From: Ikaro on
If you are looking for a method that does not alter the pitch and is
has flexible mapping I suggest also:

"Discrimination of speech processed by low-pass filtering and pitch-
invariant frequency lowering", Reed, Hick, Braida, and Durlach,J.
Acoust. Soc. Am, 74(2), August 1983

With the warping being related to all pass filters as described in :

"Discrete representation of signals", Oppenheim, Johnson, Proc. IEEE,
60, 1972

-ikaro


On May 24, 10:17 am, Jerry Avins <j...(a)ieee.org> wrote:
> On 5/24/2010 9:38 AM, emitrax wrote:
>
>    ...
>
> > Looks like I posted to the right group after all!
>
> "We do just about everything" (TM)
>
> > Thanks to both a lot.
>
> You're welcome.
>
> > I think I'm going to implement a prototype with matlab
> > DSP toolbox first and then I'll do it in C or Java.
>
> For pitch-shifting code, seehttp://www.dspdimension.com/
>
> Jerry
> --
> Engineering is the art of making what you want from things you can get.
> ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

From: Vladimir Vassilevsky on


emitrax wrote:
> Hi,

> I'm looking for a project to start learning DSP programming, and an
> idea
> that come to my mind was to implement an algorithm that, getting my
> voice as input, would output the voice with the helium effect.

I am looking for a project to learn mechanical engineering, and I have
an idea of implementing a Space Shuttle.

> So I'm wondering if any of you have any lecture to point me to?

Get a basic book such as "Understanding DSP" by Rick Lyons and prepare
for several years of learning.

VLV
From: robert bristow-johnson on
On May 24, 7:38 am, "igor b." <i...(a)REMOVEhyperglitch.com> wrote:
> On 24.05.2010 12:36, emitrax wrote:
>
>
>
> > I'm looking for a project to start learning DSP programming, and an idea
> > that come to my mind was to implement an algorithm that, getting my
> > voice as input, would output the voice with the helium effect.
>
> > I've googled a bit to see if there is already an algorithm that explain
> > how to implement this, because I certainly don't have the knowledge to
> > come with the algorithm myself (I just wanted to implement it for fun),
> > but I couldn't found anything.
>
> > So I'm wondering if any of you have any lecture to point me to?
>
>
> Helium effect is basically shifting up the pitch of a voice. Probably
> easiest way to do it is with phase vocoder.

easier than the phase vocoder is a time-domain method based on Lent's
algorithm (Computer Music Journal, Winter 1989). check out US patents
5641926 and 6336092. this time-domain method works only on a single
voice (a quasi-periodic signal) and needs a good pitch detector.

this has been done before. Roland used to have this little box called
a "voice transformer".

r b-j