[hpsdr] new architecture - FFT front-end

[Alberto I2PHD]

Sorry... I edited the message to change from Dropbox to my own Web space, but left erroneously
the Dropbox link...  here is the correct version.
=========================================================================

On 9/10/2014 12:33 AM, Erik Anderson wrote:

> /Okay, I did just try this with FreeMat.  I chalk up my results to the unfamiliarity of the program; it sounds like 
> I'm hearing the negative side of the iFFT results, and I can't understand what I'm saying (i.e. backwards, fair amount 
> of static), but it's a LOT more readable than I was expecting./


I have done the same with Matlab R2013a. First I digitized 4 seconds of my voice with Adobe Audition at a sampling rate 
of 11025 Hz.
Then I imported the WAV file into Matlab, computed an fft, then an ifft, then wrote out the result as a WAV file, using 
these commands :

[y fs]=audioread('p:\audio\wav\4sec.wav');
fy = fft(y);
iy=ifft(fy);
audiowrite('p:\audio\wav\4sec_fft_ifft.wav', iy, 11025);

I placed both files on one of my Web spaces :

http://www.i2phd.org/data/4sec.wav
http://www.i2phd.org/data/4sec_fft_ifft.wav

Judge for yourself... and please excuse my bad pronunciation.... English is not my first language...

-- 
/*73 Alberto I2PHD*
Credo Ut Intelligam



/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20140910/f2ce5568/attachment-0003.htm>