# Modulation of voice

I apologize in advance if the question is very basic. When I google modulation all I read are the techniques that are used to modulate a signal.

I start from basic. Let's say I am a reporter and the wave sound of my voice changes all the time, meaning its amplitude and frequency at the same time. Now from what I read in order to send this sound wave I need a carrier wave with high frequency. There are three types of modulation that can be used (AM,FM, and PM).

But how can I pick either FM or AM when the sound of my voice change its amplitude and frequency at the same time every second? I want to get the same wave sound at the receiver and I don't understand how a complicated signal can be transferred by just transferring its amplitude change to the carrier (in case of AM).

• Can you please try to clarify your question as at the moment it appears to be too broad? In the meantime, please see this link and this link – A_A Jul 26 '16 at 16:58
• uhm, your voice is an audio signal and the AM or FM or PM transmission of it would be a radio signal. the amount of amplitude deviation for AM or the amount of frequency deviation for FM depends not on the change, but on the instantaneous value of your voice signal at a specific instance of time. – robert bristow-johnson Jul 26 '16 at 17:57
• The following book(s) discusses a lot of mathematical and technical details of voice(speech) production and hearing mechanisms: 1-Speech and Hearing for Communications_H.Fletcher , 2-Speech and Audio Signal Processing_B.Gold – Fat32 Jul 26 '16 at 20:06

I'm not sure I understand the source of your confusion, but here's a basic explanation. Say your voice is represented by the signal $s(t)$, and that its spectrum exists over the frequency range $f=0$ to $f=B$. Then, the signal $$r(t)=s(t)\cos(2\pi f_c t)$$ has a spectrum that exists over the range $f_c-B$ to $f_c+B$. You can choose $f_c$ to suit your needs.
Now, at the receiver you want to obtain $s(t)$ from $r(t)$. So, you can calculate $$s(t)=\text{LPF}\lbrace r(t)\cos(2\pi f_c t) \rbrace,$$ where $\text{LPF}\lbrace\cdot\rbrace$ is a low-pass filter.
Let me try to get back to your question. The voice signal $s(t)$ has instantaneous frequency and amplitude, as you say. These are conveyed perfectly to the carrier when you generate $r(t)$. The term "amplitude modulation" means that the voice signal's "properties" are imprinted on the carrier's amplitude; it does not mean that the instantaneous frequency of $s(t)$ is ignored by the communication process.
The same for FM: the whole of $s(t)$, including its instantaneous amplitude and frequency, are conveyed in the frequency of the carrier. Doing frequency modulation does not mean that that amplitude of the modulating signal $s(t)$ is not transmitted.