# Speech input as stream w/o Mic?

**URL:** https://community.openconversational.ai/t/speech-input-as-stream-w-o-mic/1271
**Category:** General Discussion
**Created:** [January 27, 2017, 12:47pm UTC](https://community.openconversational.ai/t/speech-input-as-stream-w-o-mic/1271 "2017-01-27T12:47:16Z")
**Posts on this page:** 3
**Page:** 1

<div class="post-metadata">

### Author: ![duke](https://community.openconversational.ai/letter_avatar_proxy/v4/letter/d/e8c25b/32.png) [@duke](https://community.openconversational.ai/u/duke)
#### Post date: [January 27, 2017, 12:47pm UTC](https://community.openconversational.ai/t/speech-input-as-stream-w-o-mic/1271/1 "2017-01-27T12:47:16Z")

</div>

Hi there,

is there any chance to get speech input over network (eg. stream w/ websockets) into mycroft?

I don’t have a hardware Mic and wanted to use my smartphone with k6nele … works fine with kaldi + gstreamer.

- Michael

---

<div class="post-metadata">

### Author: ![Kaarel](https://community.openconversational.ai/user_avatar/community.openconversational.ai/kaarel/32/740_2.png) [@Kaarel](https://community.openconversational.ai/u/Kaarel)
#### Post date: [July 29, 2017, 11:31am UTC](https://community.openconversational.ai/t/speech-input-as-stream-w-o-mic/1271/2 "2017-07-29T11:31:32Z")

</div>

I don’t know the answer to your question but I just asked a similar one (and then found your question as a related one): [Using Mycroft without Mycroft’s speech recognition](https://community.openconversational.ai/t/using-mycroft-without-mycrofts-speech-recognition/1953)

You can use the Kõnele rewrite rules feature ([http://kaljurand.github.io/K6nele/docs/et/user\_guide.html#ümberkirjutusreeglid](http://kaljurand.github.io/K6nele/docs/et/user_guide.html#%C3%BCmberkirjutusreeglid), see the example with FetchUrlActivity) to execute a transcription as an HTTP query. This way you’d be using your Android device’s speech recognizer and would be only sending text to Mycroft. The question remains if Mycroft supports the required HTTP interface.

Your question is if Kõnele could be used as a microphone for Mycroft, i.e. Mycroft would receive the audio as websocket packages. That would be an awesome feature as well (although a bit harder to implement). Maybe it would even make sense to design the hardware in a way that the mic array becomes an optional add-on.

---

<div class="post-metadata">

### Author: ![PasabaPorAqui](https://community.openconversational.ai/letter_avatar_proxy/v4/letter/p/4af34b/32.png) [@PasabaPorAqui](https://community.openconversational.ai/u/PasabaPorAqui)
#### Post date: [August 6, 2017, 11:27am UTC](https://community.openconversational.ai/t/speech-input-as-stream-w-o-mic/1271/3 "2017-08-06T11:27:51Z")

</div>

Could be this [wiki](https://github.com/MycroftAI/mycroft-core/wiki/Mycroft-commands-from-Android-or-other-RTP-streams) page is useful to you.
