Convert YouTube Transcript to SRT: Step-by-Step Guide

1
00:00:00,000 --> 00:00:04,500
Welcome to this tutorial on YouTube transcripts

2
00:00:04,500 --> 00:00:08,200
Today we'll cover how to convert them to SRT

3
00:00:08,200 --> 00:00:12,800
This format works with most video editors

HH:MM:SS,mmm
│  │  │  │
│  │  │  └── Milliseconds (000-999)
│  │  └───── Seconds (00-59)
│  └──────── Minutes (00-59)
└─────────── Hours (00+)

1
00:00:00,000 --> 00:00:03,500
Welcome to this video

2
00:00:03,500 --> 00:00:07,200
Today I'll show you something amazing

3
00:00:07,200 --> 00:00:11,000
Let's get started with the first step

# pip
pip install yt-dlp

# Mac
brew install yt-dlp

# Windows
choco install yt-dlp

# Download auto-generated subtitles as SRT
yt-dlp --write-auto-sub --sub-lang en --convert-subs srt --skip-download "VIDEO_URL"

# Download manual subtitles as SRT
yt-dlp --write-sub --sub-lang en --convert-subs srt --skip-download "VIDEO_URL"

# Download both if available
yt-dlp --write-auto-sub --write-sub --sub-lang en --convert-subs srt --skip-download "VIDEO_URL"

# Download SRT for all videos in a playlist
yt-dlp --write-auto-sub --sub-lang en --convert-subs srt --skip-download "PLAYLIST_URL"

import re

def text_to_srt(input_file, output_file):
    """Convert timestamped text to SRT format."""
    with open(input_file, 'r') as f:
        lines = f.readlines()
    
    srt_content = []
    counter = 1
    
    for line in lines:
        # Match pattern: [0:00] or (0:00) or 0:00 - followed by text
        match = re.match(r'[\[\(]?(\d+):(\d+)[\]\)]?\s*[-–]?\s*(.+)', line.strip())
        
        if match:
            minutes, seconds, text = match.groups()
            minutes, seconds = int(minutes), int(seconds)
            
            # Calculate timestamps
            start_total = minutes * 60 + seconds
            end_total = start_total + 4  # Default 4 second duration
            
            start_time = f"{start_total//3600:02d}:{(start_total%3600)//60:02d}:{start_total%60:02d},000"
            end_time = f"{end_total//3600:02d}:{(end_total%3600)//60:02d}:{end_total%60:02d},000"
            
            srt_content.append(f"{counter}")
            srt_content.append(f"{start_time} --> {end_time}")
            srt_content.append(text.strip())
            srt_content.append("")  # Blank line
            counter += 1
    
    with open(output_file, 'w') as f:
        f.write('\n'.join(srt_content))

# Usage
text_to_srt('transcript.txt', 'output.srt')

[sequence number]
[start time] --> [end time]
[text]
[blank line]

1
00:00:00,000 --> 00:00:05,000
This is a very long subtitle that contains too much text and will be difficult for viewers to read in time

1
00:00:00,000 --> 00:00:02,500
This is a very long subtitle
that contains too much text

2
00:00:02,500 --> 00:00:05,000
and will be difficult for viewers
to read in time

1
00:00:00,000 --> 00:00:04,000
<b>Bold text</b>

2
00:00:04,000 --> 00:00:08,000
<i>Italic text</i>

3
00:00:08,000 --> 00:00:12,000
<u>Underlined text</u>

4
00:00:12,000 --> 00:00:16,000
<font color="#FF0000">Red text</font>

# Soft subtitles (can be toggled)
ffmpeg -i video.mp4 -i subtitles.srt -c copy -c:s mov_text output.mp4

# Hard subtitles (burned in)
ffmpeg -i video.mp4 -vf "subtitles=subtitles.srt" output_burned.mp4

# Styled hard subtitles
ffmpeg -i video.mp4 -vf "subtitles=subtitles.srt:force_style='FontSize=24,FontName=Arial'" output_styled.mp4

# Fix common timestamp issues
import re

def fix_srt_timestamps(content):
    # Replace periods with commas in timestamps
    content = re.sub(r'(\d{2}:\d{2}:\d{2})\.(\d{3})', r'\1,\2', content)
    # Add missing hours
    content = re.sub(r'^(\d{2}:\d{2}),(\d{3})', r'00:\1,\2', content, flags=re.MULTILINE)
    return content