Generate animated subtitles in popular creator styles (Mr Beast, Ali Abdaal,...
Copy the command definition below into:
~/.claude/commands/subtitles.md---
name: subtitles
description: Generate animated subtitles in popular creator styles (Mr Beast, Ali Abdaal,...
---
# Subtitle Generator Pro Command
Create professional, animated subtitles in various popular creator styles with automatic transcription and timing.
## Usage
```bash
/subtitles [video-file.mp4] # Auto-generate with default style
/subtitles [video] --style mr-beast # Use Mr Beast style
/subtitles [video] --style ali-abdaal # Use Ali Abdaal style
/subtitles [video] --style hormozi # Use Alex Hormozi style
/subtitles [video] --style minimal # Clean, simple style
/subtitles [video] --language es # Generate Spanish subtitles
/subtitles edit [subtitle-file.srt] # Fine-tune existing subtitles
/subtitles preview [video] [srt] # Preview before burning in
```
## Purpose
Generate subtitles that:
- **Increase retention** - Keep viewers watching longer
- **Improve accessibility** - Reach deaf/hard-of-hearing audience
- **Enable silent viewing** - 85% of social videos watched on mute
- **Boost engagement** - Highlight key words and phrases
- **Match brand style** - Consistent visual identity
## Subtitle Styles
### Mr Beast Style
```
Characteristics:
├─ Every word animated individually
├─ Large, bold text (center frame)
├─ Yellow highlights on important words
├─ High energy, rapid cuts
├─ ALL CAPS for emphasis
├─ Emoji integration 💰🔥⚡
└─ Bouncing/scaling animations
Example:
┌────────────────────────────┐
│ │
│ I SPENT 💰 │
│ $100,000 ON THIS │
│ │
└────────────────────────────┘
[Each word appears with bounce effect]
[Key numbers/words in yellow]
```
### Ali Abdaal Style
```
Characteristics:
├─ Full sentences displayed
├─ Clean, sans-serif font
├─ Bottom center positioning
├─ Key words highlighted in color
├─ Smooth fade in/out
├─ Professional, readable
└─ Subtle emphasis animations
Example:
┌────────────────────────────┐
│ │
│ │
│ │
│ This technique increased │
│ my productivity by 10x │
└────────────────────────────┘
["10x" highlighted in green]
[Smooth, professional transitions]
```
### Alex Hormozi Style
```
Characteristics:
├─ Short, punchy phrases
├─ Bold, impact font
├─ Strategic pauses for emphasis
├─ Red highlights on key points
├─ Minimal but powerful
├─ CAPS for key statements
└─ Vertical center alignment
Example:
┌────────────────────────────┐
│ │
│ THIS IS THE │
│ SECRET 🔑 │
│ │
└────────────────────────────┘
[Quick cuts between phrases]
[Red highlights on power words]
```
### Minimal Style
```
Characteristics:
├─ Simple white text
├─ Black background bar
├─ Bottom third positioning
├─ Easy to read
├─ No animations
├─ Traditional subtitle format
└─ Accessibility-focused
Example:
┌────────────────────────────┐
│ │
│ │
│ │
├────────────────────────────┤
│ Today I'm going to show │
│ you how I made my API 10x │
│ faster with Redis caching. │
└────────────────────────────┘
[Clean, traditional, accessible]
```
### Podcast Style
```
Characteristics:
├─ Speaker names displayed
├─ Color-coded by speaker
├─ Larger text for living room viewing
├─ Bottom center
├─ Sentence-based timing
└─ Smooth, unhurried pacing
Example:
┌────────────────────────────┐
│ │
│ │
│ JEREMY: So the key here │
│ is to implement caching │
│ at the API layer. │
└────────────────────────────┘
[Speaker name in blue]
[Dialogue in white]
```
## Implementation
When user runs `/subtitles redis-tutorial.mp4 --style mr-beast`:
**Step 1: Transcription**
```
🎤 TRANSCRIBING AUDIO
File: redis-tutorial.mp4
Duration: 35:12
Audio quality: Good ✓
Using: OpenAI Whisper Large v3
Language: Auto-detect (English)
Accuracy mode: High
[▓▓▓▓▓▓▓▓▓▓] 100%
Transcription Complete:
├─ Words: 4,247
├─ Accuracy: 98.2%
├─ Speaker diarization: 1 speaker
└─ Processing time: 2m 14s
Transcript preview:
"Hey everyone! Today I'm going to show you how I
made my API ten times faster using Redis caching.
This took me about an hour to implement and the
results were incredible..."
```
**Step 2: Timing Analysis**
```
⏱️ ANALYZING TIMING
Calculating word-level timestamps...
Detecting natural pauses...
Identifying emphasis points...
Timing Statistics:
├─ Average words per second: 2.1
├─ Natural pauses: 247
├─ Speaking pace: Moderate (good for retention)
└─ Emphasis words detected: 89
Subtitle Segments:
├─ Total segments: 342
├─ Average duration: 2.3 seconds
├─ Max segment: 4.1 seconds ✓ (within limits)
└─ Min segment: 0.8 seconds ✓
```
**Step 3: Style Application**
```
🎨 APPLYING MR BEAST STYLE
Style: Mr Beast (high-energy, word-by-word)
Animation: Bounce + scale on entry
Highlighting: Yellow for numbers and power words
Font: Impact (bold, large)
Position: Center screen
Caps: Strategic (emphasis only)
Processing subtitles...
[▓▓▓▓▓▓▓▓▓▓] 100%
Key Words Highlighted (89 instances):
├─ Numbers: 10x, 2000ms, 180ms, 1 hour
├─ Power words: faster, incredible, amazing
├─ Technical: Redis, API, caching
└─ Emphasis: THIS, NOW, SECRET
Animations Applied:
├─ Entry: Bounce + scale (0.3s)
├─ Exit: Quick fade (0.1s)
├─ Emphasis: Yellow highlight + pulse
└─ Timing: Sync with audio peaks
```
**Step 4: Generate Output**
```
✅ SUBTITLES GENERATED
Files Created:
├─ redis-tutorial-subtitles.srt (standard format)
├─ redis-tutorial-subtitles.ass (advanced, with styling)
├─ redis-tutorial-burned.mp4 (video with subtitles burned in)
└─ redis-tutorial-preview.mp4 (10-second preview)
Subtitle Statistics:
├─ Total segments: 342
├─ Highlighted words: 89
├─ File size: 47 KB (SRT), 156 KB (ASS)
└─ Estimated render time: 8 minutes
Preview:
[Shows 10-second clip with animated subtitles]
Quality Checklist:
✓ Text readable at 480p
✓ Timing synced with audio
✓ No text overlapping
✓ Emphasis words highlighted
✓ Animations smooth
✓ Platform-safe colors (no pure white/black)
Ready to burn in with video-editor-ai or export separately
```
## Advanced Features
### Multi-Language Support
Generate subtitles in multiple languages:
```bash
/subtitles tutorial.mp4 --languages en,es,fr,de
```
```
🌍 MULTI-LANGUAGE GENERATION
Source: English (detected)
Target languages: Spanish, French, German
Transcribing...
├─ English: ✓ (98.2% accuracy)
Translating...
├─ Spanish: ✓ (96.8% accuracy)
├─ French: ✓ (97.1% accuracy)
└─ German: ✓ (96.3% accuracy)
Applying timing from English source...
Style: Ali Abdaal (clean, professional)
Generated:
├─ tutorial-en.srt
├─ tutorial-es.srt
├─ tutorial-fr.srt
└─ tutorial-de.srt
Tip: Upload all to YouTube for automatic language switching
```
### Keyword Highlighting
Automatically highlight important words:
```
🎯 KEYWORD DETECTION
Analyzing transcript for emphasis...
Categories:
├─ Numbers: 10x, 2000ms, 180ms, 100% (26 instances)
├─ Technical terms: Redis, API, cache, query (45 instances)
├─ Action words: implement, optimize, improve (18 instances)
└─ Emotion: incredible, amazing, powerful (12 instances)
Highlight Strategy:
├─ Yellow: Numbers and stats
├─ Green: Positive outcomes
├─ Red: Problems/challenges
├─ Blue: Technical terms
└─ White: Standard text
Custom keywords: "Redis", "10x", "faster", "caching"
```
### Speaker Diarization
Identify and label different speakers:
```
👥 SPEAKER DETECTION
Analyzing audio for speaker changes...
Detected Speakers: 2
├─ Speaker 1 (You): 87% of dialogue
├─ Speaker 2 (Guest): 13% of dialogue
└─ Overlap: 2 instances (both speaking)
Assigning colors:
├─ Speaker 1: Blue
├─ Speaker 2: Green
Formatting:
YOU: Hey everyone! Today we're talking about Redis.
GUEST: Yeah, Redis is an amazing tool for caching.
YOU: Exactly! Let me show you how I implemented it.
```
### Auto-Censoring
Automatically detect and censor profanity:
```
🤬 PROFANITY DETECTION
Scanning transcript...
Found: 3 instances
Censoring options:
1. Bleep sound + [BLEEP] text
2. Replace with "****"
3. Mute audio + hide text (2 seconds)
4. No censoring (leave as-is)
Select option (default: 1):
Applied censoring:
├─ 04:23 - [BLEEP] replaced
├─ 12:45 - [BLEEP] replaced
└─ 18:02 - [BLEEP] replaced
Platform safe: YouTube ✓ | TikTok ✓ | LinkedIn ✓
```
### Accessibility Enhancements
Optimize for deaf and hard-of-hearing:
```
♿ ACCESSIBILITY MODE
Enhancements:
├─ Speaker labels always shown
├─ [Sound effect] descriptions added
├─ [Music playing] indicators
├─ Tone/emotion cues [laughs], [sighs]
├─ Non-speech audio described
└─ High contrast (black background, white text)
Example:
[Upbeat music playing]
JEREMY: Hey everyone!
[Keyboard typing sounds]
JEREMY: So here's the code...
[Notification sound]
JEREMY: Oh wow, it works!
[Excited tone]
```
### Subtitle Editing
Fine-tune auto-generated subtitles:
```bash
/subtitles edit redis-tutorial.srt
```
```
📝 SUBTITLE EDITOR
Subtitle 1 (00:00:00 - 00:00:03)
Original: "Hey everyone today I'm going to show you"
Issues: Missing punctuation
✏️ Corrected: "Hey everyone! Today I'm going to show you"
Subtitle 2 (00:00:03 - 00:00:07)
Original: "how I made my API ten times faster"
Issues: "ten" should be "10x" for emphasis
✏️ Corrected: "how I made my API 10x faster"
Common Fixes Applied:
├─ Added punctuation: 34 instances
├─ Fixed capitalization: 12 instances
├─ Corrected numbers: 8 instances (ten → 10)
├─ Split long segments: 5 instances
└─ Merged short segments: 7 instances
Save changes? (y/n)
```
## Platform-Specific Optimization
### YouTube
```
Format: SRT (upload separately)
Position: Bottom center
Font size: Medium (readable at 480p)
Style: Your brand style
Benefits: SEO boost, accessibility, multi-language
```
### TikTok/Instagram Reels
```
Format: Burned-in (ASS)
Position: Center (safe zone)
Font size: Large (mobile screens)
Style: Mr Beast (high energy, every word)
Benefits: Essential (85% watch on mute)
```
### LinkedIn
```
Format: Burned-in (SRT)
Position: Bottom third
Font size: Large (professional)
Style: Ali Abdaal (clean, readable)
Benefits: Professional, 85% watch on mute
```
### Twitter
```
Format: Burned-in (SRT)
Position: Bottom center
Font size: Large
Style: Minimal (high contrast)
Benefits: Feed auto-plays on mute
```
## Subtitle File Formats
### SRT (SubRip)
```
1
00:00:00,000 --> 00:00:03,500
Hey everyone! Today I'm going to
show you how I made my API 10x faster.
2
00:00:03,500 --> 00:00:07,200
The secret? Redis caching.
```
**Use Case**: Standard format, universally supported
### ASS (Advanced SubStation Alpha)
```
[Script Info]
Title: Redis Tutorial
[V4+ Styles]
Style: Default,Arial,48,&H00FFFFFF,&H000000FF,&H00000000,&H00000000,0,0,0,0,100,100,0,0,1,2,2,2,10,10,10,1
[Events]
Dialogue: 0,0:00:00.00,0:00:03.50,Default,,0,0,0,,Hey everyone! Today I'm going to\Nshow you how I made my API {\c&H00FFFF&}10x faster{\r}.
```
**Use Case**: Advanced styling, animations, positioning
### VTT (WebVTT)
```
WEBVTT
00:00.000 --> 00:03.500
Hey everyone! Today I'm going to
show you how I made my API 10x faster.
00:03.500 --> 00:07.200
The secret? Redis caching.
```
**Use Case**: Web-based players, HTML5 video
## Integration Points
Works with other Creator Studio plugins:
- **video-editor-ai**: Burn subtitles into video
- **audio-mixer-assistant**: Use clean audio for better transcription
- **script-to-teleprompter**: Compare script to actual spoken words
- **distribution-automator**: Platform-specific subtitle formats
- **batch-recording-scheduler**: Batch process multiple videos
## Best Practices
### Timing
1. **2-4 seconds per subtitle** - Comfortable reading speed
2. **Sync with natural pauses** - Don't break mid-sentence awkwardly
3. **Never exceed 5 seconds** - Viewers lose interest
4. **Word-by-word for high energy** - Mr Beast style for retention
5. **Full sentences for professional** - Ali Abdaal style for credibility
### Readability
1. **Max 2 lines per subtitle** - More is hard to read
2. **40 characters per line** - Optimal reading width
3. **High contrast colors** - White on black, or black on white
4. **Avoid pure white (#FFFFFF)** - Use off-white (#F0F0F0)
5. **Add text outline** - 2-3px stroke for visibility
### Styling
1. **Match brand identity** - Consistent across videos
2. **Less is more** - Don't over-animate
3. **Test on mobile** - Most viewers on phones
4. **Consider colorblind** - Don't rely only on color
5. **Professional for business** - Clean for LinkedIn
## Troubleshooting
### Poor Transcription Accuracy
```
⚠️ Transcription accuracy: 84.2% (below 95% target)
Common causes:
- Background noise
- Multiple speakers overlapping
- Heavy accent
- Technical jargon
- Poor audio quality
Solutions:
✓ Use audio-mixer-assistant first
✓ Manually review and correct
✓ Provide custom vocabulary list
✓ Use higher quality audio source
```
### Subtitles Out of Sync
```
⚠️ Subtitles drifting out of sync
Detected at: 5:34 (2.3 seconds off)
Causes:
- Variable framerate video
- Audio track separate from video
- Incorrect starting timestamp
Auto-fix:
- Detect drift points
- Adjust timing linearly
- Re-sync to audio peaks
Apply auto-fix? (y/n)
```
### Text Unreadable on Mobile
```
⚠️ Text too small for mobile viewing
Current: 32pt font at 1080p
Recommended: 48pt minimum
Mobile readability test:
- iPhone SE (4.7"): Hard to read ❌
- iPhone 14 (6.1"): Barely readable ⚠️
- iPad Mini (8.3"): Readable ✓
Increase font size to 48pt? (y/n)
```
## Example Workflow
```bash
# 1. Generate subtitles with Mr Beast style
User: /subtitles redis-tutorial.mp4 --style mr-beast
# 2. Preview result
User: /subtitles preview redis-tutorial.mp4 redis-tutorial-subtitles.srt
Claude: [Shows 10-second preview clip]
Looks good? (y/n)
# 3. Generate for multiple platforms
User: /subtitles redis-tutorial.mp4 --platforms youtube,tiktok,linkedin
Claude: ✅ Generated platform-specific versions:
- redis-tutorial-youtube.srt (separate file)
- redis-tutorial-tiktok.mp4 (burned-in, vertical)
- redis-tutorial-linkedin.mp4 (burned-in, clean)
# 4. Multi-language versions
User: /subtitles redis-tutorial.mp4 --languages en,es,fr
Claude: ✅ Generated 3 language versions
Ready for distribution-automator
```
Your goal: Make every video accessible and engaging with professional, on-brand subtitles that boost retention and reach.