← all dimensions · markdown (agent skill) · manual home
Dramatic Style VoiceNet dimension guide
Speaking Stylecode vn_S_DRAMslot 85
This dimension measures how dramatic and heightened the delivery sounds. 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.
Dramatic Style controls how dramatic and heightened the delivery sounds. The __high LoRA shifts it +0.54 at 100% (worst cost: burst-blend -0.38); the __low LoRA shifts it -0.16 at 125% (worst cost: burst-blend -0.14). The __high direction is the stronger, more reliable knob here.
▲ HIGH — Dramatic Style __high
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.386 | — | — | — | — |
| 25% | 0.441 | +0.05 | -0.06 | -0.13 | +0.08 |
| 50% | 0.492 | +0.11 | -0.05 | -0.06 | +0.04 |
| 75% | 0.766 | +0.38 | -0.10 | -0.38 | -0.14 |
| 100% ★ | 0.930 | +0.54 | -0.09 | -0.38 | -0.12 |
| 125% | 0.914 | +0.53 | -0.07 | -0.33 | -0.20 |
EmoNet emotion correlation
Elation+0.66
Anger+0.53
Sourness+0.47
Contentment-0.71
Contemplation-0.70
Relief-0.64
VoiceNet dimension correlation
Tension+0.97
Attack+0.94
Metallic Character+0.93
Warmth-0.87
Whisper-Talk Style-0.80
Vulnerability-0.76
Listen — 3 samples per dose
25%measured target: 0.61 · 0.30 · 0.28
50%measured target: 0.57 · 0.63 · 0.68
75%measured target: 0.83 · 0.66 · 0.58
100% ★measured target: 0.80 · 0.99 · 0.85
125%measured target: 0.89 · 0.98 · 0.77
Merging Dramatic Style __high pushes the dimension: at the recommended 100% dose the measured Dramatic Style score moves from a 0.39 baseline to 0.93 (a +0.54 shift). The dose-response is clean and monotonic — each step pushes the dimension further. That reach is a steep cost — the biggest hit is to burst-blend (-0.38); across the board genuineness -0.09, burst-blend -0.38, speech quality -0.12. Emotionally it reads as Elation (r 0.66), Anger (0.53) and Sourness (0.47), and it moves away from Contentment, Contemplation and Relief. On other voice dimensions it co-moves most with Tension (0.97), Attack (0.94) and Metallic Character (0.93), while pulling down Warmth, Whisper-Talk Style and Vulnerability. Practical call: merge __high at 75–100%; expect Elation and Tension to rise and burst-blend to fall. If you don’t want Elation bleeding in, offset it with a prompt cue toward Contentment to keep the dramatic style shift without the side-effect.
▼ LOW — Dramatic Style __low
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.386 | — | — | — | — |
| 25% | 0.377 | -0.01 | -0.05 | -0.03 | +0.09 |
| 50% | 0.258 | -0.13 | -0.05 | -0.13 | +0.10 |
| 75% | 0.241 | -0.14 | -0.04 | -0.06 | +0.04 |
| 100% | 0.222 | -0.16 | -0.06 | -0.12 | +0.11 |
| 125% ★ | 0.221 | -0.16 | +0.01 | -0.14 | -0.04 |
EmoNet emotion correlation
Malevolence/Malice+0.42
Jealousy & Envy+0.41
Interest+0.38
Concentration-0.44
Emotional Numbness-0.42
Contentment-0.33
VoiceNet dimension correlation
Tension+0.75
Emphasis+0.71
Volatility+0.67
Teacher/Didactic Style-0.66
Nasal Resonance-0.63
Mixed Resonance-0.63
Listen — 3 samples per dose
25%measured target: 0.44 · 0.25 · 0.28
50%measured target: 0.15 · 0.23 · 0.56
75%measured target: 0.34 · 0.17 · 0.31
100%measured target: 0.09 · 0.27 · 0.22
125% ★measured target: 0.25 · 0.26 · 0.09
Merging Dramatic Style __low dials back the dimension: at the recommended 125% dose the measured Dramatic Style score moves from a 0.39 baseline to 0.22 (a -0.16 shift). The dose-response is clean and monotonic — each step pushes the dimension further. That reach is a modest cost — the biggest hit is to burst-blend (-0.14); across the board genuineness +0.01, burst-blend -0.14, speech quality -0.04. Emotionally it reads as Malevolence/Malice (r 0.42), Jealousy & Envy (0.41) and Interest (0.38), and it moves away from Concentration, Emotional Numbness and Contentment. On other voice dimensions it co-moves most with Tension (0.75), Emphasis (0.71) and Volatility (0.67), while pulling down Teacher/Didactic Style, Nasal Resonance and Mixed Resonance. Practical call: merge __low at 100–125%; expect Malevolence/Malice and Tension to rise and burst-blend to fall. If you don’t want Malevolence/Malice bleeding in, offset it with a prompt cue toward Concentration to keep the dramatic style shift without the side-effect.