🔄 Classic Speech Recognition
For Windows 11 24H2 and forward, training your Speech Recognition is no longer available.
Microsoft decided to keep the general Speech Recognition API as-is, but forgot to keep the recognition profile training functionality…
This means that the built-in Windows Speech Recognition (WSR) is no longer able to recognize your voice prompts well, unless you already have a trained recognition profile.
To work around this, DRE now allows you to instead use the legacy >Microsoft< Speech Recognition (MSR) as its internal speech recognition engine, which doesn’t need voice training.
This adds flexibility from always using the >Windows< Speech Recognition (WSR), which supports and require training your voice recognition profile. Without training it, the recognizer will suffer greatly on recognizing your voice prompts.
Change this in
Sound -> Recognition -> RuntimeChoosing the Classic recognizer type
- Windows:
- Built-in and trainable for your voice; requires good audio quality.
- Allows for dictation commands.
- Not trainable on Windows 11 24H2 or later, so using this option on Windows 11 24H2 or later will perform sub-par to the ‘Microsoft’ option!
- Defaults to this value on Windows 10 and versions earlier than Windows 11 24H2
- Microsoft:
- Optimized for noisy/poor audio and works with many microphones and accents.
- No training needed but requires separate runtime/language pack installations.
- Disables dictation commands: "AdminChatAdmin", "AdminChatAll", "AdminChat", "NotesAdd", "ChatStart", "ChatReply", "DDChat".
- Defaults to this on Windows 11 24H2 or later
- Windows:
| What do I choose? | Option |
|---|---|
| If you’re on a Windows 10 or Windows 11 earlier than 24H2 | Windows |
| If you’re on Windows 11 24H2 or later: | |
| Have a trained recognition profile from the past | Windows |
| Have no trained recognition profile or are unsure | Microsoft |
- How do I check if I have a trained recognition profile?
- Use the shortcut Win + R and paste in:
Rundll32.exe shell32.dll,Control_RunDLL %SystemRoot%\System32\Speech\SpeechUX\sapi.cpl,,1 - Switch the tab to Speech Recognition
- If you see a Profile listed, other than the Default Speech Profile, chances are that you trained a new one
- You could also have trained the Default Speech Profile, which is fine.
- Use the shortcut Win + R and paste in:
