Web Speech API: is it possible to simultaneously record and transcribe?

[Will Koper]

Or failing that, pipe an audio source file (.wav or .mp3) into speech recognition?

[Thomas Alisi]

hey

you should be able to connect an additional javascript node to the same source and attach it to whatever the output needs to be. I'm using it to get mic levels and provide user feedback. the thing is the browser will ask for mic auth twice, check this out: https://gist.github.com/grudelsud/5442042

[Will Koper]

Thanks, I'll check it out!

I remember reading something about chrome needing to keep asking for mic permissions unless you use https.

--
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 http://groups.google.com/a/chromium.org/group/chromium-html5/?hl=en.
For more options, visit https://groups.google.com/a/chromium.org/groups/opt_out.
 
 

--
Will Koper

202-491-4869

Find me on the interweb tubes:

Twitter

Github

StackOverflow

[Robert Jones]

Here is an example that combines speech recognition with audio capture, analysis and display of a spectrogram - pretty simple, seems to work fine

https://html5-examples.herokuapp.com/web_speech_recognition_with_spectrogram.html

Hope people find it useful

--Rob Jones

--

[Robert Jones]

I have a page that combines speech recognition with capture and display of the user's audio

http://html5-examples.herokuapp.com/web_speech_recognition_with_spectrogram.html

This worked fine on Friday in stable chrome and canary

As of today it only works in Canary...

The speech recognition works in stable chrome but not the visualization

Stable is 27.0.1453.116

Canary is 29.0.1546.0 canary

It is not simply an issue of audio visualization as this works in BOTH and is pretty similar visualization code but no speech recognition:

http://html5-examples.herokuapp.com/web_audio_sonogram.html

I'm not getting any javascript errors

I know there has been a change in the way permissions are handled over the weekend - any other changes ?

Any issue with having two audioContexts ?

Any help would gratefully received

--Rob

On Wed, Jun 5, 2013 at 8:19 AM, Thomas Alisi <toma...@stinkdigital.com> wrote:

--

[Thomas Alisi]

I can see it just fine, exactly as it was last week, I'm on chrome 27

[Robert Jones]

I had tried browser reboots, etc with no luck but I just did a full system restart (Mac Mini, Mac OS X 10.7.5, cinema display) and that solved it... so I'm not sure what the cause might have been - sorry for the goose chase

--Rob

[Qiuyuan Liu]

Hi Robert, that sounds interesting but the page is no longer available. Can you post it somewhere else or send the page? Thank you!

--QY

On Monday, June 24, 2013 at 3:27:30 PM UTC-4, Robert Jones wrote:

I have a page that combines speech recognition with capture and display of the user's audio

http://html5-examples.herokuapp.com/web_speech_recognition_with_spectrogram.html

This worked fine on Friday in stable chrome and canary

As of today it only works in Canary...

The speech recognition works in stable chrome but not the visualization

Stable is 27.0.1453.116

Canary is 29.0.1546.0 canary

It is not simply an issue of audio visualization as this works in BOTH and is pretty similar visualization code but no speech recognition:

http://html5-examples.herokuapp.com/web_audio_sonogram.html

I'm not getting any javascript errors

I know there has been a change in the way permissions are handled over the weekend - any other changes ?

Any issue with having two audioContexts ?

Any help would gratefully received

--Rob

On Wed, Jun 5, 2013 at 8:19 AM, Thomas Alisi <toma...@stinkdigital.com> wrote:

hey

you should be able to connect an additional javascript node to the same source and attach it to whatever the output needs to be. I'm using it to get mic levels and provide user feedback. the thing is the browser will ask for mic auth twice, check this out: https://gist.github.com/grudelsud/5442042

On Tuesday, June 4, 2013 6:05:48 PM UTC+1, Will Koper wrote:

Or failing that, pipe an audio source file (.wav or .mp3) into speech recognition?

--
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 chromiu...@chromium.org.

[Ram Chaudhari]

Hi Qiuyuan Liu,

Di you find any workaround for this??