← manual home · markdown (agent skill)
For each of the 57 VoiceNet voice-performance dimensions, the effect of merging its __high and __low LoRA into moss-tts-4.55b-voice-acting-v2 across a 5-step scaling sweep (25 / 50 / 75 / 100 / 125%). Each dimension page shows the dose-response of the target score, the cost in genuineness / vocal-burst blend / speech quality, the strongest ± side-effects on EmoNet emotions and other VoiceNet dimensions, three audio samples per dose, and a grounded prompting tip. LoRAs: laion/moss-voicenet-dimension-loras. Cross-dimension audio overview: scaling grid.
| dimension | category | high Δ@best | high dose | high worst-cost | low Δ@best | low dose | low worst-cost |
|---|---|---|---|---|---|---|---|
Arousal AROU | Emotion & Affect | +0.73 | 100% | -0.43 | -0.09 | 50% | -0.23 |
Valence VALN | Emotion & Affect | +0.24 | 125% | -0.29 | -0.01 | 25% | -0.06 |
Volatility VOLT | Emotion & Affect | +0.28 | 125% | -0.37 | -0.15 | 75% | -0.12 |
Arousal Shift ARSH | Temporal Dynamics | +0.43 | 100% | -0.35 | -0.14 | 125% | -0.34 |
Dynamic Arc DARC | Temporal Dynamics | +0.63 | 125% | -0.42 | -0.02 | 25% | -0.01 |
Valence Shift VALS | Temporal Dynamics | +0.30 | 100% | -0.39 | -0.07 | 125% | -0.17 |
Velocity Flux VFLX | Temporal Dynamics | +0.42 | 125% | -0.41 | -0.01 | 100% | -0.14 |
Attack ATCK | Physical Production | +0.61 | 125% | -0.36 | -0.02 | 50% | -0.10 |
Cognitive Load COGL | Physical Production | +0.08 | 100% | -0.11 | -0.17 | 100% | -0.33 |
Respiration RESP | Physical Production | +0.16 | 100% | -0.29 | -0.24 | 125% | -0.36 |
Tension TENS | Physical Production | +0.58 | 125% | -0.32 | -0.08 | 25% | -0.01 |
Articulation Clarity CLRT | Rhythm & Timing | +0.34 | 125% | -0.25 | -0.19 | 75% | -0.37 |
Chunking CHNK | Rhythm & Timing | +0.28 | 125% | -0.22 | +0.00 | 50% | -0.15 |
Disfluency DFLU | Rhythm & Timing | +0.06 | 125% | -0.19 | -0.24 | 125% | -0.31 |
Emphasis EMPH | Rhythm & Timing | +0.59 | 100% | -0.30 | -0.08 | 100% | -0.15 |
Pitch Range RANG | Rhythm & Timing | +0.76 | 125% | -0.36 | -0.04 | 125% | -0.30 |
Smoothness SMTH | Rhythm & Timing | +0.21 | 125% | -0.09 | -0.16 | 125% | -0.27 |
Structure STRU | Rhythm & Timing | +0.22 | 125% | -0.05 | -0.18 | 100% | -0.27 |
Tempo TEMP | Rhythm & Timing | +0.54 | 125% | -0.33 | -0.07 | 75% | -0.14 |
Brightness BRGT | Spectral & Timbral | +0.30 | 100% | -0.20 | -0.04 | 50% | -0.07 |
Esthetics ESTH | Spectral & Timbral | +0.10 | 125% | -0.09 | -0.25 | 100% | -0.31 |
Fullness FULL | Spectral & Timbral | +0.36 | 125% | -0.36 | -0.27 | 75% | -0.22 |
Harmonicity HARM | Spectral & Timbral | +0.33 | 125% | -0.40 | -0.09 | 125% | -0.21 |
Metallic Character METL | Spectral & Timbral | +0.43 | 100% | -0.44 | -0.04 | 75% | -0.03 |
Roughness ROUG | Spectral & Timbral | +0.31 | 125% | -0.41 | -0.10 | 50% | -0.03 |
Warmth WARM | Spectral & Timbral | +0.16 | 125% | -0.13 | -0.57 | 125% | -0.33 |
Focus FOCS | Social & Interpersonal | +0.35 | 125% | -0.30 | -0.22 | 125% | -0.24 |
Stance STNC | Social & Interpersonal | +0.58 | 125% | -0.38 | -0.12 | 125% | -0.12 |
Vulnerability VULN | Social & Interpersonal | +0.05 | 125% | -0.26 | -0.40 | 125% | -0.36 |
Perceived Gender GEND | Speaker Identity | +0.14 | 125% | -0.07 | -0.51 | 125% | -0.30 |
Register REGS | Speaker Identity | +0.61 | 125% | -0.23 | -0.21 | 100% | -0.17 |
Voice Age AGEV | Speaker Identity | +0.21 | 100% | -0.18 | -0.51 | 125% | -0.28 |
Chest Resonance R_CHST | Resonance Placement | +0.49 | 125% | -0.35 | -0.29 | 50% | -0.14 |
Head Resonance R_HEAD | Resonance Placement | +0.42 | 125% | -0.41 | -0.02 | 25% | -0.06 |
Mask Resonance R_MASK | Resonance Placement | +0.39 | 125% | -0.24 | -0.09 | 50% | -0.15 |
Mixed Resonance R_MIXD | Resonance Placement | +0.04 | 125% | -0.26 | +0.00 | 25% | -0.11 |
Nasal Resonance R_NASL | Resonance Placement | +0.56 | 100% | -0.41 | -0.08 | 25% | +0.01 |
Oral Resonance R_ORAL | Resonance Placement | +0.18 | 100% | -0.39 | -0.07 | 100% | -0.19 |
Throat Resonance R_THRT | Resonance Placement | +0.56 | 125% | -0.40 | -0.12 | 75% | -0.15 |
ASMR Style S_ASMR | Speaking Style | +0.40 | 125% | -0.06 | -0.32 | 125% | -0.34 |
Authoritative Style S_AUTH | Speaking Style | +0.63 | 125% | -0.46 | -0.01 | 25% | -0.08 |
Cartoonish Style S_CART | Speaking Style | +0.40 | 125% | -0.32 | -0.15 | 50% | -0.04 |
Casual Style S_CASU | Speaking Style | +0.20 | 125% | -0.24 | -0.26 | 125% | -0.15 |
Conversational Style S_CONV | Speaking Style | +0.12 | 125% | -0.15 | -0.03 | 125% | -0.32 |
Dramatic Style S_DRAM | Speaking Style | +0.54 | 100% | -0.38 | -0.16 | 125% | -0.14 |
Formal Style S_FORM | Speaking Style | +0.44 | 125% | -0.23 | -0.10 | 100% | -0.35 |
Monologue Style S_MONO | Speaking Style | +0.10 | 75% | -0.23 | -0.37 | 125% | -0.31 |
Narrator Style S_NARR | Speaking Style | +0.39 | 125% | -0.25 | -0.06 | 125% | -0.34 |
Newsreader Style S_NEWS | Speaking Style | +0.06 | 125% | -0.04 | +0.01 | 50% | -0.16 |
Playful Style S_PLAY | Speaking Style | +0.61 | 125% | -0.40 | -0.04 | 75% | -0.07 |
Ranting/Angry Style S_RANT | Speaking Style | +0.75 | 125% | -0.40 | -0.02 | 25% | -0.02 |
Storytelling Style S_STRY | Speaking Style | +0.26 | 100% | -0.27 | -0.26 | 125% | -0.29 |
Teacher/Didactic Style S_TECH | Speaking Style | +0.25 | 125% | -0.30 | -0.08 | 25% | -0.07 |
Whisper-Talk Style S_WHIS | Speaking Style | +0.27 | 100% | -0.13 | -0.45 | 100% | -0.34 |
Background Noise BKGN | Language & Recording | +0.08 | 100% | -0.04 | -0.09 | 50% | -0.15 |
Content Appropriateness (3-point Scale) EXPL | Language & Recording | +0.11 | 125% | -0.33 | -0.03 | 50% | -0.08 |
Recording Quality RCQL | Language & Recording | +0.14 | 125% | -0.16 | -0.30 | 125% | -0.31 |