# Fullness — VoiceNet dimension guide

*Category:* Spectral & Timbral · *code:* `vn_FULL` · *slot:* 64

This dimension measures the size of the sound’s spectral footprint. Merge the `__high` / `__low` LoRA from `laion/moss-voicenet-dimension-loras` at 25/50/75/100/125%. Costs are shifts vs. the 24-sample neutral baseline.

**TL;DR.** **Fullness** controls the size of the sound’s spectral footprint. The **__high** LoRA shifts it +0.36 at 125% (worst cost: burst-blend -0.36); the **__low** LoRA shifts it -0.27 at 75% (worst cost: burst-blend -0.22). The **__high** direction is the stronger, more reliable knob here.

## ▲ HIGH — Fullness __high

| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|---|---|---|---|---|
| baseline (0%) | 0.564 | — | — | — | — |
| 25% | 0.558 | -0.01 | -0.06 | -0.14 | -0.01 |
| 50% | 0.661 | +0.10 | -0.06 | -0.10 | +0.08 |
| 75% | 0.917 | +0.35 | -0.10 | -0.30 | +0.07 |
| 100% | 0.731 | +0.17 | -0.11 | -0.34 | +0.04 |
| 125% ★ | 0.922 | +0.36 | -0.12 | -0.36 | +0.12 |

- **EmoNet emotion** — ↑ Emotional Numbness (+0.45), Malevolence/Malice (+0.40), Triumph (+0.33); ↓ Relief (-0.50), Contentment (-0.44), Fatigue/Exhaustion (-0.42)
- **VoiceNet dimension** — ↑ Formal Style (+0.83), Newsreader Style (+0.75), Chest Resonance (+0.75); ↓ Casual Style (-0.81), Disfluency (-0.74), Vulnerability (-0.60)

Merging **Fullness __high** pushes the dimension: at the recommended **125%** dose the measured Fullness score moves from a 0.56 baseline to 0.92 (a +0.36 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 -0.12, burst-blend -0.36, speech quality +0.12. Emotionally it reads as Emotional Numbness (r 0.45), Malevolence/Malice (0.40) and Triumph (0.33), and it moves away from Relief, Contentment and Fatigue/Exhaustion. On other voice dimensions it co-moves most with Formal Style (0.83), Newsreader Style (0.75) and Chest Resonance (0.75), while pulling down Casual Style, Disfluency and Vulnerability. Practical call: merge __high at **100–125%**; expect Emotional Numbness and Formal Style to rise and burst-blend to fall. If you don’t want Emotional Numbness bleeding in, offset it with a prompt cue toward Relief to keep the fullness shift without the side-effect.

## ▼ LOW — Fullness __low

| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|---|---|---|---|---|
| baseline (0%) | 0.564 | — | — | — | — |
| 25% | 0.417 | -0.15 | +0.03 | -0.07 | -0.13 |
| 50% | 0.326 | -0.24 | -0.02 | -0.23 | +0.01 |
| 75% ★ | 0.290 | -0.27 | +0.01 | -0.22 | -0.11 |
| 100% | 0.347 | -0.22 | -0.04 | -0.17 | -0.02 |
| 125% | 0.419 | -0.14 | -0.09 | -0.21 | -0.01 |

- **EmoNet emotion** — ↑ Contemplation (+0.51), Awe (+0.43), Infatuation (+0.42); ↓ Pain (-0.34), Elation (-0.34), Impatience and Irritability (-0.33)
- **VoiceNet dimension** — ↑ Warmth (+0.81), Voice Age (+0.79), Whisper-Talk Style (+0.76); ↓ Playful Style (-0.77), Casual Style (-0.69), Volatility (-0.64)

Merging **Fullness __low** dials back the dimension: at the recommended **75%** dose the measured Fullness score moves from a 0.56 baseline to 0.29 (a -0.27 shift). The dose-response is mostly ordered but noisy between doses. That reach is a real cost — the biggest hit is to **burst-blend** (-0.22); across the board genuineness +0.01, burst-blend -0.22, speech quality -0.11. Emotionally it reads as Contemplation (r 0.51), Awe (0.43) and Infatuation (0.42), and it moves away from Pain, Elation and Impatience and Irritability. On other voice dimensions it co-moves most with Warmth (0.81), Voice Age (0.79) and Whisper-Talk Style (0.76), while pulling down Playful Style, Casual Style and Volatility. Practical call: merge __low at **50–100%**; expect Contemplation and Warmth to rise and burst-blend to fall. If you don’t want Contemplation bleeding in, offset it with a prompt cue toward Pain to keep the fullness shift without the side-effect.

---
[← all dimensions](index.md) · [human/audio version](../../site/dims/FULL.html)