Cheap Spectral Analyzer

[tontoko hirorin]

Spectral Anlyzer is useful instrument, but generally a bit expensive,
isn't it? The spectral analyzer FerFT is cheap as from $29.85. The
frequency range is 0-20kHz. For details, please visit:

http://139.134.5.123/tiddler2/c3508/ad/ferftad.htm

 

[Don Pearce]

Spectral Anlyzer is useful instrument, but generally a bit expensive,
isn't it? The spectral analyzer FerFT is cheap as from $29.85. The
frequency range is 0-20kHz. For details, please visit:

http://139.134.5.123/tiddler2/c3508/ad/ferftad.htm

Too little, too late and too dear. There is plenty of free stuff out
there that does exactly this - frequently thrown in as an extra in
audio products.

d
Pearce Consulting
http://www.pearce.uk.com

 

[Tim Auton]

Spectral Anlyzer is useful instrument, but generally a bit expensive,
isn't it? The spectral analyzer FerFT is cheap as from $29.85. The
frequency range is 0-20kHz. For details, please visit:

http://139.134.5.123/tiddler2/c3508/ad/ferftad.htm

Why does it want my email address? I think not...


Tim

 

[Rene Tschaggelar]

Spectral Anlyzer is useful instrument, but generally a bit expensive,
isn't it? The spectral analyzer FerFT is cheap as from $29.85. The
frequency range is 0-20kHz. For details, please visit:

http://139.134.5.123/tiddler2/c3508/ad/ferftad.htm

Verifying the email is bull****. Either one has a web page to display
something to the public or not.

Rene

 

[Alan]

Why does it want my email address? I think not...


Tim

And insists that you turn on Javascript!

Alan


++++++++++++++++++++++++++++++++++++++++++
Jenal Communications
Manufacturers and Suppliers of HF Selcall
P O Box 1108, Morley, WA, 6943
Tel: +61 8 9370 5533 Fax +61 8 9467 6146
Web Site: http://www.jenal.com
e-mail: http://www.jenal.com/?p=1
++++++++++++++++++++++++++++++++++++++++++

 

[Guy Macon]

Spectral Anlyzer is useful instrument, but generally a bit expensive,
isn't it? The spectral analyzer FerFT is cheap as from $29.85. The
frequency range is 0-20kHz. For details, please visit:

(snip)

From [ http://linux-sound.org/dsp.html ]:

FFT/Phase Vocoder:

Enveloper
draw FFT components and amplitude waveforms with this Java-based
software from Abram Hindle

FreqTweak
Jesse Chappell's very neat "tool for FFT-based realtime audio
spectral manipulation and display"

GWave
Analyzer visualizes/analyzes WAV files via FFT, also extracts
melody to MIDI file

Kiss FFT
very small FFT library, can use either fixed or floating point
data types

NoTAM
FFT/phase vocoder packages with X11 GUI now available for Linux:

Bio
multiple graphic representations of soundfile

Ceres
fascinating spectral editor

Ceres2
Johnathan Lee's wonderfully enhanced version

Ceres2w
WAV and glibc support for Ceres2, from Reine Jonsson

Ceres3
a superb version from Stanko Juzbasic

Mammut
treats entire sound as single spectral analysis window

NoTAMProgs
the famous Institute has prepared its own sources for Linux

Sculptor
a very nice suite of realtime FFT tools

XMANA
data display program with FFT and other transforms, uses
XMAD data acquisition files


From [ http://linux-sound.org/scopes.html ]:

Alsa-Nyser
cool scope from Raphael Francois

Cthugha
"an oscilloscope on acid..."

DireShow
frequency analyzer with a nice SVGA interface

Freq51
realtime spectrum analyzer, and a patch which indicates
frequencies as musical note-names

JXSpectrum and XSpectrum
two more frequency analyzers

Sonogram
very nice audio visualization package from Christoph Lauer,
requires Java2

SoundMan
another excellent audio visualization app

Spectrogram
samples data from soundcard, displays FFT spectrogram in realtime

Waterfall
realtime spectrum analyzer for XMMS

WMscope
audio display applet for Window Maker or Afterstep

XPortShot RT
Gtk-based oscilloscope simulator for Linux

baudline
excellent realtime frequency analyzer with varied displays

eXtace
a much-improved version of this scope/visualizer

kanalyzer
KDE applet to display the spectrum or the time-domain signal
from a soundcard

xanalyser
realtime frequency spectrum analyser for X, from Arvin Schnell

xsr
oscilloscope and spectrum analyzer, output to root window


From [ http://linux-sound.org/dsp.html ]:

General Analysis/Processing:

Aglaophone
"...a system of interconnectable modules for the recording,
processing, and playback of real-time audio"

ATSH
very cool analysis/resynthesis program from Prof. Oscar Pablo Di Liscia

Dataplore
"a software tool designed for the analysis of signals and time series
data of any kind"

HASAS
HydroAcoustic Signal Analysis System from Jussi Laako

InSpect
sound analysis/resynthesis/transformation from Sylvain Marchand

LastWave
wavelet DSP software from Emmanuel Bacry

Mustajuuri
audio signal processing software with GUI and support for plug-ins

Scilab
a large package of mathematics & signal processing functions
with excellent graphic displays

Signal Applications To Audio Networks (SATAN)
"a program that is used for various audio related signal processing tasks"

SNDAN
signal analysis, graphics, modification, and synthesis

Sound Ruler
acoustic analysis software from Marcos Gridi Papp

SSPLab AudioAnalyser (SAA)
"... a tool for analyzing audio signals in terms of its content"

Yorick
another mathematics package with signal processing and data
visualization tools

dsplab Xforms-based DSP visualization and manipulation software

Convolution:

Almus VCU
"... makes a computer equipped with a multi-channel sound card
into a real-time versatile convolver unit"

 

[Jan Panteltje]

**** Post for FREE via your newsreader at post.usenet.com ****

Spectral Anlyzer is useful instrument, but generally a bit expensive,
isn't it? The spectral analyzer FerFT is cheap as from $29.85. The
frequency range is 0-20kHz. For details, please visit:

(snip)

From [ http://linux-sound.org/dsp.html ]:

"... makes a computer equipped with a multi-channel sound card
into a real-time versatile convolver unit"

The fastest FFT library I know about is fftw, http://www.fftw.org/
It is rather simple to integrate in a C program.
Linux of cause, and free.
JP

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
*** Usenet.com - The #1 Usenet Newsgroup Service on The Planet! ***
http://www.usenet.com
Unlimited Download - 19 Seperate Servers - 90,000 groups - Uncensored
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

 

[Tim Auton]

The fastest FFT library I know about is fftw, http://www.fftw.org/
It is rather simple to integrate in a C program.
Linux of cause, and free.

Linux my arse. It's vanilla ANSI C.


Tim

 

[Jan Panteltje]

**** Post for FREE via your newsreader at post.usenet.com ****

Linux my arse. It's vanilla ANSI C.

And nothing more suitable to use that with then L I N U X
With or without vanilla in your arse.

Tim

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
*** Usenet.com - The #1 Usenet Newsgroup Service on The Planet! ***
http://www.usenet.com
Unlimited Download - 19 Seperate Servers - 90,000 groups - Uncensored
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

 

[tontoko hirorin]

Spectral Anlyzer is useful instrument, but generally a bit expensive,
isn't it? The spectral analyzer FerFT is cheap as from $29.85. The
frequency range is 0-20kHz. For details, please visit:

http://139.134.5.123/tiddler2/c3508/ad/ferftad.htm

Theoretically the calculation order of FFT is O(n*log(n)) where n is
the number of data, while the method used in the above software has
the calcuration order: O(n). Practically when n=1048576, the
calculation load of FFT is six times heavier than the method.
For details of that method for Fourier transformation, please visit:

http://139.134.5.123/tiddler2/c3508/reftr.htm

 

[Rene Tschaggelar]

Theoretically the calculation order of FFT is O(n*log(n)) where n is
the number of data, while the method used in the above software has
the calcuration order: O(n). Practically when n=1048576, the
calculation load of FFT is six times heavier than the method.
For details of that method for Fourier transformation, please visit:

http://139.134.5.123/tiddler2/c3508/reftr.htm

tontoko,
your spectrum analyzer is clearly targetted at Audio frequencies.
Since the page is not really accessible to normal users, I didn't
have a closer look.
The number of ADC bits unknown, but I assume 16 bit from the sound
card. The dynamic range should be stated some where.
Nevertheless, everything above 4k points are useless since the audio
range starts at 20Hz. It does not extend to DC. Nor does the soundcard.
I guess the sub audio band is only used by very few specialists in
earth quacke research and such.

Rene

 

[Allan Herriman]

Nevertheless, everything above 4k points are useless since the audio
range starts at 20Hz. It does not extend to DC. Nor does the soundcard.
I guess the sub audio band is only used by very few specialists ...

How do you come to that conclusion? More points = more frequency
resolution. Frequency resolution is not the same thing as lower
frequency limit.

E.g. 48kHz sampling, 4096 point FFT. Frequency bin width is ~12Hz.

48kHz sampling, 1M point FFT. Frequency bin width is ~0.046Hz.

This would allow you to distinguish 1.001kHz from 1.000kHz, something
that can't be done easily with a 4k point FFT.

Regards,
Allan.

 

[Rene Tschaggelar]

How do you come to that conclusion? More points = more frequency
resolution. Frequency resolution is not the same thing as lower
frequency limit.

E.g. 48kHz sampling, 4096 point FFT. Frequency bin width is ~12Hz.

48kHz sampling, 1M point FFT. Frequency bin width is ~0.046Hz.

This would allow you to distinguish 1.001kHz from 1.000kHz, something
that can't be done easily with a 4k point FFT.

Yes, you're correct.
Just to get an idea, what can it be used for in this
specific example ?

I had a look at the paper and found the method intriguing.
Unfortunately, the frequency band of the product with 20kHz
only is toy stuff.

Rene

 

[Fred Bloggs]

Theoretically the calculation order of FFT is O(n*log(n)) where n is
the number of data, while the method used in the above software has
the calcuration order: O(n). Practically when n=1048576, the
calculation load of FFT is six times heavier than the method.
For details of that method for Fourier transformation, please visit:

http://139.134.5.123/tiddler2/c3508/reftr.htm

There is something wrong with your reasoning....

 

[Guy Macon]

I guess the sub audio band is only used by very few specialists in
earth quacke research and such.

Your geuss is wrong. It is used in motion control and temperature
servo systems.

 

[tontoko hirorin]

tontoko,
your spectrum analyzer is clearly targetted at Audio frequencies.
Since the page is not really accessible to normal users, I didn't
have a closer look.
The number of ADC bits unknown, but I assume 16 bit from the sound
card. The dynamic range should be stated some where.
Nevertheless, everything above 4k points are useless since the audio
range starts at 20Hz. It does not extend to DC. Nor does the soundcard.
I guess the sub audio band is only used by very few specialists in
earth quacke research and such.

Rene

You've written as,

...since the audio
range starts at 20Hz.

I agree. And the high-end of hardware equipped with that software is
up to 20kHz. So it might not be attractive for analyzing audio signal,
might it?
For the application to meausure & analyze bioelectric current, please
see,

http://139.134.5.123/tiddler2/bioseon/bioseon.htm

By the way if the sampling interval: d is smaller than the time for
calculating unit Fourier transformation: c, then the total processing
time for n data on the
method used that softawre is n*d, namely it coincides with the total
sampling time. On the other hand, FFT has the total processing time
greater than n*d, or total sampling time when n exceeds a certain
value no matter how small the time c is since the total prcessing time
on FFT is c*n*log(n).

 

[tontoko hirorin]

Your geuss is wrong. It is used in motion control and temperature
servo systems.

For the application to bioelectricity, please visit:

http://139.134.5.123/tiddler2/bioseon/bioseon.htm

 

[Rene Tschaggelar]

You've written as,




I agree. And the high-end of hardware equipped with that software is
up to 20kHz. So it might not be attractive for analyzing audio signal,
might it?
For the application to meausure & analyze bioelectric current, please
see,

http://139.134.5.123/tiddler2/bioseon/bioseon.htm

By the way if the sampling interval: d is smaller than the time for
calculating unit Fourier transformation: c, then the total processing
time for n data on the
method used that softawre is n*d, namely it coincides with the total
sampling time. On the other hand, FFT has the total processing time
greater than n*d, or total sampling time when n exceeds a certain
value no matter how small the time c is since the total prcessing time
on FFT is c*n*log(n).

You're right. It is great.

Rene

 

[Rene Tschaggelar]

Your geuss is wrong. It is used in motion control and temperature
servo systems.

Thanks Guy,
I wasn't aware of such slow processes of interest.

Rene

 

[tontoko hirorin]

You're right. It is great.

Rene

Thank you