[Elecraft] [K3} [KX3] SSBT like CWT ?

[Eduardo González]

Hello friends.
I have worked on algorithm about SSB auto-tune and is a reality. I
have tested code using PC simulation software (Matlab) and using
realtime coding using C language. Just now i have several students
implementing such algoritms on BeagleBoard xM, Raspberry-PI and
STM32F4 boards using external ADC and DAC (except Beagleboard xM, it
have ADC/DAC in-board)
Effectively, as said VK4JD using equidistant spectral components of
harmonics sounds (human  voice) is possible get mistuning (using
special version of Ceptrum or two cascade FFT), but great issue is
caused by fricative voices (T,X,P,G,K, etc) because algorithm require
using a audacious estimation algorithm (as histogram) requiring
additinal computational cost.

Several years ago, Robert Dick ("Tune SSB Automatically", Robert Dick,
QEX, January 1999, pp 9-18) worked succesfully autotunning, in fact
this work i used as start point.
I have a paper about it, but was written in Spanish, use a traductor
to get details: redalyc.uaemex.mx/pdf/707/70712302.pdf

Sorry, my english is horrorous

Edu (YY4GMJ)

On 2/5/13, Peter Lambert <plambert at qa.com.au> wrote:
> Hi Erik,
>
> It's R & D.  Anything could happen.
>
> I agree it's obvious but so are many things we see around us that weren't
> there a short time ago.
>
> The fundamental (vocal cord vibration frequency) changes constantly but not
> particularly fast (it is a mechanical thing we do with our muscles).  The
> constant is the original carrier frequency.  The relationship between the
> vocal cord harmonics have a computable relationship back to the original,
>
> I do think it's doable.   I've done some work on speech compression that
> gives me some level of confidence.
>
> Cheers,
> Peter
>
>
> -----Original Message-----
> From: elecraft-bounces at mailman.qth.net
> [mailto:elecraft-bounces at mailman.qth.net] On Behalf Of Erik Basilier
> Sent: Tuesday, 5 February 2013 8:05 PM
> To: Elecraft at mailman.qth.net
> Subject: [Elecraft] [K3} [KX3] SSBT like CWT ?
>
> Peter, yes, I am sure many have thought about this. For something so
> relatively obvious, one has to wonder why it hasn't already been done.
> Perhaps the rapid changes in typical contest speech makes it hard to get an
> accurate enough fix on the spectrum, unless one asks the other party to say
> a long "aaaaaah" or some such sound? Or maybe the computing speed of the
> hardware is a practical obstacle for a given rig? Maybe it has already been
> tried by several people, and they ran into some kind of obstacle? Does the
> frequency around 85 Hz stay very constant for a given speaker, as he or she
> produces a variety of sounds?
>
>
>
> 73,
>
> Erik K7TV
>
> ===============================================
>
> This is an algorithm I've contemplated writing many times.
>
>
>
> Voice results from the vocal cords snapping together in the range around
> 85Hz.  The filter effect of the various cavities, throat, mouth and nasal
> are quite high Q and filter the harmonic rich vocal chord excitation to
> produce the voice.
>
>
>
> If you watch an SSB signal on your PSK waterfall you'll clearly see the
> harmonics of the vocal chords.
>
>
>
> I think a DSP algorithm that calculates the location of the carrier based
> on
> the separation between the vocal chords is a definite doable proposition.
>
>
>
> It would give SSB the perfect clarity you seek.
>
>
>
> Love to have a go at this (I do DSP).  Do you Elecraft guys want to have a
> go or interested in using it if I do it ? (I could probably do it as a
> piece
> of PC code that processes sound card input so you could evaluate it).
>
> Doesn't feel like to be a difficult thing to do.
>
>
>
> Cheers,
>
> Peter VK4JD
>
>
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>