Subtitle Editor

AudioSRT is a free browser-based subtitle editor that handles the full subtitle workflow — from AI transcription to polished export — without uploading your files to any server. It accepts SRT, VTT, or any audio/video file; transcribes with OpenAI Whisper locally; and lets you fix timing, split lines, apply custom fonts and styling, and export to SRT, VTT, Premiere Pro JSON, or burned MP4. No account required; cloud AI transcription with 100+ language support is available on Starter.

Subtitle editing is the process of reviewing, correcting, and formatting AI-generated or imported subtitle files so they meet professional standards before publishing. It covers three areas: accuracy (are the words correct?), timing (do subtitles appear and disappear at the right moment?), and readability (are lines the right length, with the right breaks?).

Why subtitle editing matters

AI transcription — even at 95% accuracy — means errors in a 10-minute video. More importantly, raw AI output doesn't respect subtitle best practices: lines can be 80 characters long, breaks happen in the middle of phrases, and timing may not sync precisely with speech. A viewer watching a professionally edited subtitle track versus raw AI output notices the difference immediately, even if they can't articulate why.

For accessibility compliance (WCAG 2.1, ADA), subtitle files need to be accurate, properly timed, and readable. For SEO, YouTube indexes caption content — clean subtitles mean better indexing. For engagement, well-formatted captions keep viewers watching longer.

The 42-character line length rule

The broadcast and streaming industry standard is a maximum of 42 characters per line, with no more than 2 lines per subtitle. This is based on the minimum screen size a subtitle might be viewed at — a mobile phone in landscape mode. Lines longer than 42 characters may be cropped or wrapped in a way that breaks readability.

In practice, aim for 35–40 characters per line when possible. Short lines (under 20 characters) are fine — they read faster and feel cleaner. The 42-character limit is a ceiling, not a target. AudioSRT shows a character count for each subtitle line so you can monitor this as you edit.

How to fix subtitle timing

Timing issues fall into three categories:

• Start time too early or late. The subtitle appears before the speaker starts or after they've already said the word. Click the timestamp and type the correct value, or use the waveform timeline to drag the start point to the audio peak.
• End time too short. The subtitle disappears before the viewer has time to read it. A good rule: minimum 1.5 seconds, maximum 6 seconds per subtitle. Calculate reading time as approximately 150–200 words per minute.
• Gap between subtitles too small. A gap of less than 2 frames between subtitles causes them to flash. AudioSRT automatically adds a minimum gap when you adjust timing.

Split vs merge — when to use each

Split when a subtitle is too long (over 42 chars per line) or covers too much time (over 6 seconds). Splitting at a natural phrase boundary — after a comma, before a conjunction like "and" or "but" — preserves meaning while improving readability.

Merge when two adjacent subtitles are too short (under 1.5 seconds each) and cover related content. A single subtitle covering one complete thought reads better than two fragments. Click the down arrow on a subtitle's last word to move it into the next subtitle.

Typography best practices for video

Subtitle typography affects how readable your subtitles are on video. Key considerations:

• Font choice: Sans-serif fonts (Inter, Helvetica, Arial) are more readable at small sizes than serif fonts. Avoid decorative fonts for body subtitles.
• Size: For 1080p video, a font size of 40–52px works for most contexts. For vertical video (9:16), go larger — 48–60px.
• Contrast: White text on a dark semi-transparent background is the most readable combination. A drop shadow or stroke helps on light backgrounds.
• Position: Subtitles conventionally sit at the bottom of the frame. For vertical video on social platforms, consider placing them higher to avoid being covered by UI overlays.

Single line vs double line subtitles

Single-line subtitles (one line per entry) are cleaner and feel faster-paced. They work well for YouTube and fast-cut content. Double-line subtitles (two lines per entry) are standard in broadcast and streaming — they allow longer phrases to stay together while still fitting the screen. AudioSRT supports both modes in the preview panel.

Export format guide

• SRT — universal format, accepted by YouTube, Vimeo, Premiere Pro, DaVinci Resolve, Final Cut Pro, VLC. Best for broad compatibility.
• VTT — web format with slightly more styling support than SRT. Required for HTML5 video players and some web platforms.
• Premiere Pro JSON — native Premiere format that creates editable caption tracks with your styling settings. Use this if you're finishing in Premiere.
• Burned MP4 — permanently embeds subtitles into the video. Use for social media where separate caption tracks aren't supported.

Accessibility considerations

For WCAG 2.1 AA compliance, subtitle files must: identify speakers when not visually apparent, include non-speech audio information relevant to the content (e.g., [applause], [music]), and be accurate to the spoken content. AudioSRT's editor lets you add speaker labels and non-speech annotations directly in the subtitle text before export.