and a random noise generator
(preferrably a pink noise generator), you can analyze the frequency
response of electronic audio equipment. Add a calibrated
microphone, and you can also analyze loudspeakers and room
acoustics. This is where GFFT really shines. Few (if any)
analyzers are able to achieve the fine resolution or accuracy that
GFFT can achieve, or compensate for the frequency response of
ancillary equipment as flexibly as GFFT, or produce hardcopy which
looks as nice.
2. What are the features of GFFT?
GFFT is an FFT-based spectrum analysis program with many
features.
GFFT can produce high quality spectrum plots on screen,
plotter, or printer, or save to Postscript or TeX files, or output
spectral data in text form to a console or file. (GNUPLOT, a
separate program by other authors, is required for plotting
features. GFFT is designed to invoke and control GNUPLOT
automatically.) It can plot in 2D (Amplitude vs Frequency), or 3D
(Amplitude vs Frequency vs Time). It allows the selection of
logarithmic X and/or Y axes, the dB scaling of X, and arbitrary 3D
rotation. GFFT can show multiple spectra in the same plot, or
append spectral data bands from earlier sessions.
GFFT provides 8 selectable FFT window types, including
Blackman-Harris 92 dB. It can apply a special weighting for pink
noise testing. Smoothing, calibration, ranging, and quantization
are also available. There are no arbitrary limits to the number of
frequency bins or smoothing points, or to the length of data which
can be analyzed.
GFFT is intended to be fairly universal in nature.
Furthermore, since it is provided under a GNU General Public
License and distributed with source code, a user with an
understanding of programming can adapt it to more specialized
requirements. (I may also be able to help; see below.)
Please be aware that GFFT (and all the code provided with
GFFT) is not intended for use in applications where error or
inaccuracy could endanger life or property. There is ABSOLUTELY NO
WARRANTY, express or implied, concerning the performance or
accuracy of GFFT. The general no-warranty disclaimer is given in
the file named COPYING which should be included with this
distribution.
3. What are my rights to use, modify, and distribute GFFT?
This program is 'free' software; you can redistribute it and/or
modify it under the terms of the GNU General Public License as
published by the Free Software Foundation; either version 2 of the
License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
General Public License for more details.
You should have received a copy of the GNU General Public
License along with this program (in a file named 'COPYING'); if
not, write to the Free Software Foundation, Inc., 675 Mass Ave,
Cambridge, MA 02139, USA. Please do not contact them for any
other information about this program; they do not necessarily
know anything about it.
4. The trademark and copyright statements
GFFT, GFFT.HELP, and related contents in this distribution are
Copyright (C) 1994, 1996 Charles P. Peterson.
'Workbench', 'Amiga', and 'AmigaGuide' are registered trademarks of
Amiga Technologies. GNU is a name used by the Free Software
Foundation, Inc. GNUPLOT is copyright by Thomas Williams and Colin
Kelley. POST is copyright by Adrian Aylward. All other trademarks
or copyrights mentioned here or in code or supporting documentation
are property of their respective owners.
