Convert text to WAV for editing
Generate a voiceover from any text and download it as an uncompressed 16-bit PCM WAV - the format editors, DAWs and video tools prefer to work from.
Free, no credit card. Generate once and take away the WAV, the MP3, or both.
- 16-bit PCM
- No re-encoding
- No credit card
Edit the line to hear anything you like.
The honest version of what WAV gets you
Plenty of sites advertise WAV as the high-quality option. Here is what is actually happening: the neural voice engine returns MP3 at 24 kHz, 48 kbps. The WAV download decodes that MP3 and writes the samples out uncompressed.
So the WAV is not more detailed than the MP3 - it holds exactly the same audio, stored without compression. That is still worth having, but for one specific reason: every further lossy encode degrades the signal again, and starting from WAV means your final export is the second encode rather than the third.
| Your situation | Better choice |
|---|---|
| Listening on a phone or in a browser | MP3 - an eighth of the size, no audible difference |
| Dropping into a video that will be exported | WAV - avoids a third encode |
| Mixing with music or effects in a DAW | WAV - editors resample and re-render repeatedly |
| Emailing to someone | MP3 - a 21 MB attachment is a problem |
| Archiving the take you liked | WAV - no further loss if you re-export later |
File specifications
| Property | Value |
|---|---|
| Container | RIFF/WAVE (.wav) |
| Encoding | 16-bit signed PCM, uncompressed |
| Channels | Mono |
| Sample rate | 24 kHz, matching the source |
| Data rate | 48 KB per second — about 2.9 MB per minute |
| Built | In your browser, on request — the audio is not re-sent anywhere to produce it |
Effects change the picture slightly
If you apply effects or background music, the audio is processed and re-encoded before you download it. In that case the WAV is decoded from the processed file rather than the original take, so it carries one more encode than a dry render would. If you intend to mix in your own editor anyway, download the dry voiceover and do the processing there - you will end up with a cleaner result than layering both.
Questions
- Is the WAV higher quality than the MP3?
- No, and it would be dishonest to imply otherwise. The voice engine produces MP3, and the WAV is decoded from it. What you get is uncompressed audio from that point onward - the detail already lost to the first encode does not come back.
- Then why would I choose WAV?
- Because further encoding is lossy too. If the audio goes into an editor and gets exported again - into a video, a podcast, a lesson - starting from WAV means you encode once more instead of twice. For anything played straight from your phone, MP3 is the better choice.
- What are the file specifications?
- 16-bit PCM, mono, at the decoded sample rate of the source - 24 kHz. Standard RIFF/WAVE, so every editor and player opens it without a plugin.
- How large is the file?
- 48 KB per second, so about 2.9 MB per minute - eight times the MP3. A seven-minute render is around 21 MB. It is assembled in your browser when you ask for it, so nothing extra is transferred.
- Can I get both formats?
- Yes. Generate once, then download whichever you need - both are built from the same render, so asking for the second one costs no allowance.