IMPROVE TEXT-TO-SPEECH VOICE

Windows 10 saw the addition of the Cortana assistant. Whether you use it or not, the voice itself is quite a bit better in terms of quality.

However, by default, users wanting to utilize these voices using various Text-to-speech programs like DRE are not able to do this. It requires a workaround that’s fairly simple to do.

Below we will go through the few steps involved to unlock the Cortana Eva or Mark voices to be used for TTS applications.

 

Note

  • This fix is for Windows 10 only! If you have Windows 11, there’s at the time of writing this no possibility to unlock Cortana voices.
  • On non-English Windows 10, you need to install the English US language to use the voices below
  • Note: this is only needed when DRE uses local voices. Using Neural Voices, this guide has no effect

 

Making the backup

  1. Before making the required edits to the Registry Editor, we first should back up the existing registry, in case anything fails in the following steps.
  2. Hit the Windows Start key and type Registry Editor
  3. Registry Editor search
  4. In Registry Editor go to File -> Export and save the registry backup
  5. Close Registry Editor and you’re now ready to unlock the new Cortana voices

 

 

Unlock Cortana voices

  1. Download this zip file that contains two registry editor ‘reg’ files
  2. Open or extract the downloaded zip file
  3. Double click either Eva or Mark reg file
  4. That’s it! The voice is now unlocked

Cortana reg files

 

Set Cortana voice as TTS voice

  1. Hit the Windows key and type ‘Control panel’
  2. In the search box top right type ‘Speech’
  3. Open the ‘Change text to speech settings’Cortana voice selection
  4. In the Text to Speech tab under Voice selection, choose either
    1. Microsoft Eva Mobile
    2. Microsoft Mark
  5. That’s it. Now your TTS voice should be using one of the Cortana voices

The alternative

If you want even higher quality text-to-speech voices there are some pretty good payware versions out there.

Finally, DRE Essentials, Performance, and Ultimate all have neural text-to-speech, which improves the quality immensely.