netflix-subtitle-processor

name	netflix-subtitle-processor
description	Process Whisper SRT files to Netflix Timed Text specifications. Validates timing, reading speed, character limits. Supports English, Chinese, Japanese, Korean, Spanish.
user_invocable	true

Netflix Subtitle Processor

Post-process Whisper-generated SRT files to meet Netflix Timed Text Style Guide specifications.

Quick Start

# Validate subtitle file (auto-detects language)
python3 ~/.claude/skills/netflix-subtitle-processor/scripts/netflix_subs.py validate video.srt

# Fix issues automatically (keeps all entries)
python3 ~/.claude/skills/netflix-subtitle-processor/scripts/netflix_subs.py fix video.srt video_fixed.srt

# Clean: remove unfixable entries (100% compliance)
python3 ~/.claude/skills/netflix-subtitle-processor/scripts/netflix_subs.py clean video.srt video_clean.srt

# Generate detailed report
python3 ~/.claude/skills/netflix-subtitle-processor/scripts/netflix_subs.py report video.srt

Workflow Integration

This skill complements openai-whisper-guide for professional subtitle production:

Audio/Video → Whisper (transcription) → Netflix Processor (compliance) → Netflix-ready SRT

Recommended Pipeline

# Step 1: Transcribe with Whisper (use openai-whisper-guide)
whisper audio.mp3 --model turbo --output_format srt --max_line_width 42 --max_line_count 2

# Step 2: Validate against Netflix specs
python3 ~/.claude/skills/netflix-subtitle-processor/scripts/netflix_subs.py validate audio.srt

# Step 3: Auto-fix any issues
python3 ~/.claude/skills/netflix-subtitle-processor/scripts/netflix_subs.py fix audio.srt audio_netflix.srt

# Step 4: Verify compliance
python3 ~/.claude/skills/netflix-subtitle-processor/scripts/netflix_subs.py validate audio_netflix.srt

Commands

validate

Check SRT file against Netflix specifications.

python3 netflix_subs.py validate input.srt [--lang en] [--kids] [--json]

Options:

• --lang: Specify language (auto-detected if omitted)
• --kids: Use children's content limits (lower CPS)
• --json: Output results as JSON

Exit codes:

• 0: All entries pass
• 1: Issues found

fix

Auto-fix common issues and write corrected SRT. Keeps all entries.

python3 netflix_subs.py fix input.srt output.srt [--lang en]

Fixes applied:

• Extends subtitles shorter than 833ms minimum
• Splits lines exceeding character limits (smart punctuation breaks for CJK)
• Fixes overlapping subtitles
• Adjusts gaps between subtitles
• Limits to 2 lines per subtitle

clean

Remove unfixable entries, output only valid ones. Use when you need guaranteed compliance.

python3 netflix_subs.py clean input.srt output.srt [--lang en] [--kids]

Behavior:

• Applies all fixes from fix command
• Removes entries that still have issues (e.g., CPS too high)
• Re-indexes remaining entries
• Reports which entries were removed and why

report

Generate detailed validation report.

python3 netflix_subs.py report input.srt [--lang en] [--kids]

Netflix Specifications

Language	Code	Max Chars/Line	Adult CPS	Kids CPS
English	en	42	17	15
Chinese	zh	16	9	7
Japanese	ja	13	4	4
Korean	ko	16	12	10
Spanish	es	42	17	15

Common timing rules (all languages):

• Min duration: 833ms (5/6 second)
• Max duration: 7 seconds
• Min gap between subtitles: 83ms (2 frames @ 24fps)
• Max lines per subtitle: 2

Options

Language Detection

Language is auto-detected from content. Override with --lang:

# Auto-detect (default)
python3 netflix_subs.py validate video.srt

# Force specific language
python3 netflix_subs.py validate video.srt --lang zh
python3 netflix_subs.py validate video.srt --lang ja

Children's Content

Use --kids flag for stricter CPS limits:

python3 netflix_subs.py validate video.srt --kids
python3 netflix_subs.py clean video.srt output.srt --kids

JSON Output

Get machine-readable results:

python3 netflix_subs.py validate video.srt --json

Stdin/Stdout

Use - for piping:

cat video.srt | python3 netflix_subs.py validate -
python3 netflix_subs.py fix video.srt - > output.srt

Troubleshooting

"CPS too high" after fix

Some subtitles have too much text for their duration. Options:

1. Manually shorten the text
2. Use clean command to remove problematic entries
3. Extend the subtitle duration (may cause overlap)

Overlapping subtitles

The fix command automatically resolves overlaps by shortening the previous subtitle. If this causes issues, manually adjust timing.

Chinese/Japanese character counting

Full-width characters (CJK) count as 2 characters. Line breaks for CJK try to split at punctuation (。，！？) before falling back to midpoint.

Remaining issues after fix

The auto-fix handles timing and line breaks. Issues requiring manual attention:

• Content too dense (high CPS with correct timing)
• Complex multi-line reformatting

References

• Netflix Timed Text Style Guide
• Netflix English Subtitle Style Guide
• Netflix Simplified Chinese Style Guide