Can I make subtitles for a recording that contains background music?
Yes, but review the transcript carefully. Upload the recording, choose the spoken language, correct names and timestamps, then export SRT. This example supplies both clean and mixed audio so you can compare recognition under one controlled change.
Source and method
The same original English script is synthesized with eSpeak NG for both inputs. A reproducible tone sequence is mixed into one copy at a documented gain. Whisper small processes both; the raw-caption video uses the mixed recording.
These case studies use original scripts, synthetic speech, generated tones and a generated notice image. They contain no user uploads. Source media and recognized text keep their original language when the interface language changes.
Download original input ↓ · Download settings ↓ · Download source text ↓
Compare the source, recognition and review
Source text
Welcome to Field Notes.
Today Maya explains how a short recording becomes a captioned video.
Check the name Audiosubvideo, the number twenty seven, and the pause between sentences.
Review the words before you export your clip.Recognition from the clean audio
Welcome to Field Notes.
Today my rec explains how a short
recording becomes a captioned
video.
Check the name of the sub -video,
the number 27,
and the pause between sentences.
Review the words before you export
your clip.Download clean audio ↓ · Download clean-audio recognition ↓
Unedited recognition result
Welcome to Field Notes.
Today my rec explains how a short
recording becomes a captioned
video.
Check the name of the sub -video,
the number 27,
and the pause between sentences.
Renew the words before you export
your clip.Reviewed result
Welcome to Field Notes.
Today Maya explains how a short recording becomes a captioned video.
Check the name Audiosubvideo, the number twenty seven, and the pause between sentences.
Review the words before you export your clip.The reviewed version uses the original script with evenly timed sentence boundaries. It corrects the text, but does not establish word-level timing accuracy. The music remains in the output; this workflow does not perform vocal separation or noise removal.
Download video with unedited captions ↓ · Download reviewed video ↓
What to check before sharing
Both runs misrecognized Maya and Audiosubvideo. The clean recording preserved Review, while the mixed recording produced Renew. Both converted twenty seven to 27. This one controlled synthetic comparison does not establish general accuracy or isolate every cause of an error.
Measurement and limitations
Times describe a single build run, not a service-level promise. Audio cases are measured on the build computer; image OCR is measured in the Linux deployment image. Upload time and the shared processing queue are excluded.
- Recognition time
- 3.98 s
- Render time
- 2.58 s
- Processing engine
- faster-whisper small / CPU int8
This is a product demonstration, not an accuracy benchmark or customer testimonial.
Product facts
AI Free Audio to Video · audiosubvideo.com
- Price and account
- Free to use without payment or sign-up; shared queue and hourly usage limits apply.
- Speech and image recognition
- Whisper small for speech and Tesseract for images on the production server. No speaker identification, transcript translation or vocal separation.
- Downloads
- MP4 for videos; TXT, JSON and SRT for timed transcripts; TXT and JSON for image OCR.
- Who can access results
- The originating browser session and signed-in site administrators. Uploaded content is not a public case study.
- Audio and video limits
- 100 MB · 10 min
- Completed file retention
- 24 hours