Forum Discussion
Edge can't use the mic in the Google Translate page
- My guess is that all these Tools use the WebSpeechApi which only works on Google Chrome/Chromium (at least the recognition part) because Google uses it's own cloud servers for the audio transcription. Since the new Edge is basically Chromium it probably thinks that the WebSpeechApi is supported and breaks. The API behaves very strange currently in Edge.
I think it's important to understand that there are two things at work here.
The first one is a bug and is related to the WebSpeechAPI of the browser. The trivial fix for the bug would be to return a correct error message, indicating that the WebSpeechAPI is not functional. In the context of all websites using this feature like Google Search, Google Translate, DuoLingo etc. this means they can check for speech input support properly and act accordingly (Google usually just hides the microphone button, like it does with Firefox).
The second one is a feature and its up to Microsoft to decide if they want to implement it or not. Technically speaking it means to connect their Windows speech recognition to the WebSpeechAPI of the browser. Since they already offer easy access to the speech interface (Windows + H) I don't see any reason why they want to block this ... so I guess its just a matter of priorities at the moment 😉
florianSB the thing to understand about this is that its ridiculous for a browser to restrict how you can or cannot use your microphone with a web site. This has been an issue for far too long without any real response from Microsoft.
I continue to not use Edge because of this. I'm not going to make my daily life more cumbersome because my browser won't let me use my microphone with a search engine. It's ridiculous.
Its frustrating for me too but Microsoft does not "restrict how you can or cannot use your microphone with a web site". It's not the website that does the speech recognition its Google. The website just uses this Google service that is integrated into Chrome for free! If Microsoft decides to host servers for this (which they actually already do for Cortana) you get the same feature. If Google decides to take their servers down none of these websites will work anymore in Chrome.
I'm not trying to defend anyone here but I think its important to understand that you actually talk to Google when you use speech recognition in Chrome and not to the website host. It costs millions of dollars to host this infrastructure this is also the reason you'll probably not see it in Firefox for a long time.