I had some issues when installing the latest updating to the latest OpenVINO Whisper Transcription, so it took way longer than it had to… Here is what happened, maybe it will save you some time. This is using Windows (not a mac).
So the steps went like this:
I updated Audacity because I hadn’t used it in a while. 3.5 to 3.7.6.
That broke the Openvino plugins. So I went to get those, had to install the latest MUSEHUB, and create / login account. Once logged in on the web page, I was able to select the Openvino plugins, but back over to Audacity to install them, and the MUSEHUB would launch and say I need to log in. If I selected “login”, the window would just close (app crashed).
I rebooted several times, reinstalled MUSEHUB, tried looking for logs, etc., Same issue, MUSEHUB crashing. I finally rebooted again, but this time without opening my browser (or anything else), went right to the MUSEHUB app, and clicked login. Now the MUSEHUB app did the browser call, and it opened and successfully logged me in automatically.
Great, now back to Audacity and am able to click install. A few minutes later it’s done. Checked Edit:Preferences:Modules and mod-openvino is there and ENABLED. Then to Analyze: Manage Plugins and there is OpenVINO Whisper Translation listed and Checkbox Enabled.
But: When I look in dropdown under Analyze, there is no OpenVINO option. Everything else that is enabled is listed in the dropdown, but not Openvino. So I try disable / then re-enable, relaunch Audacity a few times, etc. No luck. Everything looked right, but no menu option to use the plugin. Then I saw it- The menu option was no longer available under Analyze- but I found it under Effect: Openvino Whisper .
The good news is that it works well- and seems much more stable than the early version I was using before. (That previous version would work once then crash on further attempts, until I closed and relaunched Audacity. The new version isn’t crashing at all.
Some other notes:
- After the transcription is created, If you want to get the text transcription out of audacity, first select the Transcription track, then go to File: Export Other: Export Labels. You can select txt file, or .srt (similar info but different formats).
- If you know the language in the sound file, it’s better to select it up front. I had it on auto detect, and one recording must have had some mumble in the beginning, and it auto-selected the wrong language and I got back all junk.
Good luck!