Audio and Public Announcement

From HSYCO
Jump to: navigation, search

HSYCO can play audio for public announcement, as pre-recorded files or using a text-to-speech engine that converts text to audio. Audio can be sent to the Web browser, the server’s audio line out connector, the internal speaker or audio out line of Axis cameras, SNOM's phones or PA devices, and I/O servers with audio capabilities.

The public announcement commands are the AUDIO action in EVENTS and the audioPlayFile() or audioPlayText() JavaScript or Java commands.

When playing a pre-recorded audio file, the file format should be supported by the audio destination you are using.

Destination Audio Formats
speaker most of the audio formats that are supported by Linux or Mac OS X should work
web use an audio format that is supported by the target Web browser

(only the most recent versions of Chrome and Firefox are supported)

axis use uLaw, 8 KHz, 8 bit, mono, WAV files
snom use uLaw, 8 KHz, 8 bit, mono, WAV files
io use an audio format that is supported by the target I/O server

The Text-to-Speech Engine

HSYCO uses different text-to-speech engines on Linux and Mac OS X.

On Linux, the default engine is eSpeak, with the possibility to use the optional Acapela engine that offers a wide selection of very high quality voices.

Note Go to http://espeak.sourceforge.net/ for more information about the eSpeak software.

On Mac OS X, HSYCO uses the integrated text-to-speech service and all its voices.

eSpeak

When using eSpeak, set the voice name to any of its supported voices.

Some voices are listed in the following table, but check the on-line documentation for the complete list.

Voice Description
voice:en English
voice:en-us American English
voice:it Italian
voice:fr French


The AudioServerSpeed and AudioServerVolume configuration parameters are supported.

The AudioServerSpeed default is 100: higher numbers increase the speech speed.

The AudioServerVolume default is 10, and should be set between 0 and 20, where 20 generates an audio file with the greatest possible amplitude.

Acapela

When using Acapela, it is mandatory to set the AudioServerTTS configuration parameter to acapela:

AudioServerTTS = acapela

Moreover, set the voice name to any of its installed voices.

When playing to the SNOM phones, use 8k and 22k voices.

When playing to Axis cameras, use 8kmu voices.

Mac OS X

HSYCO supports text-to-speech on Mac OS X 10.7 or later, using the integrated text-to-speech engine.

You can install additional voices through the speech system preferences panel.

Voice names in HSYCO are the same names you see in this panel.

With OS X 10.9, the text-to-speech engine needs to be executed using an operating system account with voices properly defined in the user's Preferences panel. Because of this, HSYCO needs to know the user name to play a specific voice, and you should prefix the voice name with the user id and a colon character, like user_id:voice Text-to-speech will not work otherwise.


The text-to-speech engine works only if a user's desktop is open. It is recommended to configure OSX with automatic login.


The AudioServerSpeed and AudioServerQuality configuration parameters are supported.

AudioServerSpeed is a positive number that represents speech rate, in words per minute.

AudioServerQuality sets the audio converter quality level between 0 (lowest) and 127 (highest).