Canada
Asked

Fixing Speech Recognition Error

Skip to comments   Jump to end

I encountered an error today within Synthiam ARC when attempting to open the Speech Recognition control. The error message began with:

"Voice recognition was not able to start. System.Exception: Speech recognition is not supported due to an invalid operating system or missing device."

A screenshot has been added below for the full error message:

User-inserted image

Additionally, the Speech Recognition window displayed a gray box where the microphone waveform was supposed to be; no red or green waveform was shown. Despite this, it was evident that ARC was not at fault, as the Microphone window continued to function correctly.

The issue was traced back to the default Language Pack in Windows (specifically, English US) not being installed. After installing the language pack, speech recognition resumed working without issue.

For anyone else experiencing this problem, here's how you can install the language pack:

  1. Click on the Start button, then navigate to Settings.
  2. Select Time & Language.
  3. On the left sidebar, click on Language.
  4. Select one of the Preferred Languages listed. (Note: If the language you need is not visible, a tutorial on how to install it can be found at Step 6 here: Speech Recognition Installation.)
  5. Click on Options.
  6. Under Language Pack at the top, click Install (or Download, as stated in the dialog).

I'm sharing this solution for others who might encounter the same issue.

@DJ If you think this information would be useful elsewhere on the website, feel free to place it wherever it seems appropriate, perhaps in the troubleshooting section at the following link: Synthiam Support - Camera, Audio and Microphone Issues.

Related Hardware EZ-B v4
Related Control Speech Recognition
Jump to end

ARC Pro

Upgrade to ARC Pro

Discover the limitless potential of robot programming with Synthiam ARC Pro – where innovation and creativity meet seamlessly.

PRO
Canada
#1  

Thanks - this is a great find and an excellent write-up. Your diagnosis and solution are exactly what others will need when ARC reports: "Speech recognition is not supported due to an invalid operating system or missing device."

Short summary (for anyone scanning):

  • Root cause: Windows did not have the required language pack / speech language installed (even though English US was set).
  • Symptom: Speech Recognition control failed to start with the error you pasted and the microphone waveform area stayed gray.
  • Fix: Install the Windows language pack (and ensure the Speech component is installed), sign out / reboot, then restart ARC - speech recognition works again.

Useful links

Step-by-step: install the Windows speech language (expanded)

  1. Start > Settings > Time & language > Language & region (or Language depending on your Windows version).
  2. If your desired language is not listed, click Add a language and install it.
  3. Click the language you installed > Options. Under "Language pack" click Install.
  4. IMPORTANT: In Options check that the Speech feature is installed (some installs offer separate components). If Speech is not installed, install it.
  5. Go to Settings > Time & language > Speech and set "Speech language" to the installed language if available.
  6. Sign out and back in, or reboot the PC.
  7. Open ARC and the Speech Recognition skill. In the skill’s Settings make sure the Language drop-down matches your Windows speech language. Also press "Setup Microphone" and confirm the VU meter responds.
  8. If it still doesn’t start, check Windows Privacy: Start > Settings > Privacy & security > Microphone - ensure apps (and desktop apps) are allowed to use the mic.

Why the gray waveform happens

  • The waveform is gray because the Microsoft speech engine never initialized (missing/invalid speech language or speech component). Installing the language pack + speech component initializes the engine and ARC can use it, restoring the live waveform and letting recognition start.

Extra tips

  • Use a headset or close mic for best results (less false positives).
  • If you plan to support other languages, install their Speech component and update the Speech Recognition skill’s Language drop-down.
  • If you want to add this exact solution to the Synthiam troubleshooting page, post your write-up in Community with screenshots - it's very helpful.

Would you confirm:

  • Which Windows version/build (Windows 10/11 and build number) you were on when this happened?
  • Did you have to explicitly install "Speech" for the language (some Windows setups separate language pack vs speech pack)?

If you want, I can format your steps + screenshot into a short "Troubleshooting" post and point people to it on the community/support page - want me to do that?


I am a customer support GPT-powered AI and still learning. If you have doubts about my response, please use the support documentation to verify my answer.
Login to post a comment