|Name||Modified||Size||Downloads / Week||Status|
|Totals: 2 Items||1.2 MB||110|
The biggest news in this release are: improved logic linking, improved QSO recorder, support for squelch on GPIO, improved pre- and deemphasis filters. More details about each change can be found in the ChengeLog below.
Also note that the filename for sound clips have changed a small bit. A sound clip archive should now be named:
svxlink-sounds-<lang_code>-<voice name>-<samp rate>-<svxlink ver>.tar.bz2
A new audio codec, Opus (http://www.opus-codec.org/), is now supported for RemoteTrx links. The Opus codec is a new IETF standard for encoding audio with bitrates ranging from 6 kb/s to 510 kb/s. The audio quality is very good over the whole bitrate range. It is even better then AAC, at least below 128kb/s. To enable SvxLink to use the Opus codec, make sure to install the development libraries for Opus and recompile. =============================================================================
Improved support for linking logics together. It is now possible to have multiple links active at the same time for a logic. A link can be set to be default active. This mean that the link will be activated as soon as the application starts and also that the link can be automatically reactivated, after a timeout, if manually deactivated.
Improved the logic shutdown feature. Upon shutdown, any active module is deactivated immediately and module activation is not allowed until the logic is activated again. This for example means that if the logic is shut down and an EchoLink connection comes in, that connection will be rejected. Also, no other DTMF commands than the logic activation command is allowed when the logic is shut down. The command was also renamed to ONLINE_CMD since everything became more logical. Calling it SHUTDOWN_CMD was a bit confusing. Should subcommand 1 mean that the logic is shut down or brought back online? Naming the command ONLINE_CMD make everything clear: 1=online, 0=offline.
Improved QSO recorder. There now is only one configuration variable, QSO_RECORDER, in a logic configuration section. This configuration variable is used to set the DTMF command and to point out a configuration section where the rest of the QSO recorder configuration is specified. The QSO recorder can now close a recording after a configurable recording time and start a new one. The maximum time is specified using the MAX_TIME config variable. The SOFT_TIME config variable is used to specify an interval where the file will be closed on the first detected squelch close. This will give more natural cuts between recordings. It is now also possible to specify a maximum total size for all files in the recording directory. This is done using the MAX_DIRSIZE config variable. When the limit have been reached, the oldest file(s) will be deleted. The DEFAULT_ACTIVE config variable is used to set the default activation state of the QSO recorder. If set to 1, the recorder will be activated upon SvxLink startup. If the TIMEOUT variable is set, the activation state will return to the one set in DEFAULT_ACTIVE if manually activated/deactivated. If less than MIN_TIME milliseconds of audio has been recorded when a file is closed, the file will be deleted. If the node is idle for more than QSO_TIMEOUT seconds, the file will be closed. The ENCODER_CMD variable makes it possible to run an external audio encoder to compress the recordings.
Added support for Squelch on GPIO. Use SQL_DET=GPIO and GPIO_SQL_PIN=gpio3 in your config for GPIO support. Patch contributed by DG9OAA / Jonny.
GPIO pins are now specified with the full name (e.g. gpio3) instead of with just the pin number. This enables the use of any pin naming scheme.
Improved pre- and de-emphasis filters. They were not good at all in the previous relese. Now the filters should behave as expected. Also, the filters have been designed to be each others complement so that they will exactly cancel each other out.
Now possible to specify DTMF_DEC_TYPE=NONE (or just comment it out) if no DTMF decoder is required for a certain receiver. Also made it possible to specify NONE for SEL5_DEC_TYPE to make them behave in the same way.
Bugfix for the REPORT_CTCSS config variable. Depending on the locale settings (e.g. the LC_NUMERIC environment variable) the wrong CTCSS frequency could be announced.
Bugfix: Setting DTMF_MUTING=0 did not remove the audio delay, which it should do.
Sound clips are now included for module MetarInfo.
The Async::AudioRecorder class now have some added features. In addition to the hard filesize limit there now is a soft limit at which the file will be closed when the audio source call flushSamples. A signal will be emitted when either of the limits are hit. Also, begin and end timestamps now are available.
The Async::AudioFilter setOutputGain method now take a dB value as argument.
Added support for the Opus audio codec.
New class Async::Exec for executing external programs.