SRT Generator

Upload your audio or video file (MP3, WAV, M4A, MP4, MOV, etc.) to VexaScribe (formerly NovaScribe). The AI transcribes the speech and automatically generates word-level timestamps. Review the transcript in the built-in subtitle editor, adjust any timing if needed, then click Download as SRT. The whole process takes 5-10 minutes for a one-hour file.

SRT (SubRip Text) is the most universally supported subtitle file format. It's a plain text file with the .srt extension containing numbered subtitle blocks, each with a start time, end time (HH:MM:SS,mmm), and the subtitle text. Every major video platform supports SRT — YouTube, Vimeo, Twitch, TikTok, Instagram, plus desktop video editors like Premiere Pro, Final Cut Pro, and DaVinci Resolve.

Timestamps are accurate to the millisecond. VexaScribe uses OpenAI's Whisper Large-v3 model to align each word in the transcript to the exact moment it was spoken. For best results, use clear audio with minimal background noise. You can fine-tune timestamps in the built-in editor before downloading the .srt file.

Yes, you get 30 minutes of free transcription with no credit card required — generate SRT files for short videos at no cost. Paid plans start at $2/month for 200 minutes (Starter), $5/month for 1,000 minutes (Basic), $10/month for 2,500 minutes (Pro). All paid plans include SRT export.

SRT (SubRip Text) uses a comma decimal separator (00:00:01,200) and supports basic HTML tags only. VTT (WebVTT, a W3C standard) uses a period decimal separator (00:00:01.200), requires a 'WEBVTT' header line, and supports CSS classes, positioning, and voice tags — making it the format of choice for HTML5 video and HLS streams. VexaScribe exports both formats on every paid plan.

Yes, VexaScribe includes a full subtitle editor. You can correct any text, adjust start and end timestamps, split or merge subtitle entries, and preview the timing against the audio. Changes are saved automatically. When you're satisfied, click Download as SRT to get the final file.

In YouTube Studio, go to Content, select your video, then Subtitles. Click Add Language, choose your language, then Upload File → 'With timing' → select your .srt file. YouTube applies the subtitles immediately. Captioned videos rank better in YouTube search and improve watch time.

VexaScribe accepts MP3, WAV, M4A, FLAC, OGG, AAC, AIFF, WMA, AMR, OPUS for audio, and MP4, MOV, AVI, MKV, WebM, FLV, WMV for video. Files can be up to 5 GB and 10 hours long. Video files have their audio extracted automatically; the video itself is not retained after transcription.

Yes, VexaScribe transcribes in 99 languages with automatic language detection. Upload audio in Spanish, French, German, Japanese, Arabic, or any of the 99 supported languages and download the SRT file in that language. You can also translate the transcript to a different language using the built-in translation feature (133 languages via Google Translate).

The Netflix Timed Text Style Guide recommends a maximum of 17 characters per second (CPS) for adult content and 13 CPS for children's content. The BBC Subtitle Guidelines suggest a conversational reading speed of 160-180 words per minute. VexaScribe's editor flags any cue that exceeds 17 CPS so you can split or shorten it before export.

Per Netflix's Timed Text Style Guide, the minimum duration is 5/6 of a second (approximately 833 ms) and the maximum is 7 seconds per cue. Cues shorter than 833 ms are too quick to read; cues longer than 7 seconds drift out of sync with the on-screen action. VexaScribe automatically respects these bounds when generating cues.

For prerecorded video on most public-facing websites, yes — under WCAG 2.1 Success Criterion 1.2.2 (Level A). U.S. state and local government sites have binding ADA Title II deadlines: April 26, 2027 for entities serving 50,000+ people, and April 26, 2028 for those under 50,000. Higher education and large enterprises are also commonly covered by Section 508 and equivalent international rules.

Encoding mismatch — the file was likely saved as Windows-1252 or another non-UTF-8 encoding. YouTube and most modern players require UTF-8. Re-save the file as UTF-8 (without BOM) using a code editor like VS Code or Notepad++. VexaScribe always exports UTF-8 by default, so this typically only happens after manual edits in older tools.

Yes. SRT is the universal subtitle format supported by all major video editors. Just import the .srt file using your editor's caption or subtitle import workflow. The file will sync to your video timeline using the embedded timestamps.

Open captions are burned directly into the video pixels — always visible, can't be turned off. Closed captions (SRT/VTT) live in a separate file and viewers toggle them on or off. Use open captions for social media silent autoplay, cinema accessibility, and public displays where viewers can't enable captions themselves. Use closed captions for streaming, broadcast TV, YouTube, and any context where regulation (ADA, FCC, EAA) requires user-controlled captions. VexaScribe generates SRT/VTT files; burn them in with FFmpeg, Premiere, or CapCut if you need open captions.

Yes. Two paths: (1) Upload your source audio or video, transcribe in the original language, then translate the transcript into 80+ target languages and export as a new SRT — timing preserved. (2) Upload an existing SRT and the translator preserves cue numbers and timecodes while replacing the dialogue text. Honest note: machine translation of subtitles is fine for personal viewing, draft cuts, and internal review. For broadcast or theatrical release, hire a human subtitler — idioms, proper nouns, and cultural references benefit from human review.

Yes. VexaScribe outputs timestamps in four formats: SRT-style (00:00:01,500), short form (1:30), seconds (90.5), and JSON with per-word millisecond precision. Use these for podcast chapter markers, YouTube chapter timestamps, video editor cue points, or custom downstream tooling. Word-level precision typically lands within ±50 ms on clean studio audio.

An SRT is a video transcript with timecodes formatted for a video player. If you only want the plain spoken text — for show notes, blog repurposing, citations, or feeding into an LLM — export as TXT (plain text), DOCX (Word with speaker labels), or JSON instead. The underlying transcription is identical; only the output format differs.

The original SRT format started as the output of the SubRip software in the late 1990s and has no formal spec — it's a de facto standard. Wikipedia's SubRip article is the most widely cited reference. The format is plain text: numbered cue, timecode line (HH:MM:SS,mmm --> HH:MM:SS,mmm), one or more text lines, blank line. VTT (WebVTT), in contrast, has a formal W3C specification at w3.org/TR/webvtt1/.