← all dimensions · markdown (agent skill) · manual home
Oral Resonance VoiceNet dimension guide
Resonance Placementcode vn_R_ORALslot 77
This dimension measures resonance originating from the oral cavity. Below: the merge-scaling dose-response for the __high and __low LoRA (laion/moss-voicenet-dimension-loras) at 25/50/75/100/125%, the measured target value at each dose, the top-3 ± correlated EmoNet emotions and VoiceNet dimensions, three audio samples per dose, and a grounded prompting tip. Costs are shifts vs. the 24-sample neutral baseline in genuineness, vocal-burst blend and speech quality.
Oral Resonance controls resonance originating from the oral cavity. The __high LoRA shifts it +0.18 at 100% (worst cost: burst-blend -0.39); the __low LoRA shifts it -0.07 at 100% (worst cost: burst-blend -0.19). The __high direction is the stronger, more reliable knob here.
▲ HIGH — Oral Resonance __high
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.310 | — | — | — | — |
| 25% | 0.325 | +0.01 | -0.05 | -0.08 | +0.04 |
| 50% | 0.380 | +0.07 | -0.07 | -0.16 | +0.07 |
| 75% | 0.345 | +0.04 | -0.09 | -0.35 | +0.02 |
| 100% ★ | 0.489 | +0.18 | -0.12 | -0.39 | -0.02 |
| 125% | 0.443 | +0.13 | -0.12 | -0.38 | -0.04 |
EmoNet emotion correlation
Sourness+0.55
Disgust+0.45
Contempt+0.44
Fear-0.45
Longing-0.45
Sexual Lust-0.38
VoiceNet dimension correlation
Mask Resonance+0.76
Valence+0.70
Head Resonance+0.68
Vulnerability-0.50
Monologue Style-0.47
Cognitive Load-0.45
Listen — 3 samples per dose
25%measured target: 0.30 · 0.30 · 0.23
50%measured target: 0.21 · 0.35 · 0.33
75%measured target: 0.31 · 0.21 · 0.23
100% ★measured target: 0.23 · 0.33 · 0.48
125%measured target: 0.66 · 0.31 · 0.46
Merging Oral Resonance __high pushes the dimension: at the recommended 100% dose the measured Oral Resonance score moves from a 0.31 baseline to 0.49 (a +0.18 shift). The dose-response is mostly ordered but noisy between doses. That reach is a steep cost — the biggest hit is to burst-blend (-0.39); across the board genuineness -0.12, burst-blend -0.39, speech quality -0.02. Emotionally it reads as Sourness (r 0.55), Disgust (0.45) and Contempt (0.44), and it moves away from Fear, Longing and Sexual Lust. On other voice dimensions it co-moves most with Mask Resonance (0.76), Valence (0.70) and Head Resonance (0.68), while pulling down Vulnerability, Monologue Style and Cognitive Load. Practical call: merge __high at 75–100%; expect Sourness and Mask Resonance to rise and burst-blend to fall. If you don’t want Sourness bleeding in, offset it with a prompt cue toward Fear to keep the oral resonance shift without the side-effect.
▼ LOW — Oral Resonance __low
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.310 | — | — | — | — |
| 25% | 0.308 | +0.00 | +0.01 | +0.06 | +0.04 |
| 50% | 0.265 | -0.05 | -0.02 | +0.02 | +0.01 |
| 75% | 0.259 | -0.05 | +0.04 | -0.06 | -0.15 |
| 100% ★ | 0.244 | -0.07 | +0.05 | -0.19 | -0.18 |
| 125% | 0.248 | -0.06 | +0.02 | -0.22 | -0.22 |
EmoNet emotion correlation
Contentment+0.41
Contemplation+0.38
Affection+0.28
Fear-0.39
Fatigue/Exhaustion-0.34
Anger-0.33
VoiceNet dimension correlation
Warmth+0.57
Recording Quality+0.57
Mask Resonance+0.50
Tension-0.49
Respiration-0.47
Playful Style-0.46
Listen — 3 samples per dose
25%measured target: 0.35 · 0.32 · 0.23
50%measured target: 0.17 · 0.43 · 0.25
75%measured target: 0.26 · 0.18 · 0.19
100% ★measured target: 0.36 · 0.16 · 0.27
125%measured target: 0.23 · 0.17 · 0.30
Merging Oral Resonance __low lowers the dimension: at the recommended 100% dose the measured Oral Resonance score moves from a 0.31 baseline to 0.24 (a -0.07 shift). The dose-response is clean and monotonic — each step pushes the dimension further. That reach is a real cost — the biggest hit is to burst-blend (-0.19); across the board genuineness +0.05, burst-blend -0.19, speech quality -0.18. Emotionally it reads as Contentment (r 0.41), Contemplation (0.38) and Affection (0.28), and it moves away from Fear, Fatigue/Exhaustion and Anger. On other voice dimensions it co-moves most with Warmth (0.57), Recording Quality (0.57) and Mask Resonance (0.50), while pulling down Tension, Respiration and Playful Style. Practical call: merge __low at 75–100%; expect Contentment and Warmth to rise and burst-blend to fall. If you don’t want Contentment bleeding in, offset it with a prompt cue toward Fear to keep the oral resonance shift without the side-effect.