← all dimensions · markdown (agent skill) · manual home
Monologue Style VoiceNet dimension guide
Speaking Stylecode vn_S_MONOslot 96
This dimension measures how much the voice sounds like a sustained monologue. 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.
Monologue Style controls how much the voice sounds like a sustained monologue. The __high LoRA shifts it +0.10 at 75% (worst cost: burst-blend -0.23); the __low LoRA shifts it -0.37 at 125% (worst cost: burst-blend -0.31). The __low direction is the stronger, more reliable knob here.
▲ HIGH — Monologue Style __high
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.693 | — | — | — | — |
| 25% | 0.685 | -0.01 | -0.01 | -0.03 | -0.01 |
| 50% | 0.772 | +0.08 | -0.05 | +0.03 | +0.09 |
| 75% ★ | 0.788 | +0.10 | -0.08 | -0.23 | +0.08 |
| 100% | 0.776 | +0.08 | -0.08 | -0.17 | +0.13 |
| 125% | 0.781 | +0.09 | -0.10 | -0.20 | +0.17 |
EmoNet emotion correlation
Contemplation+0.53
Emotional Numbness+0.44
Concentration+0.37
Fatigue/Exhaustion-0.59
Fear-0.56
Relief-0.45
VoiceNet dimension correlation
Whisper-Talk Style+0.79
Recording Quality+0.77
Narrator Style+0.72
Disfluency-0.68
Genuineness-0.64
Casual Style-0.62
Listen — 3 samples per dose
25%measured target: 0.77 · 0.90 · 0.56
50%measured target: 0.87 · 0.72 · 0.85
75% ★measured target: 0.77 · 0.82 · 0.81
100%measured target: 0.82 · 0.78 · 0.80
125%measured target: 0.83 · 0.78 · 0.77
Merging Monologue Style __high lifts the dimension: at the recommended 75% dose the measured Monologue Style score moves from a 0.69 baseline to 0.79 (a +0.10 shift). The dose-response is monotonic but it saturates early, so higher doses add little. That reach is a real cost — the biggest hit is to burst-blend (-0.23); across the board genuineness -0.08, burst-blend -0.23, speech quality +0.08. Emotionally it reads as Contemplation (r 0.53), Emotional Numbness (0.44) and Concentration (0.37), and it moves away from Fatigue/Exhaustion, Fear and Relief. On other voice dimensions it co-moves most with Whisper-Talk Style (0.79), Recording Quality (0.77) and Narrator Style (0.72), while pulling down Disfluency, Genuineness and Casual Style. Practical call: merge __high at 50–100%; expect Contemplation and Whisper-Talk Style to rise and burst-blend to fall. If you don’t want Contemplation bleeding in, offset it with a prompt cue toward Fatigue/Exhaustion to keep the monologue style shift without the side-effect.
▼ LOW — Monologue Style __low
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.693 | — | — | — | — |
| 25% | 0.686 | -0.01 | -0.02 | -0.12 | +0.01 |
| 50% | 0.702 | +0.01 | -0.01 | -0.21 | +0.02 |
| 75% | 0.636 | -0.06 | -0.02 | -0.19 | +0.04 |
| 100% | 0.488 | -0.20 | -0.06 | -0.27 | -0.05 |
| 125% ★ | 0.320 | -0.37 | -0.03 | -0.31 | -0.21 |
EmoNet emotion correlation
Contemplation+0.59
Affection+0.46
Contentment+0.45
Impatience and Irritability-0.58
Anger-0.53
Sourness-0.50
VoiceNet dimension correlation
Whisper-Talk Style+0.82
Warmth+0.79
Narrator Style+0.74
Dynamic Arc-0.77
Ranting/Angry Style-0.77
Arousal-0.68
Listen — 3 samples per dose
25%measured target: 0.87 · 0.44 · 0.67
50%measured target: 0.77 · 0.91 · 0.66
75%measured target: 0.80 · 0.69 · 0.79
100%measured target: 0.73 · 0.36 · 0.19
125% ★measured target: 0.63 · 0.29 · 0.07
Merging Monologue Style __low dials back the dimension: at the recommended 125% dose the measured Monologue Style score moves from a 0.69 baseline to 0.32 (a -0.37 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 (-0.31); across the board genuineness -0.03, burst-blend -0.31, speech quality -0.21. Emotionally it reads as Contemplation (r 0.59), Affection (0.46) and Contentment (0.45), and it moves away from Impatience and Irritability, Anger and Sourness. On other voice dimensions it co-moves most with Whisper-Talk Style (0.82), Warmth (0.79) and Narrator Style (0.74), while pulling down Dynamic Arc, Ranting/Angry Style and Arousal. Practical call: merge __low at 100–125%; expect Contemplation and Whisper-Talk Style to rise and burst-blend to fall. If you don’t want Contemplation bleeding in, offset it with a prompt cue toward Impatience and Irritability to keep the monologue style shift without the side-effect.