# Tension — VoiceNet dimension guide

*Category:* Physical Production · *code:* `vn_TENS` · *slot:* 58

This dimension measures the perceived muscular tightness in throat, jaw and neck. 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.** **Tension** controls the perceived muscular tightness in throat, jaw and neck. The **__high** LoRA shifts it +0.58 at 125% (worst cost: burst-blend -0.32); the **__low** LoRA shifts it -0.08 at 25% (worst cost: genuineness -0.01). The **__high** direction is the stronger, more reliable knob here.

## ▲ HIGH — Tension __high

| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|---|---|---|---|---|
| baseline (0%) | 0.163 | — | — | — | — |
| 25% | 0.245 | +0.08 | -0.02 | -0.02 | -0.02 |
| 50% | 0.588 | +0.42 | +0.00 | -0.14 | -0.17 |
| 75% | 0.582 | +0.42 | -0.03 | -0.31 | -0.13 |
| 100% | 0.654 | +0.49 | -0.01 | -0.33 | -0.28 |
| 125% ★ | 0.742 | +0.58 | -0.03 | -0.32 | -0.29 |

- **EmoNet emotion** — ↑ Anger (+0.74), Sourness (+0.65), Triumph (+0.65); ↓ Contemplation (-0.71), Contentment (-0.71), Affection (-0.62)
- **VoiceNet dimension** — ↑ Dramatic Style (+0.96), Metallic Character (+0.96), Stance (+0.95); ↓ Warmth (-0.80), Whisper-Talk Style (-0.76), Teacher/Didactic Style (-0.74)

Merging **Tension __high** drives the dimension: at the recommended **125%** dose the measured Tension score moves from a 0.16 baseline to 0.74 (a +0.58 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** (-0.32); across the board genuineness -0.03, burst-blend -0.32, speech quality -0.29. Emotionally it reads as Anger (r 0.74), Sourness (0.65) and Triumph (0.65), and it moves away from Contemplation, Contentment and Affection. On other voice dimensions it co-moves most with Dramatic Style (0.96), Metallic Character (0.96) and Stance (0.95), while pulling down Warmth, Whisper-Talk Style and Teacher/Didactic Style. Practical call: merge __high at **100–125%**; expect Anger and Dramatic Style to rise and burst-blend to fall. If you don’t want Anger bleeding in, offset it with a prompt cue toward Contemplation to keep the tension shift without the side-effect.

## ▼ LOW — Tension __low

| dose | target | Δtarget | Δgenu | Δblend | Δqual |
|---|---|---|---|---|---|
| baseline (0%) | 0.163 | — | — | — | — |
| 25% ★ | 0.087 | -0.08 | -0.01 | +0.03 | +0.09 |
| 50% | 0.126 | -0.04 | -0.05 | +0.01 | +0.08 |
| 75% | 0.091 | -0.07 | -0.01 | +0.01 | +0.03 |
| 100% | 0.106 | -0.06 | -0.01 | +0.04 | +0.04 |
| 125% | 0.119 | -0.04 | -0.04 | -0.05 | +0.02 |

- **EmoNet emotion** — ↑ Jealousy & Envy (+0.42), Fear (+0.39), Distress (+0.23); ↓ Contentment (-0.48), Contemplation (-0.43), Concentration (-0.23)
- **VoiceNet dimension** — ↑ Dramatic Style (+0.82), Stance (+0.80), Metallic Character (+0.76); ↓ Teacher/Didactic Style (-0.63), Register (-0.53), ASMR Style (-0.52)

Merging **Tension __low** lowers the dimension: at the recommended **25%** dose the measured Tension score moves from a 0.16 baseline to 0.09 (a -0.08 shift). The dose-response is noisy and non-monotonic — the merge is an unreliable knob here. That reach is nearly free — the biggest hit is to **genuineness** (-0.01); across the board genuineness -0.01, burst-blend +0.03, speech quality +0.09. Emotionally it reads as Jealousy & Envy (r 0.42), Fear (0.39) and Distress (0.23), and it moves away from Contentment, Contemplation and Concentration. On other voice dimensions it co-moves most with Dramatic Style (0.82), Stance (0.80) and Metallic Character (0.76), while pulling down Teacher/Didactic Style, Register and ASMR Style. Practical call: merge __low at **a light 25–50%**; expect Jealousy & Envy and Dramatic Style to rise and genuineness to fall. If you don’t want Jealousy & Envy bleeding in, offset it with a prompt cue toward Contentment to keep the tension shift without the side-effect.

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