voice-localization

What this skill does

# AI Voice Localization

> Scale your brand voice across multiple languages using AI voice synthesis, maintaining consistent character and quality for global content.

## When to Use This Skill

- Expanding video content to new language markets
- Creating multilingual courses or training
- Localizing ads and marketing videos
- Dubbing existing content for international audiences
- Building consistent global brand voice
- Deciding between dubbing vs. subtitles

## Methodology Foundation

**Source**: ElevenLabs Multilingual + Global Content Best Practices

**Core Principle**: True localization means the same perceived person speaks each language natively—not a translated voice, but a voice that sounds local while maintaining brand character. AI voice synthesis enables this at scale by preserving voice identity while adapting pronunciation and rhythm to each language.

**Why This Matters**: Global content traditionally required separate voice actors per language, losing brand consistency. AI voice localization maintains the same "person" across 29+ languages, creating unified brand experience worldwide while reducing production costs 70-90%.


## What Claude Does vs What You Decide

| Claude Does | You Decide |
|-------------|------------|
| Structures production workflow | Final creative direction |
| Suggests technical approaches | Equipment and tool choices |
| Creates templates and checklists | Quality standards |
| Identifies best practices | Brand/voice decisions |
| Generates script outlines | Final script approval |

## What This Skill Does

1. **Maintains voice identity across languages** - Same character, different language
2. **Handles cultural adaptation** - Beyond translation to localization
3. **Manages multilingual production** - Efficient workflows for many languages
4. **Ensures quality per market** - Native speaker validation
5. **Calculates ROI** - Traditional dubbing vs. AI localization costs

## How to Use

### Plan Localization Project
```
Help me plan voice localization for [content].
Source language: [original]
Target languages: [list]
Content type: [video/audio/course]
Volume: [duration/number of assets]
```

### Evaluate Localization Approach
```
Should I use AI voice localization or traditional dubbing?
Content: [describe]
Markets: [target countries]
Budget: [range]
Timeline: [deadline]
```

## Instructions

When localizing voice content, follow this methodology:

### Step 1: Assess Localization Needs

Determine the right approach for your content.

```
## Localization Decision Matrix

### When to Use AI Voice Localization

✓ Same brand voice needed across markets
✓ Frequent content updates (efficiency matters)
✓ Educational/informational content
✓ Budget constraints
✓ Quick turnaround needed
✓ 5+ languages needed

### When to Use Traditional Dubbing

✓ Character-driven content (emotions critical)
✓ One-time major production
✓ Markets expect dubbed content (Germany, France)
✓ Complex lip-sync requirements
✓ Budget allows $1,000+ per language

### When to Use Subtitles Instead

✓ Documentary/interview content
✓ Authenticity of original voice matters
✓ Lowest budget option
✓ Markets prefer subtitles (Nordics, Netherlands)
✓ Legal/compliance content (exact words matter)

### Hybrid Approach
Hero content → Traditional dubbing
Supporting content → AI localization
Supplementary → Subtitles
```

---

### Step 2: Select Languages Strategically

Prioritize languages based on market opportunity.

```
## Language Prioritization Framework

### Tier 1: High Volume Languages (1B+ speakers)
| Language | Global Speakers | Key Markets |
|----------|----------------|-------------|
| English | 1.5B | Global |
| Mandarin | 1.1B | China |
| Spanish | 550M | LATAM, Spain |
| Hindi | 600M | India |

### Tier 2: High Value Languages
| Language | Economic Value | Markets |
|----------|---------------|---------|
| German | High GDP | DACH |
| French | Colonial reach | France, Africa |
| Japanese | High spending | Japan |
| Portuguese | Large market | Brazil |

### Tier 3: Strategic Languages
| Language | Strategic Value | Markets |
|----------|----------------|---------|
| Arabic | Growing middle class | MENA |
| Korean | Tech-forward | South Korea |
| Italian | Fashion/luxury | Italy |
| Dutch | High English | Benelux |

### ElevenLabs Supported Languages (29+)
English, Spanish, French, German, Italian, Portuguese,
Polish, Dutch, Hindi, Arabic, Chinese, Japanese, Korean,
Turkish, Swedish, Indonesian, Filipino, Malay, Russian,
Czech, Danish, Finnish, Greek, Romanian, Ukrainian,
Vietnamese, Norwegian, Hungarian, Tamil, and more.
```

---

### Step 3: Prepare Content for Localization

Translation alone isn't enough—prepare for voice adaptation.

```
## Content Preparation Checklist

### Script Adaptation

**Text expansion/contraction**:
| Language | vs English |
|----------|-----------|
| German | +30% longer |
| French | +15-20% longer |
| Spanish | +15-25% longer |
| Chinese | -30% shorter |
| Japanese | Variable |

**Implications**:
- Video may need re-timing
- Allow flexibility in pacing
- Consider sentence splitting for longer languages

**Localization notes to provide**:
□ Brand terms (don't translate, keep English)
□ Product names (pronunciation guide)
□ Numbers (format varies by locale)
□ Dates (format varies by locale)
□ Currency (localize amounts)
□ Cultural references (adapt or explain)

### Voice Consistency Notes

**Preserve across languages**:
- Character/personality
- Energy level
- Authority/warmth balance
- Pace relative to content

**Adapt per language**:
- Natural rhythm and cadence
- Pronunciation of brand terms
- Formal/informal register (varies by culture)
```

---

### Step 4: Production Workflow

Efficient process for multilingual voice production.

```
## Multilingual Production Pipeline

### Phase 1: Source Production
1. Finalize English script
2. Record/generate English voice
3. Lock timing and pacing
4. Create master video/audio

### Phase 2: Translation
1. Professional translation (not machine)
2. Localization review (cultural adaptation)
3. Timing adaptation (fit original duration)
4. Brand term glossary enforcement

### Phase 3: Voice Generation

**Per language**:
```
1. Load translated script
2. Apply same voice settings as source
3. Generate voice in target language
4. Check pronunciation of brand terms
5. Adjust pacing if needed
6. Review for naturalness
```

### Phase 4: Quality Control

**Native speaker review checklist**:
□ Natural pronunciation
□ Correct emphasis and intonation
□ Brand terms handled correctly
□ No awkward phrasing
□ Appropriate formality level
□ Cultural appropriateness

### Phase 5: Integration
1. Replace audio track in video
2. Re-sync if timing changed
3. Update text overlays
4. Localize captions/subtitles
5. Final review per language
```

---

### Step 5: Quality Assurance

Ensure each language meets standards.

```
## Localization QA Framework

### Technical QA
□ Audio levels consistent across languages
□ No clipping or distortion
□ Background music balanced correctly
□ Transitions smooth
□ Sync with video acceptable

### Linguistic QA
□ Translation accuracy (spot check 10%)
□ Natural flow and rhythm
□ Brand voice maintained
□ Technical terms correct
□ No machine-translation artifacts

### Cultural QA
□ No offensive content for market
□ References appropriate
□ Humor/idioms adapted correctly
□ Visual content appropriate
□ Call-to-action localized

### Native Speaker Sign-Off
For each language:
- [ ] Spanish (Reviewer: _____) ☐ Approved
- [ ] French (Reviewer: _____) ☐ Approved
- [ ] German (Reviewer: _____) ☐ Approved
- [ ] [Add languages...]
```

---

### Step 6: Calculate ROI

Compare AI localization to traditional approaches.

```
## Localization Cost Comparison

### Traditional Dubbing (per language)

| Component | Cost |
|-----------|------|
| Translation | $0.15/word |
| Voice talent | $300-1,000/hour finished |
| Studio time | $100-200/hour |
| Direction | $50-100/hour |
| Engineering