Misc Effects #
Auto-Tune #
Doing autotune in code is well beyond the scope of a small comment. Outside of just throwing AI at the problem - witch is likely to sound god awful - there’s no magic bullet here. It’s spectral analysis hell.
Given it’s basically a meme now, I assume you know what Auto-Tune is. What you may not realize is that not all auto-tune is, well, auto. A lot of the time it’s done manually, painstakingly correcting the tuning. If the original sound is close enough, it’s usually pretty easy to get it to sound natural, too.
Also keep in mind pitch correction isn’t just for vocals, you might want to pitch correct your guitar playing, change a note in an already recorded sequence when you change keys, or even just use it as a glitchy effect on pitch bends.
Vocoders #
Traditionally, It’s a lot of band pass filters with envelope followers. That’s it. That said, actuall implementiton, use an FFT, not a fuck load of filters. See https://dsp.stackexchange.com/questions/1232/programming-a-vocoder
Vocoders are a bit weird to understand, and really need a visual explanation. This first video is really trying to sell you on the ‘OVOX’ vocoder, but does start with a decently good explanation. The second video is for a particularly neat vocoder-y, pitch-shifty effect that I think does the vocoder effect better than your traditional vocoder. It is kinda pricey though.
Note, that when you hear a vocoder you may confuse the sound for a talkbox, the way they work is pretty dramatically different. There’s an extraordinary good video on talkboxes embedded at the bottom of this page.
Pitch, Glitch, & Granular #
Most of these come down to recording a buffer, occasionally playing back from that buffer, and doing weird things with the sample rate or reversing the order. Some pitch effects, like the PitchFork+ below, are real time and are doing much crazier voodoo. If you just want to do monophonic octave up/down, that’s easy though. Just use a PLL (or code to simulate one) and multiply it for octave up or just use a digital counter and output on every-other above-threshold point for octave down. This just tends to sound like ass with polyphony.
Rather than try to explain all of these, I’m just gonna throw a grid of videos at you. Skip around, hear some neat sounds. Get inspired.
Spectral Editing #
De-Noising & De-Essing #
(Make sure you didn’t gloss over Noise Gates in the Dynamics Chapter)
Ring Modulators #
Talk-boxes #
There’s really no way to emulate this, as it’s less a matter of the sound and more of the interface - using a real mouth as a filter. Vocoders are basically the closest you’ll get, but you lose out on the infinite sustain. Even digital has limits, I’m afraid.