## PSN-L Email List Message

Subject: Re: Most Recent Discover Magazine pp66-67 WaterFall Display
From: "Geoffrey" gmvoeth@...........
Date: Sat, 14 Jul 2007 03:53:53 -0700

```Thanks for the Info.

I searched and searched and found only
one PDF file that explained things in
ways I could somewhat understand.

It seems the data fed into an fft routine needs
to be symmetrical +/- between one and zero
in magnitude as samples voltages or whatever
and the results is given as a series of
freq points that are the original sample points / two.
You start out with 1024 samples and end up
with 512 freq points that are in two arrays of
Sin and Cosin data. to get the final results
you need to use the pythagoras formula
to combine the two a = sqr(r^2 + I^2) maybe x 0.7071
to get the RMS value.

The data can be any sample rate and the results will
be interpreted according to the sample rate of the
original data. The freq spread will be sample rate
divided by two.
DC to that freq.

It will be a slow process to calculate a decent length of time. But it will let you look at the overall
response of your system especially when you record
storm noise of a passing storm which is like
infrared noise. The noise will show the passband

regards;
geoff

----- Original Message -----
From: "Mike Price"
To:
Sent: Sunday, July 08, 2007 11:39
Subject: Re: Most Recent Discover Magazine pp66-67 WaterFall Display

> Geoff,
> The FFT algorithm is well documented and implementations abound. The FFT
> is widely used by anyone interested in exploring data in the frequency
> domain and it would be hard to find an algorithm better explained or
> more readily available. A Google search for "fft source code" generates
> 1.3M hits; indeed, the first one is www.fftw.org which has downloadable
> C source and links to many other sites containing source in several
> languages. You can find a basic implementation at:
> http://logosfoundation.org/fft/fft.html for example.
>
> The notion that access to FFT code is somehow restricted is misguided.
>
> Mike
>
>
> Geoffrey wrote:
>> This program seems to not use data files
>> but wants you to build a custom A/D that
>> will be read by the program itself.
>> Not sure a sound card will take sample rates
>> as low as 20Hz because most sound hardware
>> has a capacitor input that rejects DC or freqs
>> below the sense of hearing.
>>
>> It would be so nice to have this feature in Winquake to
>> do an Analysis of file data in TXT/8/12/16/24 bit form
>> or text form all you have to do is limit the frq
>> between Dc and 5 Hz or so and make atleast
>> 1024 samples per spectrum line.
>>
>> I have seen the routine in the library at ASU
>> to FFT and there seems to be several ways.
>> The best i like seems to result in the final
>> results with the number of samples to be worked
>> setting the resolution over the band.
>>
>> These people who write these books on FFT are oriented
>> only to their own kind so if you do not know calculus or higher
>> transforming their ideas into a computer program is very difficult.
>>
>> But I have in the past done this successfully myself.
>> On a very small scale with a VIC20 machine running 10MHz.
>>
>> You need to take the first 1024 samples then make an imaginary
>> set from the same ones then do a butterfly operation
>> (some kind of fancy math) that results in a backward
>> spectrum with each sample now representing a frequency band.
>>
>> Then you straighten out the results to get the low freq at the low
>> sample number position.
>>
>> The final results is the freq range represented within the sample set
>> chosen
>> and that would be line one.
>>
>> Then you advance one sample in the data and do the same for the
>> next contiguous 1024 samples.
>>
>> This is repeated until you have built a picture of your entire data file.
>>
>> The sample rate determins the freq limits I think.
>>
>> Why do not more people take an interest
>> in the math and programming and hardware
>> because seismology is just a fine specialty
>> and I think more serious PROs should take an
>> interest in fostering human interests in all lines of Science.
>>
>> You got to be an MD or equiv before you can
>> get into forensic medicine ot Toxicology but this
>> Seismology stuff is non license stuff and great way to
>> learn many facets of science.
>>
>> Please someone with Math background make us a
>> Basic Program like QBASIC or POWERBASIC
>> to do one line of FFT and I will take it from there
>> to make the whole program.
>>
>> I did this once myself but did not record the final
>> program and would have to learn everything all over
>> again to do this program again.
>> A Burgler stole the VIC20 laptop with all my programs
>> several years ago so those programs are lost.
>>
>> There is more to amateur stuff then just watching the waves
>> and building things with your hands. It is a great chance
>> to play general science and bring most everything together
>> to see how it works with NO MYSTRIES.
>>
>> There seems to be effort to compartmentize information
>> in the USA because of its capitalistic nature forcing one
>> to become a college goon before you are able to
>> do anything academic. The truth of the matter
>> is to get a formal degree the system studies you as much
>> as you study your subject so there is no real privacy
>> government awareness. The Counter Intelligence peoples
>> do not want Academic Freedom in the USA.
>> Even tho we are playing with this harmless passive seismology stuff.
>>
>> I have come to realize that in some European Countries
>> they have laws amounting to NEPLEONIC LAW which basically
>> says everything is illegal except that which is specified by government.
>>
>> These kinds of forces are what we up against even in this
>> relatively harmless area of seismology.
>>
>> If there is no common place to get knowledge we must make it ourselves
>> even if it angers the powers that be.
>>
>> It is no accident we can not find a simple BASIC FFT source code
>> with logic trees etc in laymen terms for our interest area.
>> Everyone is trying to keep secrets or make money.
>>
>> For the sake of the evolution of science lets
>> please gather together a collection of utility programs
>> free for universal use in this PSN group.
>>
>> I am assuming PSN means Public like WORLD PUBLIC and not just
>> California Public.
>>
>> One LAST thing for NOW.
>> Black and White and shades of gray is by far the best for resolution
>> If a person is lucky to get a totally black and white monitor or green
>> you will see no sharper images than with such an analog electrostatic
>> scanning device.
>> The military used to preferr these over color since it used to be all
>> male
>> (for warships)
>> and males have the highest incident of color blindness.
>>
>> Regards;
>> geoff
>> __________________________________________________________
>>
>> Public Seismic Network Mailing List (PSN-L)
>>
>> To leave this list email PSN-L-REQUEST@.............. with the body of
>> the message (first line only): unsubscribe
>>
> __________________________________________________________
>
> Public Seismic Network Mailing List (PSN-L)
>
> To leave this list email PSN-L-REQUEST@.............. with
> the body of the message (first line only): unsubscribe