Web Speech API - Access to Audio

[Abraham Guerra]

Is it possible for me to get the audio data from the Web Speech API? I'd like to let Web Speech API do speech-to-text but I'd like to store the audio and perform some additional processing of the audio data. Is this possible with the Web Speech API?

[PhistucK]

What about invoking media capture (navigator.getUserMedia, or its unfortunate Chrome equivalent navigator.webkitGetUserMedia) when you invoke the Web Speech API? It looks like they can capture simultaneously.

Then you can use the Media Recorder API or the Web Audio API to further process the captured audio.

However, if your intention is to feed the Web Speech API some pre-processed (or post-processed, it depends on your perspective ;)) audio, then that is not currently possible, I believe.

Note that there was an intent to implement (and this may actually be implemented, but disabled by default) to support something similar (but not quite) to this, but according to the comments on the W3C bug (linked within that thread), it is intended for original microphone sources only, so post processing is not supposed to work, I guess.

☆PhistucK

On Wed, May 11, 2016 at 8:53 PM, Abraham Guerra <awa...@gmail.com> wrote:

Is it possible for me to get the audio data from the Web Speech API? I'd like to let Web Speech API do speech-to-text but I'd like to store the audio and perform some additional processing of the audio data. Is this possible with the Web Speech API?

--
You received this message because you are subscribed to the Google Groups "Chromium HTML5" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chromium-html...@chromium.org.
To post to this group, send email to chromiu...@chromium.org.
Visit this group at https://groups.google.com/a/chromium.org/group/chromium-html5/.
For more options, visit https://groups.google.com/a/chromium.org/d/optout.

[Abraham Guerra]

Thanks! Your suggestion worked like a charm.