Emphasis & Prosody¶
FaceCue reads the character of the voice, works out which moments carry the weight, and carries that continuously through the head and the brows, so the performance moves with the delivery instead of sitting still underneath it.
A mouth on its own reads as a mouth on its own. This is most of what makes it read as a person talking.
On the character, this is the Prosody driver. Prosody is the rhythm, pitch and stress of a voice, and emphasis is the largest thing the driver draws out of it, though not the only one. The inspector says Prosody, and this chapter uses both words: emphasis for what you see, Prosody wherever it points at the inspector.
What It Drives¶
The brows lift on the moments being leant on, and settle back between them.
The head nods with the voice. It dips into the moments that carry weight and comes back out.
The head drifts. A slow side-to-side turn that eases off in the pauses and picks a fresh direction when the voice returns. A little keeps a character alive while it talks, and a lot reads as restless.
None of it is a motion placed on a beat. All of it is a continuous response to the voice, which is why it reads as the character meaning something and not as an animation played on a keyword.
It Runs Continuously¶
Nothing here waits for an event and then plays a clip. The motion is a proportional response to the voice as it goes, so it is always the right size for what the voice is doing. A flat read produces almost no motion, which is correct. An animated read produces a lot.
A moment that stands out a little moves the head and the brows a little. A moment that stands out a lot moves them a lot. There is no point at which the response switches on, and none at which it switches off.
Baked & Live¶
How much the driver can read depends on the line. A baked line comes with its words, so the driver knows where the syllables fall, which ones are stressed, and where a phrase ends. A live line, from Live Viseme or Live Amplitude, has only the voice as it arrives.
| Baked | Live | |
|---|---|---|
| Speech modes | Baked Phoneme | Live Viseme, Live Amplitude |
| Reads | The voice and the words | The voice alone, as it arrives |
| Drive modes | Emphasis, Signal, None | Signal, None |
| Questions & phrase endings | Yes | No |
Under Mode, Drive Mode picks how the driver plays:
- Emphasis, the default for baked lines, finds the emphasised syllables and places a brow lift and a head nod on each, with the character's own pose on questions and phrase endings. It needs the words, so it is baked only. See Baked.
- Signal follows the voice moment to moment, the brows and the head rising and falling with it. It is the one live lines use, and a baked line can use it too. See Live.
- None leaves the brows and the head alone.
The driver keeps one choice for baked lines and one for live lines, and each line plays with the one that matches it. The inspector shows the choice for the Speech driver's mode, so switch the Speech Mode to see the other.
A Fallback Plays Its Line Live
A clip with no bake that falls back to Live Viseme or Live Amplitude is a live line, and it plays with the live choice. So a character that nods on every stressed word in its baked lines follows the voice more loosely on a line that fell back. See The Fallback Chain.
Setting It Up¶
The OneClick Wizard does all of this when it adds the driver. What each piece is for:
- Prosody Rig Map, under References, says which blendshapes the brows use. Without one the brows stay still and the head still moves. What goes in it is set out on the Prosody Channel.
- Enable, under Settings, switches the three parts on their own: Brow, Head Dip and Head Turn. All three start on. Turn one off when your own animation already drives that part, see Animator Coexistence.
- Forward Source, under Setup ▸ Head, is how the driver knows which way the head faces, so a nod dips straight down. Eyes, the default, takes it from the Eyes driver's setup. A character whose eyes are not set up can press Recapture Head Frame instead. See The Head Nods at an Angle.
What It Does Not Do¶
Worth knowing before you tune.
What is authored is the shape, not the moment. The brows do have recipes of their own: the character's prosody rig map says which blendshapes they use and what full emphasis looks like, set out on the Prosody Channel. The head has no recipe at all. What is never authored, for either, is when it plays or by how much. You cannot ask for a particular head movement on a particular word. The voice decides that, continuously, and what you tune is the shape at full and how strongly the character responds.
It does not read meaning. It reads how something was said and, on a baked line, how it was built, not what it means. Sarcasm delivered flatly produces a flat performance, correctly.
It does not replace emotion. Emphasis is about weight and rhythm. What the character feels is a separate layer, authored instead of measured. See Emotion.