Hoo boy, that was a tough one! Fun though! Which I suppose is a common enough theme, the tougher something is, the more fun it is to do it. When I first started mapping out this episode in my head I thought to use the Prometheus Trailer Theme for the summoning ritual. But I didn’t feel quite right about it – it was a very recent movie, so it felt fad-ish, and it was a pretty bad movie to boot. I spent way too long searching for other options. There were many I liked (the Inception theme was almost used – it has a very similar sound, and comes from a much better movie. But it sounds less horror-esque, and the music is actually an important part of the plot of the movie, I didn’t want to bring in all that baggage). Ultimately nothing really hit me as just over-the-top enough with a horror feel like the Prometheus Trailer.
Of course the full trailer theme is only a minute and a half long, so I had to do some looping to stretch it out. Fortunately the song already has a lot of abrupt swells and falls, which made good break points. And it allowed me to time the fade-out very nicely to the narration. The seams aren’t perfect, but they work.
The voice-masking effect was tricky. Generally for 60-Minute-type shows they just drop the pitch drastically. The “blurred into unidentifiability by a buzzing undertone” is easy to imagine, I’m sure we all know what that should sound like, but how the hell does one actually do it? Blurring too much makes the words hard to understand, which is probably OK if you’re beating up 11-year-old girls, but it’s a big no-no for an audio book. Underlaying a buzzing doesn’t really mask the voice at all, it just creates a buzzing in the background. I tried mixing the audio with white noise via the vocodor option, as was suggested at various sites, but that produced too much of a robotic auto-tuned sound.
I sometimes adjust the tempo of speech a bit for various nefarious purposes, but I quickly learned you can’t alter it by more than a few percent, or you get a “clicking” effect. I decided to use that to my advantage in this episode, first dropping the tempo by 70%, then speeding it back up to normal again. The clicking is an irremovable artifact of the slowing process, so bringing the voice back up to speed left the distortion. It’s not the world’s best masking, but it provides plausible deniability, and it sounds similar to how it was described.