How to give speech a Donald duck effect?

[dexterdev]

Hi PF,
I read that in the communication engineering book by simon haykin saying that if the phase response of the voice signal is altered we can get the donald duck effect. But how should I play with the phase response. Any advice.

-Devanand T

 

[sophiecentaur]

A poor SSB demodulator will give a good 'donald duck' effect. If you have a DSP program then you could arrange that quite easily. However, I'm not sure that a simple 'messing about' with phases would do what you want - more like frequency shifting would be needed. Inverting the frequency spectrum of the audio could do it, perhaps - easy to do, once you have DFT'd a short passage of speech.

 

[Enthalpy]

I have the strongest doubt about our ears detecting any phase. I had made a software to synthesize a sound from its harmonics (...hence only periodic sounds... I had been fooled by Helmholtz' initial mistake) that could adjust each phase individually and that made zero difference for the ear, nada, niente, nitchevo, walu.

Though, I haven't still checked if we locate a noise source by the phase difference between our ears, so I could still accept this allegation.

De-tuned SSB let's the local oscillator beat, hacking the received signal quickly. This deforms the voice for sure. >30 years ago a hobbyist paper proposed a phase inverter driven by an oscillator to do this on voice and instruments. On electronic pianos then, the "chorus" effect was obtained this way; it sounded more or less like several instruments playing not exactly in tune.

 

[sophiecentaur]

I have the strongest doubt about our ears detecting any phase. I had made a software to synthesize a sound from its harmonics (...hence only periodic sounds... I had been fooled by Helmholtz' initial mistake) that could adjust each phase individually and that made zero difference for the ear, nada, niente, nitchevo, walu.

Though, I haven't still checked if we locate a noise source by the phase difference between our ears, so I could still accept this allegation.

De-tuned SSB let's the local oscillator beat, hacking the received signal quickly. This deforms the voice for sure. >30 years ago a hobbyist paper proposed a phase inverter driven by an oscillator to do this on voice and instruments. On electronic pianos then, the "chorus" effect was obtained this way; it sounded more or less like several instruments playing not exactly in tune.

If it's tuned the 'wrong side' it will invert the audio frequency spectrum, too.

 

[alphysicist]

.

Hi Devanand T,

Thank you for bringing up this interesting topic. The "Donald Duck effect" is a well-known phenomenon in the field of communication engineering, where the human voice is altered to sound like the famous cartoon character. As you mentioned, this effect can be achieved by manipulating the phase response of the voice signal.

To understand how to manipulate the phase response, it is important to first understand what it is. The phase response refers to the relationship between the input signal and the output signal in terms of their phase shifts. In simpler terms, it is the delay or advancement of different frequencies in a signal.

To achieve the Donald Duck effect, the phase response of the voice signal needs to be altered in a specific way. This can be done by using digital signal processing techniques such as phase shifting, filtering, and time-domain manipulation. These techniques allow you to manipulate the phase response of the voice signal, which in turn alters the sound of the voice.

However, it is important to note that achieving the Donald Duck effect requires precise control and manipulation of the phase response. It is recommended to consult with a communication engineer or an expert in digital signal processing for specific techniques and advice on how to achieve this effect.

I hope this information helps. Best of luck in your experiments!

Sincerely,