Change emphasis of syllables in a voice?

Hi all,
I have some audio files and I would like to alter the emphasis of certain syllables to change the meaning of what is being said. Here are some samples to illustrate what I mean:

So I imagine that in order to get some of the changes that I would make small changes to the volume, maybe pitch slightly higher and perhaps even the speed but only on certain parts of words and many times these changes would be gradual and sometimes quite sudden. Any advice on how to change (for instance) the first wave file into a question through audacity? Or would it be easier using a plugin or some other program? Or isn’t it even a feasible endeavor?

Any help would be great.


Windows 8.1
Audacity 2.1.2

Next to impossible to produce something which is convincingly human-sounding.
Audacity does have a time-track feature which warps the speed, changing both pitch & tempo …
Time track in Audacity, warping pitch & speed.png
And a similar looking thing called envelope tool which just changes volume …

But the results will sound like a tape being chewed-up on playback.

I wouldn’t say “impossible”, but “very difficult”.
To learn about voice analysis, have a look at this website:

You need the Acting Filter or Theatrical Expression Modifier.

I’m making all that up. This is one of those jobs that it’s tempting to say it’s impossible. It’s not, but it’s very highly impractical. So you’ll need Envelope tool (two white arrows and bent blue line). That will let you change volume on a fractional second basis.

Then Effect > Sliding Time Scale/Pitch Shift. That will let you move tonal quality around, but note you only get one shift. So Sliding up to a question and back down is two shifts—minimum.

Pauses are harder to make sound natural. “And THAT [pause] would be why.”

Effect > Change Pitch, speed, and tempo.

The change tools notoriously fail if you try to push, say, pitch more than one or two piano notes. All the sounds of your voice don’t change during expression. When they do (by accident) it has the effect of putting your finger on the record while it’s playing.

When you need a longer pause, you can’t just cut a hole the right size. Sudden Blackness of Space holes sound like the player or radio just died. You have to put Room Tone (normal background sound) in the hole. And it has to match.

There was an AudioBook reader who was prepared to edit each word of his presentation before submission. We figured he would still be editing his work when he retired. How old are you?

Regard the list of Very Highly Paid Actors doing voice performances behind animations and cartoons. This is why.

I heard a radio show devoted to the extreme measures some companies were going to in order to heave a script into a big bin and have convincing voices come out the other end. We’ll notify you when they get one to work.


Rocket Raccoon in Guardians of the Galaxy was voiced by Bradley Cooper.


Great responses from everyone!

I wasn’t aware of the Time Track feature so that may help and I will read over the speech analysis website which may offer clues.

I’m concluding though that the short answer is that it’s not IMPOSSIBLE but certainly IMPRACTICAL and very difficult to get it to sound close to realistic. A far better option is to record it properly to start with and just clean it up a bit if needed.

THanks again,

Windows 8.1
Audacity 2.1.2

Yes, that’s the concise answer :wink:

There’s a cousin problem called the Uncanny Valley. You very rarely see a live actor animated on the screen. We’re so sensitive to what faces look like that if the animation isn’t absolutely, totally, perfectly correct, the audience sees through it immediately and the performance falls apart. Studios go through the expensive and bother of a live reshoot rather than try to fake it with pen and paper (computers).

That’s also why game actors don’t look any more natural. It’s better to make a face unquestionably fake than get too close to real, but not quite.

Your head doesn’t like close but no cigar. It triggers suspicion, fight or flight and other nasty impulses.