No a Issue but Need some details regarding streaming app.

[higunjan]

I'm developing speech to text web app. As per recording and generate wave file I converted that wave into google api supported .raw file and sent to API it successfully converted into text.

Now I have to develop like streaming when I speak send chunk to google API and bind text output on web page.

How should I develop. record direct into .raw format is that any way or google API support with direct chunk file.

Technology : NodeJS

[jmdobry]

You can find the encodings supported by the Google Cloud Speech API here: https://cloud.google.com/speech/reference/rest/v1beta1/RecognitionConfig#AudioEncoding

It may be that your .wave file matches the MULAW encoding supported by the Cloud Speech API. You'll have to try it and see. More details on encodings: https://cloud.google.com/speech/docs/encoding

You can access the Cloud Speech API from the browser, but only via HTTP (REST) and only for the SyncRecognize and AsyncRecognize methods.

To do streaming, you need to use Speech#createRecognizeStream on the server in Node.js, then send the results back up to the browser.