← all dimensions · markdown (agent skill) · manual home
Mixed Resonance VoiceNet dimension guide
Resonance Placementcode vn_R_MIXDslot 81
This dimension measures balanced resonance across multiple cavities. 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.
Mixed Resonance controls balanced resonance across multiple cavities. The __high LoRA shifts it +0.04 at 125% (worst cost: burst-blend -0.26); the __low LoRA shifts it +0.00 at 25% (worst cost: burst-blend -0.11). The __high direction is the stronger, more reliable knob here.
▲ HIGH — Mixed Resonance __high
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.426 | — | — | — | — |
| 25% | 0.424 | +0.00 | -0.03 | +0.07 | +0.02 |
| 50% | 0.429 | +0.00 | -0.04 | -0.03 | +0.05 |
| 75% | 0.436 | +0.01 | -0.02 | -0.05 | +0.04 |
| 100% | 0.444 | +0.02 | -0.05 | -0.05 | +0.05 |
| 125% ★ | 0.462 | +0.04 | -0.12 | -0.26 | +0.08 |
EmoNet emotion correlation
Concentration+0.39
Emotional Numbness+0.32
Contentment+0.24
Longing-0.46
Sadness-0.39
Helplessness-0.38
VoiceNet dimension correlation
Velocity Flux+0.68
Head Resonance+0.65
Teacher/Didactic Style+0.61
Genuineness-0.43
Cognitive Load-0.40
Disfluency-0.38
Listen — 3 samples per dose
25%measured target: 0.42 · 0.39 · 0.44
50%measured target: 0.47 · 0.40 · 0.40
75%measured target: 0.41 · 0.46 · 0.42
100%measured target: 0.47 · 0.45 · 0.41
125% ★measured target: 0.48 · 0.46 · 0.46
Merging Mixed Resonance __high ramps the dimension: at the recommended 125% dose the measured Mixed Resonance score moves from a 0.43 baseline to 0.46 (a +0.04 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 (-0.26); across the board genuineness -0.12, burst-blend -0.26, speech quality +0.08. Emotionally it reads as Concentration (r 0.39), Emotional Numbness (0.32) and Contentment (0.24), and it moves away from Longing, Sadness and Helplessness. On other voice dimensions it co-moves most with Velocity Flux (0.68), Head Resonance (0.65) and Teacher/Didactic Style (0.61), while pulling down Genuineness, Cognitive Load and Disfluency. Practical call: merge __high at 100–125%; expect Concentration and Velocity Flux to rise and burst-blend to fall. If you don’t want Concentration bleeding in, offset it with a prompt cue toward Longing to keep the mixed resonance shift without the side-effect.
▼ LOW — Mixed Resonance __low
Dose-response
| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|
| baseline (0%) | 0.426 | — | — | — | — |
| 25% ★ | 0.422 | +0.00 | +0.03 | -0.11 | -0.10 |
| 50% | 0.447 | +0.02 | -0.02 | -0.17 | +0.00 |
| 75% | 0.450 | +0.02 | -0.08 | -0.25 | +0.04 |
| 100% | 0.466 | +0.04 | -0.10 | -0.39 | +0.05 |
| 125% | 0.461 | +0.03 | -0.10 | -0.39 | -0.07 |
EmoNet emotion correlation
Teasing+0.36
Concentration+0.34
Hope/Enthusiasm/Optimism+0.28
Fatigue/Exhaustion-0.52
Distress-0.40
Pain-0.38
VoiceNet dimension correlation
Velocity Flux+0.73
Valence+0.63
Head Resonance+0.63
Cognitive Load-0.69
Genuineness-0.57
Disfluency-0.55
Listen — 3 samples per dose
25% ★measured target: 0.37 · 0.45 · 0.40
50%measured target: 0.50 · 0.47 · 0.41
75%measured target: 0.47 · 0.47 · 0.46
100%measured target: 0.49 · 0.49 · 0.47
125%measured target: 0.46 · 0.46 · 0.44
Merging Mixed Resonance __low damps the dimension: at the recommended 25% dose the measured Mixed Resonance score moves from a 0.43 baseline to 0.42 (a +0.00 shift). The dose-response is mostly ordered but noisy between doses. That reach is a modest cost — the biggest hit is to burst-blend (-0.11); across the board genuineness +0.03, burst-blend -0.11, speech quality -0.10. Emotionally it reads as Teasing (r 0.36), Concentration (0.34) and Hope/Enthusiasm/Optimism (0.28), and it moves away from Fatigue/Exhaustion, Distress and Pain. On other voice dimensions it co-moves most with Velocity Flux (0.73), Valence (0.63) and Head Resonance (0.63), while pulling down Cognitive Load, Genuineness and Disfluency. Practical call: merge __low at a light 25–50%; expect Teasing and Velocity Flux to rise and burst-blend to fall. If you don’t want Teasing bleeding in, offset it with a prompt cue toward Fatigue/Exhaustion to keep the mixed resonance shift without the side-effect.