WAV transcription
Check your WAV export, upload one file within the limits, and turn its spoken audio into text you can review and download.
Choose a file, or drop it here.
WAV · Max 100 MB per file · Checking the file length limit…WAV commonly stores uncompressed PCM audio. Duration, sample rate, bit depth and channel count all affect its size, so a recording can reach the 100 MB cap before reaching your minute limit. Export an audio file from your editor, rather than uploading its project file. Keep your original master; a smaller working copy may be enough for transcription.
The per-file length limit depends on your plan and available transcription minutes. The uploader shows the applicable limit before you choose a recording.
Try up to 10 cumulative guest minutes. A free account has up to 30 lifetime minutes in total, including linked guest use. These minutes do not renew monthly; all transcription pages share the same balance.
Prepare, upload, review
In your audio editor, choose an audio export such as PCM WAV. Check the selected range and listen to the exported file. If speakers use separate channels, listen to each channel before mixing; lowering the channel count can change what is audible.
Use the actual file size, not just the length shown in your editor. Select one WAV no larger than 100 MB and within the uploader’s current minute limit. If it is too large, keep the master and export a smaller supported copy of the needed section.
Start transcription, then replay important words and figures in the recording. Correct the transcript and save TXT. Keep formatting, speaker attribution and notes in your document editor when you need more than plain text.
Export reference: Audacity audio export guide. Menu names depend on your app and version.
A look at the result
Owned synthetic speech decoded from our MP3 demo with FFmpeg on October 7, 2026: 7.34 seconds, 234,816 bytes, 16-bit PCM, 16 kHz, mono. It is not an original studio master. The text is a manual edit of the spoken script, not a new WAV provider result.
Review the example
This illustrative edit changes “nine” and “ten” to digits after checking the script. The recording does not specify 9 a.m. or 9 p.m., so adding either would invent information. Decoding MP3 to WAV preserves this source’s limitations; it does not add detail. This derivative has been checked locally for decoding, not newly transcribed through the provider.
VoxText returns editable text for speech. Add speaker names, summaries or note formatting yourself. It does not generate automatic speaker labels, timestamps or subtitle files. Copy and TXT download use the latest text in the editor.
If something goes wrong
A larger sample rate, bit depth or extra channels makes a PCM file larger. Preserve the master and export only the relevant section, or make an MP3/M4A working copy. A smaller copy still uses the same recording minutes and must meet your duration allowance.
WAV is a container, not a guarantee of one encoding. Check that the file plays, then export a fresh PCM WAV from your editor. A corrupted export, project file or renamed recording needs a real audio export, not an extension change.
Inspect the source channels and exported selection in your editor. If a voice is only on one channel, confirm it survives your chosen mix. Replay the exported audio before retrying; a high-resolution file alone does not make every voice intelligible.
A smaller file does not necessarily fit your remaining minutes. Check both size and duration before uploading. Review plans and file limits →
Common uses
Compare the words in a spoken take with the script while keeping the original WAV master.
Use an exported interview section as a text draft and verify names, figures and quoted passages.
Check channels in a saved WAV and make a readable reference for manual review and note-taking.
Questions
No. Clear, audible speech and the source recording matter. A WAV converted from MP3 cannot restore lost detail, and a large WAV can still contain noise or quiet voices.
A supported file within the limits does not need to be changed just for this workflow. Keep the original and listen to any new export before using it; changing sample rate alone is not a way to recover missing speech.
Yes. Correct the text in the editor, then copy it or download your latest version as a UTF-8 TXT file. Save it before leaving the page; transcript retrieval is temporary.
No. All pages share your balance: up to 10 cumulative guest minutes, or up to 30 lifetime minutes in total with a free account, including linked guest use. The account allowance is not an additional 30 minutes and does not renew monthly.