WAV is the format you get from field recorders, interview rigs and audio editors. It holds uncompressed audio, so the quality is as good as the recording and the files are big. Both facts shape how you should approach transcribing one.
Open JotVox →A WAV normally stores uncompressed audio, so nothing has been discarded to save space. That is why editors, field recorders and archives use it, and why an hour of stereo audio is a substantially bigger file than the same hour as MP3.
For transcription the practical consequence is upload time, not accuracy. The audio has to reach the server before any transcription starts, so a large WAV over a slow connection spends most of its wait in transfer.
Upload from your device, or select the file in Google Drive or Dropbox so a large recording does not have to be downloaded and re-uploaded. AI transcription reads the audio at 98%+ accuracy across 99 languages and uses credits.
If the same recording is also published with a public link, pasting the link can be free, since JotVox grabs captions where a platform already has them and that needs no sign-up. Publicly hosted audio is covered on SoundCloud transcripts and across all supported platforms.
Audio is processed in memory and deleted right after transcription.
Only up to a point. Starting from the master WAV avoids stacking a second round of compression on top of whatever the recorder already did, which is a good habit. But an uncompressed file of a bad recording is still a bad recording.
If a WAV is impractically large to upload, exporting it to a compressed format at a sensible quality first is a reasonable trade. You lose very little that matters to a transcript.
Multitrack sessions are the exception worth planning for. A bounce with every speaker on their own clean track transcribes better than a room mix, because the loud voice no longer masks the quiet one.
Processing time scales with the runtime of the recording, so a full day of interviews takes considerably longer than a single session, and the upload adds to that on WAV in particular.
AI transcription and AI summaries use credits. Signup includes 3 free credits and one free AI transcription per day after email verification. Credit packs start at $4.99 for 20 credits and subscriptions start at $9 per month.
Summaries turn long recordings into key takeaways, chapters, quotes or a timeline, which is usually what you want from hours of raw tape.
TXT for the readable transcript, SRT and VTT when the audio is going back to picture as timed captions, and MP3 when you want a compact, easily shared copy of the sound.
The MP3 export is handy on this page in particular, since it gives you a small shareable version of a recording that started out too large to send anywhere.
Other source formats are covered on MP3 to text and M4A to text, and the general case on audio to text.
1. Decide where the file lives WAV files are large. If yours is already in Google Drive or Dropbox, select it there rather than downloading it first, which saves a full transfer of an uncompressed recording.
2. Upload and transcribe Add the WAV and run AI transcription. It reads the audio at 98%+ accuracy across 99 languages and uses credits. Expect the upload itself to take a noticeable share of the total wait.
3. Review the difficult passages Use the timestamps to check crosstalk, distant speakers and any proper nouns. Uncompressed audio helps, but it cannot fix echo, clipping or two people talking at once.
4. Export what you need Download TXT, SRT, VTT or MP3, or generate an AI summary with takeaways, chapters, quotes or a timeline. Summaries use credits.
Slightly, at best. WAV holds uncompressed audio, so working from the master avoids adding a second layer of compression, which is good practice. But at normal bitrates the difference is small compared with recording conditions. Microphone distance, room echo, clipping and overlapping speakers affect accuracy far more than whether the file was compressed.
Because uncompressed audio is large. A WAV can be several times the size of the same recording exported as MP3 or M4A, and the whole file has to reach the server before transcription begins. Selecting the file from Google Drive or Dropbox avoids downloading it locally first, and exporting to a compressed format is a reasonable option for very long recordings.
Only if the file size is a problem. Transcribing the WAV directly is fine and avoids an unnecessary step. If an hours-long uncompressed recording is slow to upload, exporting it to a compressed format at a sensible quality loses very little that matters to a transcript, since speech survives ordinary compression well.
Yes. The transcript is timestamped in the app, so you can jump to any moment in a long recording. Export SRT or VTT to keep those timings as caption cues, which is what you want if the audio belongs to a video edit. TXT gives you the continuous reading version instead.