the voice of evil?

[Josh Bass]

I'm working on an animated project and want to give my super villainess an unearthly quality to her voice. Any fX/techniques you guys know to augment a regular human voice and push it into the supernatural/demonic/etc. range? Or at least starting points?

I have at my disposal FCP, soundtrack pro, and Logic. Prefer to use FCP since that's where I'm assembling everything, but if one of the others has an essential effect that FCP lacks, that's fine.

[Adrian Frearson]

Well presuming that your voice artist gives a great performance, you could try adding some pitch correction or gender bending effects. I would prefer to do this in STP or Logic rather than FCP, as you have a more control over the audio in these programs.

[Josh Bass]

Thanks. I played around with pitch, flanger, and reverb in STP, got a decent result. She is also doing an Irish accent and I'm worried, with all the FX, about intelligibility. So I've got to find the compromise between sounding super evil and being understood.

I've seen "Aqua Teen Hunger Force" get into trouble with this with the ghost of christmas future robot character. . .could barely understand a word.

[Jon Fairhurst]

EQ and compression are your friends.

I like to combine a heavily compressed track (to fill in the vowels) with an uncompressed track (to preserve the dynamics of consonants.)

Here's a quick tour of voice EQ regions:

* 200-300 Hz: This is the voice fundamental. Vary this to make the voice thin, full, or muddy.
* 600 Hz area: You can knock this broad area down to make room for music and fx.
* 1.2 kHz: This is the key frequency for consonants. You can boost the heck out of a narrow band in this area for dialog without problems. Move the frequency around if needed to avoid resonance.
* 2.4 kHz: This broad area can be manipulated to make things dull or nasal. For instance, with an oboe, messing with this region can make the instrument as smooth as a flute, or piercing and annoying. I think of this as the "character" region. If you have two people that sound similar, boost this region for at least one person to help differentiate their sound.
* 5k - 15k: This is the "air" of the voice. You can boost this to make the voice sound magical, but be aware that too much air causes sibilance problems.

If after the pitch shifting, you have intelligibility problems, try boosting a narrow band around 1k. If it's still hard to understand, try layering a 1k band from the unprocessed voice. Ideally, it will just add clear consonant sounds. There is a risk of phasing problems, but it's probably nothing compared to the phasing issues that you will already get from the other processing.

[Josh Bass]

Thanks. When I was playing last night I was using an already compressed and EQd piece of audio (using settings I'd applied to all the other dialogue). . .now I'm thinking I should bring in "raw" audio, process it, and then play with the EQ and stuff in FCP to see how it mixes, if it's understandable, etc.

[Robert Turchick]

one of my favs from music producing is the reverse reverb. not sure if FCP or soundtrack could do this but in Pro Tools it's easy.

reverse your VO track so it plays backwards. apply reverb...record the reverb to a new track...then reverse the new reverb track and your VO to play forward. very spooky!

[Josh Bass]

I'm not quite sure I understand "record reverb to a new track". Do you mean take the tail reverb (what's left after the person is done talking and it's just pure reverb) and put that on a new track? Or do you mean something else?

[Robert Turchick]

It's a little tricky unless the software can buss effects. That's why Pro Tools works for this effect. Essentially your "session" would have the following tracks:

1) VO (reversed) with aux send to reverb (track 2)
2) Reverb Aux track bussed to new audio track (track 3)
3) Audio track to record reverb

What you are trying to do is make the reverb an actual audio file rather than a plugin effect. That way you can reverse it.

Later tonight I'll take some screen grabs and post an example.

[Josh Bass]

Ah. That's what I thought you meant. I'm sure there's a way to do that either with STP or Logic, but it is way outside the scope of my knowledge of either program. I'm working around this particular character's lines 'til I get it all resolved.

[Robert Turchick]

I haven't used logic in a while but I'm sure you could do it in that.

[Josh Bass]

My brief research into the STP manual suggests that you can create busses, but not route just an effect to it.

OH, PS, it's Logic Express, not Studio.

[Josh Bass]

Ok. . .I've gotten to the point in logic where I have a bus with what I THINK is just the reverb effect (sounds like the audio itself, but weak, very metallic, and of course, reverby), but it's just an empty track, there's no region/clip or anything that I can reverse.

[Josh Bass]

Ok, I had a "duh" moment and tried something. Is the effect you describe similar to this: I take a clip in STP, add reverb, export as new AIF, bring that AIF back into STP, and re-reverse it (making it play normally again, that is). So now you have the weird backwards reverb on a forward clip? It's definitely cool, but makes it very difficult to understand. I guess you like the separate bus thing 'cause you can control the amount of the effect? I could likely do something similar by controlling the plugin in STP, how much wet/dry/etc. I add in the first place.

[Robert Turchick]

yup! That will do it too. And yes, the nice thing about having them on separate tracks is you can easily adjust.
The other tip I'll give is keep the reverb length fairly short. Maybe .5 - 1 sec decay
Have fun and be sure to post results!

[Josh Bass]

Well, funny thing. I would post the results, but the piece is fairly unPC short about Jewish ninjas. And the only lines that will have the FX are fairly inflammatory.