Dragon voice matures (Janne)

Dragonbane Diary
Saturday, June 3rd, 2006
Virrat, Finland

Last weekend saw a spur of activity concerning the Dragon voice process. Friday evening, I travelled to Turku, where Heiko picked me up and took me to Loimaa. In Loimaa, the Dragon is being summoned from nothingness into existence. I'm not exaggerating a lot by stating that I saw the light, it was simply impressing. A set of materials arranged into a skeleton, soon to be cast into life and into a living creature.

On Saturday, I went through two laptops trying to squeeze the power hungry process into their puny existence, until I was told about an unused PC, which could handle the magic. Luckily the owner of the PC, one of the main dragon wizards and the owner of the place, hadn't been able to install Linux on the resisting machine (he had it on all the other computers already). Of course the installation had failed, because the voice processing software runs on Windows and Mac, and the all so lovable Windows was now all the Red One needed.

The base was now secured, and it was time to make the air vibrate. We gathered the elements of the Dragon's vocal machinery, and layed them out at the corner of the large yard. After initial sound checks and the joy of playing music loud the Dragon system was connected, and for the first time, the sound was travelling through reality. And getting fed back into the microphone, which resulted in the quick zeroing of the power amplifier for preventing everyone getting permanent hearing damage, and more importantly, to not destroy the earthly ingredients of the Dragon voice chords. A few adjustments later we continued our session and made notes on how to improve the voice further.

During the evening and some way on to the night I then mumbled incomprehensibly into the microphone of a headset, and built the process deeper. For some reason the other people, who were not able to hear the resulting sound, but my voice only, gave me a few odd looks. Luckily for everyone, my station was in the dressing room of the sauna, and it had this air lock style two door system connecting it with the rest of the house.

The next day went a little smoother, and the voice sounded initially really nice. We did some balance checks and changed equalization for more fluid reproduction via the powerful vocal system. The voice is still in progress, but we are close now to please the Red One.

What can I say about the actual process then? The main element is a four-voice pitch shifter, to which original voice is fed in, the fine tune is modulated by two out-of-sync lfo's. Parallel to that process there are two pitch shifters, the other taking it down an octave, the other taking it up by 4 semitones. A biquad filter, with frequency and bandwidth modulated by three freely oscillating lfo's. A vocoder where the original voice is fed in as the carrier, which is modulated by low frequency wind, steady full fire, violent whooshes, a high pitched undulating oscillator, the pitched down jaguars, wolwes, raccoons and a a female scream. The vocoder output is then put through a freaky phaser. All this is then mixed in a software mixer, the output of which is harmonically enhanced, equalized and alongside this the ready sound now pitched up three semitones, equalized for high frequencies and mixed back with the original processed voice. The lot is then dynamically limited for protecting the innocent from the sudden bursts of sound energy.

More earthly ingredients are on their way, and the final combination for unleashing the ready Dragon voice awaits in the very near future.

Janne Särkelä
Dragonbane audio master