Whats in a voice?

With the release of my latest Orc sound pack, I thought I’d get into a little (or maybe a lot) about the key processes behind designing a voice.

There are some powerful tools on the market today. Full package voice changing software like 'Voxpat 2' and 'Dehumaniser' are renowned for their ability to make powerful and extreme changes. Yet while such massive differences from the initial recordings to the end product is certainly fascinating to hear, I find myself having little need for such radical software, instead drawing smaller elements from a variety of plugins based on what they do best... or even worst...

Welcome to the world of Voice design.

As with nearly everything in the world of sound, the first part of your chain is almost always the most important. And in this case that’s the voice. You may be surprised at how much, or more precisely how little digital voice processing goes into the game voices and the vocal libraries I've made...

Let’s jump in...



Plugins much?

For my latest Orc release, the idea was for the sound not to be a generic humanoid “beast” sound, but distinctly Orcish. Save for cleanup and EQ, Only one plugin was needed for each vocal. Sometimes "SoundToys Little Alter Boy" with it’s clean clear pitch shifting made an appearance, but for a little extra spice "Antares Mutator" was overwhelmingly the winner. Using some downward pitch shifting, the extra spice was added with small tweaks to the throat width and length options the plugin has to get it sounding just right.

The rest was up to my ability to express an Orc.


If it’s fixed, break it!

People often ask me if I used some sort of vocoder to make the Bleep library. No I didn't use a vocoder. In fact I didn't even shift the pitch. All I did was change the formant. With a bit of a twist...
"WAVES Ultrapitch" as an older plugin, doesn’t have quite the speed modern pitch plugins have to adjust the pitch or formant. As such sudden or quick changes of the pitch in the source material with the plugins slower processing can result in some strange robotic artifacts. With all the heavy purry warbles I did with my voice, the plugin had a difficult time with the shifts resulting in some beautiful “mistakes” that modern plugins more easily avoid. It's voice mode option allows you to choose the speed at which it makes changes. This gives you greater control to introduce robotic effects as a result of slower shifts.

Bleep DEMO

Mini Ninjas - Square Enix
When I worked on The Mini Ninjas franchise, the only real processing done for the small enemy ninjas was resampling the voices by varying degrees to be faster. That’s it. It really came down to the voice actors abilities to make zany style sounds.


Intel Warrior Wave.
If it’s not broken, don’t fix it...

This was a lemmings style where you had to save Roman as many little Roman hoplite characters as you could using your hands as the controller through motion sensors. To add to the difficulty they included Good guy Hopites and Bad Guy Hoplites that if they came into contact with each other, so I had to do similar voices but in good style and evil style.

It ended up being a matter of fine tuning the performance while recording rather than using any processing at all.

We opted for a kind of simpleton funny voice for the good guys, and the same sound with a angry rough edge for the bad guys.




Romans from Mars


And I use Romans from Mars as the final example because it contained a little bit of everything. A touch of re sampling, a little pitch and formant changing, and of course a lot of expression.

With that said...
To maximize your voice designing abilities you really need to know all the nuances, quirks, strengths, and even weaknesses of your plugins. But the foundation of all voice design will come from the emotion, expression, and character that is imbued just before it hits your microphone...

“So how do you create the character behind the mic?” you may be asking.
That my friends is another story, I'll save for another time...

© 2018 Yarron Katz - Sonic Worlds