← all dimensions · markdown (agent skill) · manual home
Pitch Range VoiceNet dimension guide
Rhythm & Timingcode vn_RANGslot 44
This dimension measures the vertical movement of pitch across the segment. 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.
Pitch Range controls the vertical movement of pitch across the segment. The __high LoRA shifts it +0.76 at 125% (worst cost: burst-blend -0.36); the __low LoRA shifts it -0.04 at 125% (worst cost: burst-blend -0.30). The __high direction is the stronger, more reliable knob here.
▲ HIGH — Pitch Range __high
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.103 | — | — | — | — |
| 25% | 0.226 | +0.12 | -0.03 | -0.09 | +0.02 |
| 50% | 0.287 | +0.18 | -0.04 | -0.14 | +0.07 |
| 75% | 0.665 | +0.56 | -0.04 | -0.34 | -0.14 |
| 100% | 0.841 | +0.74 | -0.09 | -0.39 | -0.18 |
| 125% ★ | 0.860 | +0.76 | -0.08 | -0.36 | -0.23 |
EmoNet emotion correlation
Triumph+0.64
Anger+0.64
Elation+0.62
Contemplation-0.72
Relief-0.61
Contentment-0.52
VoiceNet dimension correlation
Arousal+0.96
Dynamic Arc+0.95
Metallic Character+0.94
Warmth-0.90
Whisper-Talk Style-0.82
Vulnerability-0.82
Listen — 3 samples per dose
25%measured target: 0.47 · 0.62 · 0.09
50%measured target: 0.27 · 0.07 · 0.21
75%measured target: 0.85 · 0.34 · 0.83
100%measured target: 0.84 · 0.75 · 0.86
125% ★measured target: 0.82 · 0.83 · 0.82
Merging Pitch Range __high ramps the dimension: at the recommended 125% dose the measured Pitch Range score moves from a 0.10 baseline to 0.86 (a +0.76 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 (-0.36); across the board genuineness -0.08, burst-blend -0.36, speech quality -0.23. Emotionally it reads as Triumph (r 0.64), Anger (0.64) and Elation (0.62), and it moves away from Contemplation, Relief and Contentment. On other voice dimensions it co-moves most with Arousal (0.96), Dynamic Arc (0.95) and Metallic Character (0.94), while pulling down Warmth, Whisper-Talk Style and Vulnerability. Practical call: merge __high at 100–125%; expect Triumph and Arousal to rise and burst-blend to fall. If you don’t want Triumph bleeding in, offset it with a prompt cue toward Contemplation to keep the pitch range shift without the side-effect.
▼ LOW — Pitch Range __low
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.103 | — | — | — | — |
| 25% | 0.148 | +0.05 | -0.04 | -0.11 | -0.08 |
| 50% | 0.115 | +0.01 | -0.04 | -0.27 | -0.19 |
| 75% | 0.109 | +0.01 | -0.01 | -0.31 | -0.23 |
| 100% | 0.156 | +0.05 | -0.02 | -0.28 | -0.22 |
| 125% ★ | 0.067 | -0.04 | +0.02 | -0.30 | -0.30 |
EmoNet emotion correlation
Elation+0.39
Interest+0.36
Malevolence/Malice+0.35
Fatigue/Exhaustion-0.42
Sadness-0.32
Disappointment-0.27
VoiceNet dimension correlation
Metallic Character+0.80
Arousal+0.78
Tension+0.74
Vulnerability-0.40
Oral Resonance-0.32
Disfluency-0.31
Listen — 3 samples per dose
25%measured target: 0.03 · 0.04 · 0.02
50%measured target: 0.02 · 0.00 · 0.01
75%measured target: 0.31 · 0.00 · 0.28
100%measured target: 0.11 · 0.00 · 0.68
125% ★measured target: 0.00 · 0.45 · 0.00
Merging Pitch Range __low dials back the dimension: at the recommended 125% dose the measured Pitch Range score moves from a 0.10 baseline to 0.07 (a -0.04 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 (-0.30); across the board genuineness +0.02, burst-blend -0.30, speech quality -0.30. Emotionally it reads as Elation (r 0.39), Interest (0.36) and Malevolence/Malice (0.35), and it moves away from Fatigue/Exhaustion, Sadness and Disappointment. On other voice dimensions it co-moves most with Metallic Character (0.80), Arousal (0.78) and Tension (0.74), while pulling down Vulnerability, Oral Resonance and Disfluency. Practical call: merge __low at 100–125%; expect Elation and Metallic Character to rise and burst-blend to fall. If you don’t want Elation bleeding in, offset it with a prompt cue toward Fatigue/Exhaustion to keep the pitch range shift without the side-effect.