Metadata-Version: 2.4
Name: ovos-skill-audio-recording
Version: 0.2.13a3
Summary: OVOS audio recording skill
Author-email: JarbasAi <jarbasai@mailfence.com>
License: Apache-2.0
Project-URL: Homepage, https://github.com/OpenVoiceOS/ovos-skill-audio-recording
Keywords: ovos,skill,plugin
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: ovos-utils<1.0.0,>=0.0.28
Requires-Dist: ovos-workshop<10.0.0,>=9.5.0a1
Provides-Extra: test
Requires-Dist: pytest>=5.2.4; extra == "test"
Requires-Dist: pytest-cov>=2.8.1; extra == "test"
Provides-Extra: end2end
Requires-Dist: ovoscope<2.0.0,>=1.6.8a1; extra == "end2end"
Requires-Dist: ovos-bus-client; extra == "end2end"
Requires-Dist: ovos-core[lgpl,plugins]>=2.0.0a1; extra == "end2end"
Requires-Dist: ovos-padatious>=2.0.4a1; extra == "end2end"
Dynamic: license-file

# Audio Recording Skill

This skill records audio to a file. It needs [ovos-dinkum-listener](https://github.com/OpenVoiceOS/ovos-dinkum-listener).

## About

The skill records audio to a file and disables wake words and speech-to-text while it is active. It needs [ovos-dinkum-listener](https://github.com/OpenVoiceOS/ovos-dinkum-listener).

A similar skill, [OpenVoiceOS/ovos-skill-dictation](https://github.com/OpenVoiceOS/ovos-skill-dictation), saves text transcriptions instead of audio.

To avoid trapping a user in recording mode, you can configure a *stop hotword*. This special wake word only works during recording mode. When detected, it restores the listener to its default state. By default, no *stop hotword* is set.

A recording started by this skill times out after 4 minutes. Change this limit with the `max_recording_seconds` setting.

If a `mycroft.stop` bus message arrives (for example, "stop" on the CLI), the skill takes dinkum out of recording mode, but only if this skill started the recording.

Dinkum does not yet have a native, optional timeout that uses voice activity detection to stop a recording after a period of silence.

## Examples

- "new recording"
- "start recording"
- "new recording named {file_name}"

## Entity hints

The skill ships `locale/<lang>/intents/name.entity`, a list of example recording titles ("voice memo", "meeting", "interview", ...) for the `{name}` slot in `start_recording.intent`. These are hints, not a closed list: any title you say, including one not on the list, still fills the slot and is used as the recording's file name; listed titles simply match with more confidence. `ovos-workshop` (>=9.5.0a1) registers every shipped `.entity` file automatically when the skill's language resources are loaded, so nothing needs to be configured for this.

## Credits

[NeonGecko](https://github.com/NeonGeckoCom/skill-audio-recording)
