Double incoming frequency through soundcard

[Kway]

Hi,

I'm having a problem with my program...

I'm trying to make a frequency doubler. (Look at the vi in the attachment)

But when I run this vi, the sound has an delay and the frequency aint doubled...

Anyone knows what's wrong?

(written in labview 8.0)

[SteelStonehenge]

The most this VI  could possibly do is multiply the amplitude.  If you want to multiply the frequency, you have to sample into a buffer and then  read back out at a higher sample rate.

Eric

[Mellobuck]

I agree with Steel, you are only multiplying the time-domain signal in the existing V I. Doubling the frequency will probably involve a fourier transform into the frequency domain and some operations there. With your comments it seems like you are doing this and making a real-time voice changer? If this is your goal, I can blow some of the dust off of my systems and signals references and step you through the frequency and/or time domain math...give me a day to search through my library.

[Kway]

Yeah that's thrue, I'm making a real-time voice changer... I'm going to try to find out how and with what, but I will always be glad if you could help me 😉

greetings

[Mellobuck]

Hello Kway,

Sorry for the long response time, I was going back a few years and through a few books in the attic. First off, despite this being a LabVIEW discussion, there are single chip designs out there which will perform these functions. If you are concerned about time delay (as it seemed) solid state is really the way to go. Check out the HT8950 single chip CMOS LSI voice modulator. However, for the edification of the signal processing crowd - what we want to do here is basically modulate and then demodulate the signal. This is what your AM band radio does. But, what is to be done is modulate using one carrier wave, and then demodulate using a higher freqency carrier wave. This will give you the chipmunk voice without speeding it up. Using a lower frequency demodulation would give you the Barry White voice.

How to do this in LabVIEW.

What you'll need is the convolution and deconvolution operations. Convolve at 50mHZ and Deconvolve at 100MHz and you will get a signal with twice the frequency, but same time scale, ie the voice with be higher, but not sped up. You won't really notice the delay either.

Peace

Message Edited by Mellobuck on 11-29-2006 09:46 AM

[InstaAxEToast]

I was wondering if anybody could help me out with doing this method of frequency doubling?  So do you simply take a 50 MHz sin wave and input it into the convoluter as well as the audio input? Also I tried to look for the deconvoluter but it says that it is not in the standard libraries. Does anybody know where I can get the deconvoluter?

[InstaAxEToast]

Anybody?

[Mellobuck]

Hey There InstaAx,

So, let's say your audio signal is X, and your carrier is Y, you would put X and Y into the convolution function and get X * Y (X 'convolve' Y, not X times Y). To deconvolve, you input X*Y, and also need to input Y so that you can retrieve X. In your case, you would need to convolve X * Y1, then deconvolve giving it Y2 (higher or lower freq), to get a sped-up or slowed down X.

Looking through my library, I have it in my Signal Processing>Sig Operation, but I think that came with the Signal processing toolkit, which doesn't sound like it is on your computer. If you do have the ability to have MathScript, there are equivalents in Matlab M code to do this processing, which might be found on another forum. Otherwise, the math gets very messy to do it explicitly with the more basic LabVIEW functions.

-Mello

Message Edited by Mellobuck on 04-11-2007 09:39 AM

[InstaAxEToast]

Thanks!

Well I sat here with the tech and we went through the 27 cds. We managed to find the signal processing toolkit. We installed it, but still no deconvolution. Oi

[Mellobuck]

Hmm, De-Convolution is right next to my Convoution function, strange that you have one and not the other...