Recording checks, cleanup and transcription
AUDIO TO TEXT.REVIEW EVERY WORD.
Inspect channels and loudness locally. Choose noise reduction or normalization when needed, compare the result, then transcribe your recording and review the words, timestamps and optional speaker labels.
Stop fighting
your recordings.
Free Tools
Tools for your recording
Measure and inspect locally. Use cloud processing when you need cleanup or a transcript.
Audio to Text
Transcribe a recording, review the text and export captions.
MP3 to Text
Turn MP3 recordings into editable text with timestamps.
SRT Generator
Generate subtitles from speech and review the timing before export.
Audio Normalizer
Even out quiet and loud passages to a consistent loudness.
Remove Background Noise
Reduce steady noise and compare the result with the original.
Mic Test
Check your microphone in the browser before you record.
LUFS Meter
Measure loudness and peaks locally, without uploading audio.
Channel Inspector
Find silent channels and potential phase cancellation.
Subtitle Validator
Catch malformed timing, empty cues and caption overlaps.
Browse all audio toolsCloud processing is also available through the API. Read the docs
How it works
Check. Compare. Review.
Check the recording
Inspect channels and loudness locally, or choose a supported recording in the transcript workbench.
Choose cleanup
Use noise reduction or normalization if needed. Listen to the processed audio and original before choosing what to transcribe.
Transcribe and review
Check words and caption timing against the recording. Request optional speaker labels and correct them where needed.
Export
Download reviewed text or captions as TXT, SRT, VTT or DOCX. Cleanup tools offer separate processed-audio downloads.
Use Cases
Who it's for
Podcasters
Clean up remote guest audio and get timestamped transcripts for show notes and chapters.
Journalists
Review interview transcripts and check quoted words against the original recording.
Students
Turn spoken lectures into editable text, then check names, terms and unclear passages.
Researchers
Transcribe interviews, review optional speaker labels and export text for analysis.
Voice-memo users
Clean up quick voice memos and get readable text you can drop into any doc.
FAQ
Public tools do not require an account or credit card. Local checks run on your device. Cloud transcription and cleanup have recording-duration, upload and shared-network processing limits. Sign in for saved transcripts and account allowances.
Local tools do not upload your files. Cloud tools store audio on SoundHalo and send recordings to speech providers when transcribing. Audio access expires after 30 days, but expiry does not confirm storage deletion. Use Delete stored audio before leaving the tool. Account transcripts and provider copies have separate handling. Read /privacy for details.
Not always. Inspect the recording first. If you choose noise reduction or normalization, listen to the original and result before transcribing. Cleanup can change speech details and does not guarantee better recognition. Caption times follow the recording you upload, so finish timeline edits before generating captions.
The transcript workbench offers automatic language detection and several language choices. Results depend on the recording and speech provider. Review words and timestamps before sharing, and request speaker labels explicitly when you need them. Labels are not verified identities.
Browser uploads are capped at 100 MB. Public cloud processing accepts recordings up to 30 minutes and uses shared-network daily allowances. A displayed allowance is a snapshot, not a reservation. Signed-in plans have separate file-count and processing allowances; Pro supports recordings up to ten hours. Local tools show their own size and duration limits.
Yes. Cloud audio operations are available through the HTTP API. Browser-local checks and converters do not all have equivalent API endpoints. See /docs for implemented operations, formats and examples.