More organizing of manual section
This commit is contained in:
Zachary Klosko
parent
d42f076daf
commit
b072db38c5
17 changed files with 7 additions and 2 deletions
|
|
@ -1,250 +0,0 @@
|
|||
LibreTime has a feature which enables your station's show and schedule information to be displayed on remote websites. This feature is included in LibreTime because you would not usually invite the general public to access your LibreTime server directly. If you had very large numbers of people requesting data from the LibreTime server at once, the burst of network traffic might overload the server, potentially disrupting your broadcasts. If carried out maliciously, this network overload is known as a *denial of service attack*.
|
||||
|
||||
Instead, your public-facing web server can retrieve the schedule information from the LibreTime API. This information can then be displayed on your broadcast station or affiliate websites by a content management system, such as Sourcefabric's **Newscoop** (<http://newscoop.sourcefabric.org/>). It can be presented using Javascript widgets and styled with CSS, in any format that you require. The **Broadcaster** theme for Newscoop (<https://github.com/newscoop/theme-Broadcaster>) integrates these widgets with ready-to-use styles.
|
||||
|
||||
There are two kinds of information that can be retrieved remotely from the LibreTime API without authentication; the metadata for the current show plus the following show (live-info), or the schedule for the current week and the week ahead (week-info). The week-info metadata includes show names, times, and individual show URLs on your public website. That way, the audience for your station can click through from the schedule information to find out more about a particular show, or download a previous show recording that you might have made available.
|
||||
|
||||
If your LibreTime server was accessible at https://libretime.example.com the live show information could be retrieved by your web server using this URL:
|
||||
|
||||
https://libretime.example.com/api/live-info/?callback
|
||||
|
||||
The comma-separated text metadata returned to your web server might be something like this:
|
||||
|
||||
({"env":"production",
|
||||
"schedulerTime":"2013-07-03 14:07:31",
|
||||
|
||||
"previous":{"name":"Earth, Wind & Fire - Boogie Wonderland",
|
||||
"starts":"2013-07-01 22:53:00",
|
||||
"ends":"2013-07-01 22:55:00",
|
||||
"type":"track"},
|
||||
"current":{"name":"Deodato - Also Sprach Zarathustra",
|
||||
"starts":"2013-07-03 13:07:06",
|
||||
"ends":"2013-07-03 13:16:05",
|
||||
"media_item_played":true,
|
||||
"record":0,
|
||||
"type":"track"},
|
||||
|
||||
"next":{"id":8,
|
||||
"instance_id":10,
|
||||
"name":"Rock Show",
|
||||
"url":"https:\/\/rock.example.com\/",
|
||||
"start_timestamp":"2013-07-03 14:00:00",
|
||||
"end_timestamp":"2013-07-03 15:00:00",
|
||||
"starts":"2013-07-03 14:00:00",
|
||||
"ends":"2013-07-03 15:00:00",
|
||||
"record":0,
|
||||
"type":"show"},
|
||||
|
||||
"currentShow":[{"start_timestamp":"2013-07-03 14:07:00",
|
||||
"0":"2013-07-03 13:07:00",
|
||||
"end_timestamp":"2013-07-03 15:00:00",
|
||||
"1":"2013-07-03 14:00:00",
|
||||
"name":"Funk Show",
|
||||
"2":"Funk Show",
|
||||
"id":7,
|
||||
"3":7,
|
||||
"instance_id":9,"4":9,
|
||||
"record":0,
|
||||
"5":0,
|
||||
"url":"https:\/\/funk.example.com\/",
|
||||
"6":"",
|
||||
"starts":"2013-07-03 14:07:00",
|
||||
"7":"2013-07-03 13:07:00",
|
||||
"ends":"2013-07-03 15:00:00",
|
||||
"8":"2013-07-03 14:00:00"}],
|
||||
|
||||
"nextShow":[{"id":8,
|
||||
"instance_id":10,
|
||||
"name":"Rock Show",
|
||||
"url":"https:\/\/rock.example.com\/",
|
||||
"start_timestamp":"2013-07-03 15:00:00",
|
||||
"end_timestamp":"2013-07-03 16:00:00",
|
||||
"starts":"2013-07-03 15:00:00",
|
||||
"ends":"2013-07-03 16:00:00",
|
||||
"record":0,
|
||||
"type":"show"}],
|
||||
|
||||
"timezone":"BST",
|
||||
"timezoneOffset":"3600",
|
||||
"AIRTIME_API_VERSION":"1.1"})
|
||||
|
||||
The information for the current week's schedule could be retrieved using the URL:
|
||||
|
||||
https://libretime.example.com/api/week-info/?callback
|
||||
|
||||
In this case, the metadata returned would be in a different format from the above example, something like the following. To keep the example short, this particular schedule export only contains four shows on a Monday. A full weekly schedule export would contain a great deal more text.
|
||||
|
||||
({"monday":[
|
||||
|
||||
{"start_timestamp":"2013-07-01 12:05:00",
|
||||
"end_timestamp":"2013-07-01 13:00:00",
|
||||
"name":"Elvis Show",
|
||||
"id":2,
|
||||
"instance_id":2,
|
||||
"record":0,
|
||||
"url":"https:\/\/elvis.example.com\/",
|
||||
"starts":"2013-07-01 12:05:00",
|
||||
"ends":"2013-07-01 13:00:00"},
|
||||
|
||||
{"start_timestamp":"2013-07-01 13:00:00",
|
||||
"end_timestamp":"2013-07-01 14:00:00",
|
||||
"name":"News",
|
||||
"id":3,
|
||||
"instance_id":4,
|
||||
"record":0,
|
||||
"url":"https:\/\/news.example.com\/",
|
||||
"starts":"2013-07-01 13:00:00",
|
||||
"ends":"2013-07-01 14:00:00"},
|
||||
|
||||
{"start_timestamp":"2013-07-01 14:00:00",
|
||||
"end_timestamp":"2013-07-01 15:00:00",
|
||||
"name":"Funk Show",
|
||||
"id":4,
|
||||
"instance_id":6,
|
||||
"record":0,
|
||||
"url":"https:\/\/funk.example.com\/",
|
||||
"starts":"2013-07-01 14:00:00",
|
||||
"ends":"2013-07-01 15:00:00"},
|
||||
|
||||
{"start_timestamp":"2013-07-01 15:00:00",
|
||||
"end_timestamp":"2013-07-01 17:30:00",
|
||||
"name":"Rock Show",
|
||||
"id":5,
|
||||
"instance_id":7,
|
||||
"record":0,
|
||||
"url":"https:\/\/rock.example.com\/",
|
||||
"starts":"2013-07-01 15:00:00",
|
||||
"ends":"2013-07-01 17:30:00"},
|
||||
|
||||
],
|
||||
|
||||
"tuesday":[],
|
||||
"wednesday":[],
|
||||
"thursday":[],
|
||||
"friday":[],
|
||||
"saturday":[],
|
||||
"sunday":[],
|
||||
"AIRTIME_API_VERSION":"1.1"})
|
||||
|
||||
If you see the message *You are not allowed to access this resource* when attempting to display schedule information in your web browser, log in to the LibreTime administration interface, click *System* in the main menu, then *Preferences*. Set **Allow Remote Websites To Access "Schedule" Info?** to **Enabled**, click the **Save** button, then refresh the browser window opened on the schedule export URL. If you do not wish to make schedule information available to the public, set this option to **Disabled** instead.
|
||||
|
||||
|
||||
|
||||
Caching schedule information
|
||||
----------------------------
|
||||
|
||||
If the LibreTime server is behind a firewall, or you want to protect the LibreTime server from large numbers of schedule requests, you may wish to cache the schedule information on a public-facing or intermediate server. You can then create a firewall rule that only allows the schedule server to connect to the LibreTime server, in addition to any remote users of the LibreTime web interface.
|
||||
|
||||
Your system administrator can set up schedule caching on a standard Apache and PHP enabled web server with the *curl* program installed, using the following steps:
|
||||
|
||||
1. Create a shell script on the schedule server (schedule.example.com) that polls the remote LibreTime server (libretime.example.com), and writes the metadata returned into a pair of local temporary files:
|
||||
|
||||
sudo nano /usr/local/bin/libretime-schedule.sh
|
||||
|
||||
The content of this file should be like the following script, replacing libretime.example.com with the name of your LibreTime server:
|
||||
|
||||
#!/bin/sh
|
||||
|
||||
curl -s "https://libretime.example.com/api/live-info/?callback=***" > /tmp/live-info
|
||||
|
||||
curl -s "https://libretime.example.com/api/week-info/?callback=***" > /tmp/week-info
|
||||
|
||||
2. Make the script executable:
|
||||
|
||||
sudo chmod +x /usr/local/bin/libretime-schedule.sh
|
||||
|
||||
3. Create an Apache VirtualHost configuration for the schedule server:
|
||||
|
||||
sudo nano /etc/apache2/sites-available/schedule
|
||||
|
||||
containing a definition like the following, replacing *schedule.example.com* with the name of your schedule server:
|
||||
|
||||
<VirtualHost *:80>
|
||||
ServerName schedule.example.com
|
||||
DocumentRoot /var/www/schedule/
|
||||
</VirtualHost>
|
||||
|
||||
4. In the schedule server's DocumentRoot folder, create the folders *api/live-info/* and *api/week-info/*
|
||||
|
||||
sudo mkdir -p /var/www/schedule/api/live-info/
|
||||
sudo mkdir -p /var/www/schedule/api/week-info/
|
||||
|
||||
5. Create an index.php file in the *api/live-info/* folder:
|
||||
|
||||
sudo nano /var/www/schedule/api/live-info/index.php
|
||||
|
||||
containing the following code:
|
||||
|
||||
<?php
|
||||
$filename = '/tmp/live-info'; // define here the path and name of uploaded live-info file
|
||||
|
||||
header('Content-Type: text/javascript');
|
||||
header("Expires: Thu, 01 Jan 1970 00:00:00 GMT");
|
||||
header("Cache-Control: no-store, no-cache, must-revalidate");
|
||||
|
||||
$callback = empty($_GET['callback']) ? null : $_GET['callback'];
|
||||
$content = file_get_contents($filename);
|
||||
$content = str_replace('***', $callback, $content);
|
||||
echo $content;
|
||||
?>
|
||||
|
||||
6. Create an index.php file in the *api/week-info/* folder:
|
||||
|
||||
sudo nano /var/www/schedule/api/week-info/index.php
|
||||
|
||||
containing the following code:
|
||||
|
||||
<?php
|
||||
$filename = '/tmp/week-info'; // define here the path and name of uploaded week-info file
|
||||
|
||||
header('Content-Type: text/javascript');
|
||||
header("Expires: Thu, 01 Jan 1970 00:00:00 GMT");
|
||||
header("Cache-Control: no-store, no-cache, must-revalidate");
|
||||
|
||||
$callback = empty($_GET['callback']) ? null : $_GET['callback'];
|
||||
$content = file_get_contents($filename);
|
||||
$content = str_replace('***', $callback, $content);
|
||||
echo $content;
|
||||
?>
|
||||
|
||||
7. Enable the new configuration and reload the Apache web server:
|
||||
|
||||
sudo a2ensite schedule
|
||||
sudo /etc/init.d/apache2 reload
|
||||
|
||||
8. Create a cron job to run the shell script each minute:
|
||||
|
||||
sudo nano /etc/cron.d/libretime-schedule
|
||||
|
||||
containing the line:
|
||||
|
||||
* * * * * www-data /usr/local/bin/libretime-schedule.sh
|
||||
|
||||
The schedule server will now be serving the same show information as the LibreTime server, with a cache lifetime of one minute. You can adjust the cache lifetime by altering the frequency of the cron job that polls the LibreTime server.
|
||||
|
||||
Pushing schedule information via FTP or SSH
|
||||
-------------------------------------------
|
||||
|
||||
If there is no inbound access to the LibreTime server at all, an FTP script can be used to push cached schedule data from LibreTime to an external web server. The standard ftp command should be available on the LibreTime server and the external web server should have a suitably restricted FTP account set up. After following steps 1 and 2 above to export schedule data to a pair of temporary files on the LibreTime server, create a new script on the LibreTime server to automatically make the upload:
|
||||
|
||||
sudo nano /usr/local/bin/upload-schedule-data.sh
|
||||
|
||||
Replace host, user and password values with appropriate values for your external web server:
|
||||
|
||||
#!/bin/sh
|
||||
HOST='website.example.com'
|
||||
USER='ftp_user'
|
||||
PASSWD='ftp_password'
|
||||
|
||||
ftp -n -v $HOST << EOT
|
||||
user $USER $PASSWD
|
||||
ascii
|
||||
prompt
|
||||
put /tmp/airtime-week-info
|
||||
put /tmp/airtime-live-info
|
||||
bye
|
||||
EOT
|
||||
|
||||
Then make the new script executable and create a cron job to launch it every minute, as in step 8 above. Steps 3 to 7 above should be carried out on the external web server so that it can convert the two temporary files uploaded via FTP into public schedule data.
|
||||
|
||||
If you have secure shell access (SSH) to the remote web server, you could write a script to use the secure copy command (scp) instead of ftp.
|
||||
Binary file not shown.
|
Before Width: | Height: | Size: 6.2 KiB |
Binary file not shown.
|
Before Width: | Height: | Size: 14 KiB |
Binary file not shown.
|
Before Width: | Height: | Size: 38 KiB |
Binary file not shown.
|
Before Width: | Height: | Size: 42 KiB |
|
|
@ -1,45 +0,0 @@
|
|||
In a typical radio station configuration, the live output from the broadcast studio and the scheduled output from LibreTime are mixed together before being sent further along the broadcast chain, to a transmitter or streaming media server on the Internet. (This may not be the case if your LibreTime server is remote from the studio, and you are using the **Show Source Mount Point** or **Master Source Mount Point** to mix live and scheduled content. See the *Stream Settings* chapter for details).
|
||||
|
||||
If your **Icecast** server is hosted in a remote data centre, you may not have the option to handover the streaming media source manually, because you have no physical access to connect a broadcast mixer to the server. Disconnecting the stream and beginning another is less than ideal, because the audience's media players will also be disconnected when that happens.
|
||||
|
||||
The **Icecast** server has a *fallback-mount* feature which can be used to move clients (media players used by listeners or viewers) from one source to another, as new sources become available. This makes it possible to handover from LibreTime output to a show from another source, and handover to LibreTime again once the other show has ended.
|
||||
|
||||
To enable fallback mounts, edit the main Icecast configuration file to define the mount points you will use, and the relationship between them.
|
||||
|
||||
sudo nano /etc/icecast2/icecast.xml
|
||||
|
||||
The example *<mount>* section provided in the *icecast.xml* file is commented out by default. Before or after the commented section, add three mount point definitions. The default mount point used by LibreTime is */airtime\_128* which is shown in the */etc/airtime/liquidsoap.cfg* file. You must also define a mount point for the live source (called */live.ogg* in this example) and a mount point for the public to connect to (called */stream.ogg* in this example).
|
||||
|
||||
<mount>
|
||||
<mount-name>/airtime_128</mount-name>
|
||||
<hidden>0</hidden>
|
||||
</mount>
|
||||
|
||||
<mount>
|
||||
<mount-name>/live.ogg</mount-name>
|
||||
<fallback-mount>/airtime_128</fallback-mount>
|
||||
<fallback-override>1</fallback-override>
|
||||
<hidden>0</hidden>
|
||||
</mount>
|
||||
|
||||
<mount>
|
||||
<mount-name>/stream.ogg</mount-name>
|
||||
<fallback-mount>/live.ogg</fallback-mount>
|
||||
<fallback-override>1</fallback-override>
|
||||
<hidden>0</hidden>
|
||||
</mount>
|
||||
|
||||
These mount point definitions mean that a client connecting to a URL such as *http://icecast.example.com:8000/stream.ogg* will first fall back to the */live.ogg* mount point if it is available. If not, the client will fall back in turn to the */airtime\_128* mount point for LibreTime playout.
|
||||
|
||||
Setting the value of *<fallback-override>* to 1 (enabled) means that when the */live.ogg* mount point becomes available again, the client will be re-connected to it. If you wish to hide the */airtime\_128* and */live.ogg* mount points from the public Icecast web interface, set the value of *<hidden>* in each of these definitions to 1.
|
||||
|
||||
Source configuration
|
||||
--------------------
|
||||
|
||||
Connect the other source to the Icecast server with the same parameters defined in the */etc/airtime/liquidsoap.cfg* file, except for the mount point. This should one of the mount points you have defined in the */etc/icecast2/icecast.xml* file, such as */live.ogg* in the example above.
|
||||
|
||||
To configure **Mixxx** for streaming to Icecast, click *Options*, *Preferences*, then *Live Broadcasting*. For server *Type*, select the default of *Icecast 2* when streaming to Debian or Ubuntu servers, as this is the current version of Icecast supplied with those GNU/Linux distributions.
|
||||
|
||||
|
||||
|
||||
By default, Icecast streams are buffered to guard against network problems, which causes latency for remote listeners. When monitoring the stream from a remote location, you may have to begin the live stream a few seconds before the previous stream ends to enable a smooth transition.
|
||||
Binary file not shown.
|
Before Width: | Height: | Size: 44 KiB |
|
|
@ -1,129 +0,0 @@
|
|||
Piwik (<http://piwik.org/>) is an open source web analytics platform which supports Icecast's log format, in Piwik version 2.0 or later. For your convenience, packages of Piwik for Debian and Ubuntu are provided in the <http://apt.sourcefabric.org> repository. If you have already configured your LibreTime server for this repository, you can install the **piwik** package and its dependencies with the command:
|
||||
|
||||
sudo apt-get install piwik php5-geoip php5-cli mysql-server
|
||||
|
||||
For security reasons, you should set a strong root password for the MySQL server. If you have not set a MySQL root password already, you should be promoted to do this during the installation of the **mysql-server** package.
|
||||
|
||||
Database and web server configuration
|
||||
-------------------------------------
|
||||
|
||||
Piwik uses a MySQL database which must be created manually before you can use the Piwik web interface.
|
||||
|
||||
1. Log into the database management system with the root password that you set for the MySQL server:
|
||||
|
||||
mysql -uroot -p
|
||||
Enter password:
|
||||
|
||||
The server should respond with:
|
||||
|
||||
Welcome to the MySQL monitor. Commands end with ; or \g.
|
||||
|
||||
Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
|
||||
|
||||
mysql>
|
||||
|
||||
2. At the mysql command prompt, create a database named *piwik* with the command:
|
||||
|
||||
CREATE DATABASE `piwik` /*!40100 DEFAULT CHARACTER SET utf8 */;
|
||||
|
||||
3. Create a MySQL user and a strong password with access to the database previously created:
|
||||
|
||||
CREATE USER piwik@localhost IDENTIFIED BY 'my-strong-password';
|
||||
GRANT ALL PRIVILEGES ON piwik.* to piwik@localhost;
|
||||
GRANT FILE ON *.* TO piwik@localhost;
|
||||
QUIT
|
||||
|
||||
4. Create a virtual host for the piwik web interface in */etc/apache2/sites-available/piwik-vhost.conf* setting the *ServerName* as appropriate for your server:
|
||||
|
||||
<VirtualHost *:80>
|
||||
ServerName stats.example.com
|
||||
DocumentRoot /usr/share/piwik
|
||||
Include "conf.d/piwik.conf"
|
||||
</VirtualHost>
|
||||
|
||||
5. Enable the virtual host and reload the web server:
|
||||
|
||||
sudo a2ensite pwiki-vhost.conf
|
||||
sudo invoke-rc.d apache2 reload
|
||||
|
||||
6. Open the *ServerName* that you set for Piwik in your browser. You should see the Piwik setup pages. Provide the MySQL database and Piwik *Super User* details when requested. If you intended to use Piwik for Icecast statistics only, you can skip the step *JavaScript Tracking Code*.
|
||||
|
||||
|
||||
|
||||
7. Uncomment the last line in the crontab file */etc/cron.d/piwik* and set your Piwik *ServerName* to enable automatic archiving every five minutes:
|
||||
|
||||
5 * * * * www-data [ -e /usr/share/piwik/misc/cron/archive.php ]
|
||||
&& [ -x /usr/bin/php ]
|
||||
&& /usr/bin/php /usr/share/piwik/misc/cron/archive.php
|
||||
-- "url=http://stats.example.com/" >/dev/null 2>&1
|
||||
|
||||
See <http://piwik.org/docs/> for more details of Piwik installation and configuration.
|
||||
|
||||
Configure Piwik for geolocation
|
||||
-------------------------------
|
||||
|
||||
By default, Piwik uses browser language settings to guess the countries in which your station's audience is physically located. In order to satisfy broadcast regulators and music royalty collection societies, a more accurate method is required, based on the location of individual IP addresses.
|
||||
|
||||
1. Edit the file */etc/php5/apache2/conf.d/20-geoip.ini* to set the path to your Piwik installation:
|
||||
|
||||
extension=geoip.so
|
||||
geoip.custom_directory=/usr/share/piwik/misc
|
||||
|
||||
Then restart the web server for the change to take effect:
|
||||
|
||||
sudo invoke-rc.d apache2 restart
|
||||
|
||||
2. Download GeoLite data from MaxMind's website <http://dev.maxmind.com/geoip/legacy/geolite/> under the [Creative Commons Attribution-ShareAlike 3.0 Unported License](http://creativecommons.org/licenses/by-sa/3.0/), for example:
|
||||
|
||||
wget http://geolite.maxmind.com/download/geoip/database/GeoLiteCity.dat.gz
|
||||
|
||||
2. Unzip and set permissions on the data file, then move it to the *misc* directory under the Piwik install directory and rename it to *GeoIPCity.dat* :
|
||||
|
||||
gunzip GeoLiteCity.dat.gz
|
||||
chmod 644 GeoLiteCity.dat
|
||||
sudo mv GeoLiteCity.dat /usr/share/piwik/misc/GeoIPCity.dat
|
||||
|
||||
4. In the Piwik web interface click **Settings** in the top right corner, then **Geolocation** in the left side menu. Change the setting from *Default* to *GeoIP (PECL)* .
|
||||
|
||||
5. Optionally, configure *Setup automatic updates of GeoIP databases* at the end of the Geolocation page.
|
||||
|
||||
Importing an Icecast log file
|
||||
-----------------------------
|
||||
|
||||
In addition to the JavaScript tracking code used for web analytics, Piwik includes a Python script which can import a server log file directly. This script accepts the parameter *--log-format-name=icecast2* which enables support for the connection duration field logged by Icecast when a client disconnects. It is this field which enables stations to track the duration of individual audience connections, and calculate both the average connection duration and the 'aggregate tuning hours' figure required by some broadcast regulators.
|
||||
|
||||
For an Icecast log file at */var/log/icecast2/access.log* you can run the script as follows:
|
||||
|
||||
python /usr/share/piwik/misc/log-analytics/import_logs.py --show-progress
|
||||
--url=http://stats.example.com --idsite=1 --recorders=8 --enable-http-errors
|
||||
--log-format-name=icecast2 --strip-query-string /var/log/icecast2/access.log
|
||||
|
||||
where the parameters are:
|
||||
|
||||
--url=http://stats.example.com
|
||||
|
||||
*The ServerName of your Piwik installation.*
|
||||
|
||||
--idsite=1
|
||||
|
||||
*Number of the default site configured in Piwik.*
|
||||
|
||||
--recorders=8
|
||||
|
||||
*How many threads to use while parsing the log file. To begin with, set this value to the number of CPU cores you can spare on the Piwik server.*
|
||||
|
||||
--enable-http-errors
|
||||
|
||||
*Collect statistics for errors, such as stream links not found.*
|
||||
|
||||
--strip-query-string
|
||||
|
||||
*Used because any characters requested by the client after the Icecast mounpoint URL can confuse the statistics.*
|
||||
|
||||
If the Piwik server is remote, you can use an additional parameter *--token\_auth=* to authenticate the request. You can find the token to use for your installation by logging in to Piwik and clicking **API** in the main navigation menu at the top of the page. Further details of the import script are shown at <a href="http://piwik.org/docs/log-analytics-tool-how-to/" class="uri" class="moz-txt-link-freetext" title="http://apt.sourcefabric.org">http://piwik.org/docs/log-analytics-tool-how-to/</a>
|
||||
|
||||
The script will report the number of records processed and the time that has been required to parse the Icecast log file. If the import script is taking too long to execute, there are performance and configuration tips in the <a href="https://github.com/piwik/piwik/blob/master/misc/log-analytics/README.md" class="uri" class="moz-txt-link-freetext">https://github.com/piwik/piwik/blob/master/misc/log-analytics/README.md</a> file.
|
||||
|
||||
Open the Piwik dashboard in your browser, and you should now see a summary of the imported data, including the most popular streams (listed under *Entry Pages*), *Visits by Server Time*, *Referrer Websites* and a *Visitor Map*.
|
||||
|
||||
|
||||
Binary file not shown.
|
Before Width: | Height: | Size: 54 KiB |
Binary file not shown.
|
Before Width: | Height: | Size: 75 KiB |
|
|
@ -1,95 +0,0 @@
|
|||
Before uploading media to an LibreTime server, there are a number of factors which should be considered. Getting your ingest workflow right will save you a lot of time later.
|
||||
|
||||
Metadata quality
|
||||
----------------
|
||||
|
||||
LibreTime automatically imports any metadata that is in the files' ID3 tags. If these tags are incorrect or are missing information, you will have to either edit the metadata manually, or suffer the consequences. For example, if the files have creator or genre metadata missing, it will be impossible to search for, create playlists or generate smart blocks according to these criteria until you add it.
|
||||
|
||||
There are a number of programs available which can be used to correct mistakes or incomplete information in ID3 tags. On GNU/Linux, the program **Ex Falso** ([http://code.google.com/p/quodlibet/](http://code.google.com/p/quodlibet/) "http://sourceforge.net/projects/dr14tmeter/")) can be useful for batch setting and editing ID3 tags before importing files into your LibreTime server. On a Debian or Ubuntu desktop machine, you can install this program with the command:
|
||||
|
||||
sudo apt-get install exfalso
|
||||
|
||||
After installation, you can run this program with the command:
|
||||
|
||||
exfalso
|
||||
|
||||
or from the desktop menu. The *Tags From Path* feature of this program is a particularly useful time saver if you have a large archive of untagged files. Sometimes there is useful creator or title information in the file name or directory path structure, which can be converted into an ID3 tag automatically.
|
||||
|
||||
|
||||
|
||||
Metadata in legacy character sets
|
||||
---------------------------------
|
||||
|
||||
LibreTime expects file tag metadata to be stored in the international *UTF-8* character set. Programs such as **Ex Falso** (described above) encode metadata in UTF-8 by default. If you have an archive of files encoded with metadata in a legacy character set, such as the Cyrillic encoding *Windows-1251*, you should convert these files before import.
|
||||
|
||||
The program **mid3iconv** (part of the **python-mutagen** package in Debian and Ubuntu) can be used to batch convert the metadata character set of files on the command line. You can install **python-mutagen** with the command:
|
||||
|
||||
sudo apt-get install python-mutagen
|
||||
|
||||
<span id="Convert_MP3_Tags_using_mid3iconv" class="mw-headline"> For example, to preview the conversion of tags from Windows-1251 </span><span id="Convert_MP3_Tags_using_mid3iconv" class="mw-headline">(CP1251)</span><span id="Convert_MP3_Tags_using_mid3iconv" class="mw-headline"> character set to UTF-8 for a whole archive of MP3 files, you could use the command: </span>
|
||||
|
||||
find . -name "*.mp3" -print0 | xargs -0 mid3iconv -e CP1251 -d -p
|
||||
|
||||
in the base directory of the archive. The **-d** option specifies that the new tag should be printed to the server console (debug mode), and the **-p** option specifies a preview run. This preview will enable you to confirm that the metadata is being read and converted correctly before writing the new tags.
|
||||
|
||||
To actually convert all of the tags and strip any legacy ID3v1 tag present from each file at the same time, you could use the command:
|
||||
|
||||
find . -name "*.mp3" -print0 | xargs -0 mid3iconv -e CP1251 --remove-v1
|
||||
|
||||
The name of the original character set follows the **-e** option. Other legacy character sets that mid3iconv can convert to UTF-8 include:
|
||||
|
||||
KOI8-R: Russian
|
||||
KOI8-U: Ukrainian
|
||||
|
||||
GBK: Traditional Chinese
|
||||
GB2312: Simplified Chinese
|
||||
|
||||
EUC-KR: Korean
|
||||
EUC-JP: Japanese
|
||||
|
||||
CP1253: Greek
|
||||
CP1254: Turkish
|
||||
CP1255: Hebrew
|
||||
CP1256: Arabic
|
||||
|
||||
Audio loudness
|
||||
--------------
|
||||
|
||||
On file ingest, LibreTime analyzes each Ogg Vorbis, MP3, AAC or FLAC file's loudness, and stores a *ReplayGain* value for that file in its database. At playout time, the ReplayGain value is provided to Liquidsoap so that gain can be automatically adjusted to provide an average output of -14 dBFS loudness (14 decibels below full scale). See <http://www.replaygain.org/> for more details of ReplayGain.
|
||||
|
||||
Because of this automatic gain adjustment, any files with average loudness higher than -14 dBFS will not sound louder than quieter files at playout time, but the lower crest factor in the louder files (their relatively low peak-to-average ratio) may be apparent in the output, making those files sound less dynamic. This may be an issue for contemporary popular music, which can average at -9 dBFS or louder before ReplayGain adjustment. (See <http://www.soundonsound.com/sos/sep11/articles/loudness.htm> for a detailed analysis of the problem).
|
||||
|
||||
Your station's producers should therefore aim for 14dB between peak and average loudness to maintain the crest factor of their prepared material (also known as *DR14* on some dynamic range meters, such as the command-line DR14 T.meter available from <http://sourceforge.net/projects/dr14tmeter/>). If the producers are working to a different loudness standard, the ReplayGain modifier in LibreTime's Stream Settings page can be adjusted to suit their material.
|
||||
|
||||
Large transient peaks in otherwise quiet files should be avoided, to guard against the need for peak limiting when ReplayGain is applied to those quieter files.
|
||||
|
||||
The **vorbisgain** command-line tool, available in the **vorbisgain** package in Debian/Ubuntu, can be used to indicate the ReplayGain of an individual Ogg Vorbis file before ingest into LibreTime. (A similar tool for MP3 files is available in the **mp3gain** package in Debian/Ubuntu).
|
||||
|
||||
Here is an example of a very quiet file where the use of ReplayGain would make the output more than 17dB louder:
|
||||
|
||||
$ vorbisgain -d Peter_Lawson-Three_Gymn.ogg
|
||||
Analyzing files...
|
||||
|
||||
Gain | Peak | Scale | New Peak | Track
|
||||
----------+------+-------+----------+------
|
||||
+17.39 dB | 4536 | 7.40 | 33585 | Peter_Lawson-Three_Gymn.ogg
|
||||
|
||||
And here is an example of a very loud file, with lower crest factor, where the output will be more than 7dB quieter with ReplayGain applied:
|
||||
|
||||
$ vorbisgain -d Snoop_Dogg-Doggfather.ogg
|
||||
Analyzing files...
|
||||
|
||||
Gain | Peak | Scale | New Peak | Track
|
||||
----------+-------+-------+----------+------
|
||||
-7.86 dB | 36592 | 0.40 | 14804 | Snoop_Dogg-Doggfather.ogg
|
||||
|
||||
In the output from vorbisgain, *Peak* is the maximum sample value of the file before any ReplayGain has been applied, where a value of 32,767 represents full scale when decoding to signed 16 bit samples. Note that lossy compressed files can have peaks greater than full scale, due to encoding artifacts. The *New Peak* value for the Snoop Dogg file may be relatively low due to the hard limiting used in the mastering of that piece of music.
|
||||
|
||||
Silence in media files
|
||||
----------------------
|
||||
|
||||
Before importing media, it is good practice to check for any silent sections in the media files. While LibreTime compensates for leading and trailing silence with the use of automatic cue-in and cue-out points****, it may be preferable to trim these files to the intended length before upload. This is because media in the LibreTime library could potentially be re-used in many different systems. **Audacity** is a cross-platform editor suitable for the task of trimming audio files, available from [http://audacity.sourceforge.net/](http://audacity.sourceforge.net "http://sourceforge.net/projects/dr14tmeter/")
|
||||
|
||||
Very quiet introductions or over-long fades can also lead to apparent gaps in your broadcast playout. Some audio CDs feature a 'hidden track' at the end, which in fact uses a long period of silence within the final track, rather than an actual separate track on the disc. This means that CD encoding programs will often encode both the hidden material and the silence in the media file. For example, the track *Debra* from the CD *Midnite Vultures* by *Beck* includes hidden material preceded by seven minutes of silence, as shown in the screen shot from Audacity below.
|
||||
|
||||
|
||||
Binary file not shown.
|
Before Width: | Height: | Size: 101 KiB |
Binary file not shown.
|
Before Width: | Height: | Size: 108 KiB |
|
|
@ -1,31 +0,0 @@
|
|||
If you have an Icecast server, you can put a link to the Icecast status page (by default at port 8000) on your station's homepage, to provide an overview of available streams. See the chapter *Interface customization* for tips on theming the Icecast status page. You can also use Now Playing widgets (see the chapter *Exporting the schedule*) or HTML5 stream players (see the chapter *Stream player for your website*) to help grow your audience.
|
||||
|
||||
On an Icecast server, you can uncomment the *<directory>* section in the */etc/icecast2/icecast.xml* file to have your station automatically listed on the Icecast directory website <http://dir.xiph.org> which could help you pick up more listeners.
|
||||
|
||||
<!-- Uncomment this if you want directory listings -->
|
||||
|
||||
<directory>
|
||||
<yp-url-timeout>15</yp-url-timeout>
|
||||
<yp-url>http://dir.xiph.org/cgi-bin/yp-cgi</yp-url>
|
||||
</directory>
|
||||
|
||||
The Indymedia stream directory at <http://radio.indymedia.org/en/yp> links to grassroots independent radio projects around the world. You can add your station to their list with an additional *<directory>* section, as follows:
|
||||
|
||||
<directory>
|
||||
<yp-url-timeout>15</yp-url-timeout>
|
||||
<yp-url>http://radio.indymedia.org/cgi-bin/yp-cgi</yp-url>
|
||||
</directory>
|
||||
|
||||
Another stream directory service is provided by the Liquidsoap Flows! site <http://flows.liquidsoap.fm/>. The following section can be added to the file */usr/lib/airtime/pypo/bin/liquidsoap\_scripts/ls\_script.liq* after *add\_skip\_command(s)* on line 174, for a stream named '*ourstation*':
|
||||
|
||||
ourstation = register_flow(
|
||||
radio="Rock 'n Roll Radio",
|
||||
website="http://radio.example.com/",
|
||||
description="Canada's most rockin' radio!",
|
||||
genre="Rock",
|
||||
user="",
|
||||
password="",
|
||||
streams=[("ogg/128k","http://streaming.example.com/libretime_128")],
|
||||
ourstation)
|
||||
|
||||
For the time being, a stream can be registered on the Liquidsoap Flows! site with any username and password. Authenticated services may be offered in future.
|
||||
Loading…
Add table
Add a link
Reference in a new issue