SHOUTcast DNAS Transcoder 2
- 1 Introduction
- 2 Getting Started
- 2.1 Running The Transcoder
- 2.2 Windows
- 2.3 Linux / Mac OS X / BSD
- 2.4 Additional Signals
- 2.5 Registering for MP3 Stream Encoding
- 2.6 Supported File Formats
- 3 Configuration File
- 4 Administration Weblet Commands
- 5 Metadata Support
- 6 Filename Metadata Extraction
- 7 Playlists Support
- 8 DJing Information
- 9 Virtual Memory Footprint (Linux / Mac OS X / BSD)
- 10 Locale Error (Linux / Mac OS X / BSD)
- 11 Example Configurations
The purpose of this document is to show you the different configuration options supported by sc_trans along with basic and more advanced example configurations enabling you to get started with using sc_trans and the various features it can offer.
To get started using sc_trans you need to have some media to play or to have a stream you want to relay via sc_trans. With a 'source' you can use then you will need to setup your configuration file for the transcoder so it can process the input source and provide an output which can be used by the DNAS server (see sc_serv2.txt in the sc_serv distribution for more details especially those for stream configurations as non-matching details will prevent sc_trans from being able to connect to it).
The following sections will show how to get sc_trans up and running along with details on the configuration options it provides for handling different sources and outputs it can support.
Running The Transcoder
The transcoder is able to be run either as a console application or it can be run as service (Windows) or daemon (Linux / Mac OS X / BSD). Detailed below is how to get the transcoder running on the different operating systems supported by it.
The Windows version of sc_trans is designed to run on fully updated and patched versions of Windows 2000, XP, Vista and Windows 7.
Please note that the Windows versions of sc_trans are built with a dependency against the Microsoft Visual C++ 2008 SP1 Redistributable Package. If sc_trans is unable to start due to a dependency issue then you will need to install the correct version of the package so it can run which depends on the version of sc_trans you are attempting to run:
Install as a Service
sc_trans.exe install <servicename> <username> <password> <conf>
<servicename> - Name of service
Typically enter this as 'sc_serv' though you can use a different name but you will need to remember it as it is required to be the same when using the 'uninstall' mode.
<username> - User under which to run the service as or '0' for the local system
<password> - Password for user or '0' for the local system or with no password
<conf> - File path to the configuration file
If no file / an invalid file is specified then sc_trans will abort loading.
When run as a service there is not a good sense of a current working directory so requires all paths in the configuration and playlist files to be fully qualified otherwise lookups will fail when sc_serv is run.
To run sc_serv with a configuration file in the same folder as the server as the current system user account you would enter into the console:
sc_trans.exe install sc_trans 0 0 sc_trans.conf
Uninstall the Service
sc_trans.exe uninstall <servicename>
<servicename> - Name of service as used when the service was registered
To uninstall sc_serv assuming it was installed as detailed in the install section above then you would enter into the console:
sc_trans.exe uninstall sc_trans
Run as a Non-Service in the Console
<conf> - File path to the configuration file (can be relative or absolute)
If no file / an invalid file is specified then sc_trans will abort loading.
Linux / Mac OS X / BSD
Remember to enable the required access on the sc_trans file by doing 'chmod a+x sc_trans' after extracting it from the distribution file otherwise the OS is likely to not run it and show the error message './sc_trans: Permission denied'.
Run as a Daemon
./sc_trans daemon <conf>
<conf> - File path to the configuration file (required in all cases)
If no file / an invalid file is specified then sc_trans will abort loading.
./sc_trans daemon ./sc_trans.conf
When run this should output the following:
'sc_trans going daemon with PID [XXXX]' where XXXX is the <pid> of the process.
End a Daemon
kill -SIGTERM <pid> or kill -15 <pid> or kill -s TERM <pid>
<pid> - The PID of the daemon instance (reported when the daemon started or can be found with 'ps ax | grep sc_serv' as long as sc_serv was the file run otherwise you can just use 'ps ax' if the filename isn't known). Additionally the PID of sc_trans is listed in the log file.
Run as a Non-Daemon
<conf> - File path to the configuration file (can be relative or absolute)
If no file / an invalid file is specified then sc_trans will abort loading.
When run on Linux / Mac OS X / BSD then some additional signals are supported to allow for additional control over the running daemon instance of sc_trans.
The following signals can be used with the 'kill' command (in the manner of your choosing for using the kill command) along with the <pid> of the daemon instance to do one of the following actions:
SIGKILL - Stops sc_trans (also SIGTERM, SIGINT and SIGQUIT will work) SIGHUP - Rotates logfile SIGWINCH - Skips to next song in playlist SIGUSR1 - Reload playlist SIGUSR2 - Toggle playlist shuffle
The result of SIGHUP is that the current log file contents will be moved into <logfile>_1 e.g. sc_trans_1.log, <logfile>_1 will be moved into <logfile>_2 e.g. sc_trans_2.log and so on for all log files which can be found which match the current log file's name. This is useful if timed to have it create day specific log files.
These signals are not supported by the Windows version of sc_trans which will only respond to the Ctrl + C / Ctrl + Break / console close commands the OS provides.
Registering for MP3 Stream Encoding
In order for compliance with all broadcasting laws there is a legal requirement for the need to purchase an MP3 license key in-order to be able to encode your stream in MP3 (see section 3.10 for the related configuration entries). The license key should cost $5.
To obtain the license key you need to go to the following page which will takes you to a secure page for purchasing the MP3 encoding license key:
Once the license key has been obtained (this should be a quick process only taking a few minutes when done) then you just need to add the 'name' and 'key' you have obtained into the 'unlockkeyname' and 'unlockkeycode' values in your configuration file (see section 3.10) and to change the 'encoder' type for the stream(s) you require from 'aacp' to 'mp3' to use it (see section 3.4).
So if your key is '123456' and you have used 'Bob Cratchit' as the registered name during the billing process, you would add the following to your configuration file:
unlockkeyname=Bob Cratchit unlockkeycode=123456
If the details do not exactly match then the Transcoder will show a message and will not allow the encoding of MP3 streams until the configuration issue has been resolved. The common cause for failure is not entering the unlockkeyname exactly as specified for the billing name of the method used during the purchase.
Why Does AAC Encoding Not Require a License Key?
Encoding to an AAC stream does also require a license like the MP3 encoding though this has already been paid for you so there is no need for you to do so yourself when using sc_trans. This setup may seem strange but it is just how things happen with the different nature of licensing encoding of the AAC and MP3 formats.
Supported File Formats
As sc_trans does transcoding of audio from one format to another for then being fed to a server or another sc_trans instance as a relay, the following file extensions are handled as valid input file sources:
MP3 + MPEG (treated as mp3) WAV (uncompressed) OGG FLAC
Here you can find a complete list of all of the configuration options which are provided by sc_trans which ranges from logging to networking configuration to control over how the media is encoded before transmission to the server.
The entries stored in the configuration file are processed case insensitively.
Configuration entries labelled as <MULTI> can be used to set up specific encoder types or for specifying multiple playlists for use with the calendar system. These are specified by adding _# to the end of the option's name as shown below where # begins with one. If you are only working with a single instance then you do not need to add the _# part as any instances of a configuration option will assume it is for _1.
Note: The <MULTI> system is not hierarchical and all values beyond the default must be specified for all connections. So for example, if you wanted to connect to two servers on the same port you must do:
encoder_1=aacp encoder_2=mp3 bitrate_1=32000 bitrate_2=32000
Note that you CANNOT do it like this as it leads to not all values being set:
encoder_1=aacp encoder_2=mp3 bitrate=32000
The configuration files also allow for notes or options to be disabled by the use of a comma (;) at the start of a line which can be seen in all of the configuration examples.
Known options in the configuration files are recognised irrespective of the case they are entered in the configuration file so djport and dJpOrT will be handled the same way.
Any items found in the configuration file which are not known (as detailed in following sections) or is not processed as a comment will be reported in the following manner:
<date + time> W msg:[CONFIG] Invalid item on line XX
Where <date + time> is the date and time the event happens at and XX is the line in the configuration file where the error has been found to occur at.
adminport : Specify the port used to access the administrative control page [Default = 0]
If no valid port is specified (i.e. zero) then this feature is disabled.
adminuser : Specify the username used to access the administrative control page [Default = admin]
adminpassword : Specify the password to access the administrative control page [Default = goaway]
calendarfile : Specify the path to the xml based calendar file [Default = calendar.xml]
calendarrewrite : Re-write the 'calendar file' on exit [Default = 1]
Note: See calendarxml.txt for more information on the event format and what it allows you to do with playlist, DJ and relay access.
djport : Specify the port to listen on for DJ access using the SHOUTcast 1 protocol [Default = 0 (no listening)]
This will actually open 'djport + 1' as the listening port.
djport2 : Specify the port to listen on for DJ access using the SHOUTcast 2 protocol [Default = 0 (no listening)]
Note: It is possible to use 'djport' and 'djport2' at the same time though you need to ensure that there is no overlap of the ports used in what is specified.
This is because 'djport2' will open the port specified unlike 'djport' which opens 'djport + 1' (as is the way with SHOUTcast 1 connections). Keep this in mind as you would not be able to do the following as it causes a duplication of the port being opened (901 in this case) :
djport=900 ; i.e. will open port 901 djport2=901
Whereas this would work as we are opening two different ports (901 + 902):
djcipher : Specify the key used to obfuscate the initial handshaking with the SHOUTcast 2 protocol [Default = foobar]
This is a YP2 feature and it matches the 'uvoxcipherkey' value in the sc_serv configuration file (see sc_serv2.txt - section 4.14).
Only change this if you really need to do so as not all SHOUTcast 2 clients will allow you to edit this value from the default value. If using the Source DSP plug-in then see dsp_sc.txt - section 5.0 for details on how to change the plug-in to use a differnet value.
autodumpsourcetime : Specify the maximum idle time for the DJ in seconds before being disconnected [Default = 30]
djbroadcasts : Specify the directory in which DJ broadcasts will be recorded into [Default = recorded\ or recorded/ - relative to sc_trans[.exe] ]
djcapture : Enable to allow the recording of the DJ broadcast [Default = 1]
djfilepattern : Specify how the DJ archive filenames will be created in the form of <djname><date time pattern>suffix [Default = _%Y_%m_%d_%H_%M_%S. (year_month_day_hour_minute_second)]
The date and time information are formatted using strftime().
<MULTI> (one set for each DJ):
djlogin : Specify the username required for the DJ to get access [Default = dj]
When connecting as a SHOUTcast 1 source the password has to be specified as <djlogin>:<djpassword> e.g. dj:noise
djpassword : Specify the password required for the DJ to get access [Default = noise]
djpriority : Specify the priority for the DJ when multiple DJ's are connected [Default = 1]
<MULTI> (one set for each encoder):
encoder : Specify the encoder to use, either (mp3 or aacp) [Default = aacp]
bitrate : Specify the encoding bitrate in bits per second [Default = 96000]
samplerate : Specify the encoding sample rate to use [Default = 44100]
channels : Specify the number of channels to encode [Default = 2]
mp3quality : Specify the MP3 encoder quality to use when encoding [Default = 0] Available options are :
0 - fast (preferred) 1 - high quality (changes sample rate and does additional filtering)
mp3mode : Specify the MP3 encoder mode (CBR / VBR) to use when encoding [Default = 0] Available options are :
0 - CBR 1 - VBR High Quality 2 - 3 - VBR Medium Quality 4 - 5 - VBR Low Quality]
flashpolicyfile : Specify the name of file which contains the flash crossdomain policies [Default = crossdomain.xml]
flashpolicyserverport : Specify the port used for the flash policy server [Default = 0]
Specify '843' if you want to turn this on.
capture : Enable to allow the use of live capture as an input [Default = 0]
capturedevice : Specify the OS dependent device used for capture [Default = <no value>]
captureinput : Specify the OS dependent input used for capture [Default = <no value>]
capturesamplerate : Specify the sample rate used for capture [Default = 44100]
capturechannels : Specify the number of channels to capture [Default = 2]
Note: When run under Windows on an operating system newer than XP there is a possibility for the device name obtained from Windows to be longer than 32 characters. This is a problem as the api's used only support device names upto 31 characters in length so sc_trans will attempt to clip the device name specified with 'capturedevice' to 31 characters to try to get a match with the devices the OS is reporting.
If you are having issues with the live capture then add 'capturedebug=1' to allow you to see the device names reported back to sc_trans by the OS.
log : Enable logging of the transmission output [Default = 1]
screenlog : Enable logging of output to the console [Default = 1]
If log=0 or when running this as a daemon / service then this option is ignored as it is not applicable.
logfile : Specify a different logfile to save the transmission logs into [Default = %temp%\sc_trans.log or /tmp/sc_trans.log]
If you are using a folder for saving the logs into then you need to ensure that the folder already exists as sc_serv will not attempt to the create the folder for you.
streamtitle : Specify the general name for the radio station or stream [Default - Misc]
streamurl : Specify the url of the radio station or stream [Default = http://www.shoutcast.com]
genre : Specify the genre for the radio station or stream [Default - N/A]
aim : Specify the AIM contact ID to show for the radio station or stream [Default - N/A]
irc : Specify the IRC contact ID to show for the radio station or stream [Default - N/A]
icq : Specify the ICQ contact ID to show for the radio station or stream [Default - N/A]
public : Enable to register the radio station or stream with the YP (shoutcast.com) [Default = 0]
If connected to sc_serv then this is used by the 'publicserver' value unless it is being overriden (see sc_serv2.txt - section 4.14).
usemetadata : Enable to use use the in file metadata i.e. stored in file tags [Default = 1]
metadatapattern : Specify the pattern used for extracting metadata from the filename [Default = *\%N.* or */%N.*]
See section 6.0 for more details about this)
displaymetadatapattern : Specify a pattern to use to form the metadata output string for SHOUTcast 1 metadata [Default = %R[ - ]%N]
nextlookahead : Specifies the number of tracks ahead of the current track playing to be reported in the stream's metadata to the server [Default = 3]
This is the maximum limit attempted to be reported and will vary based on the source in use e.g. if there enough entries in the current playlist being played. Setting to zero will disable this.
configrewrite : Re-write the 'config file' on exit [Default = 0]
timemultiplier : Used to shift reported time in certain uses [Default = 1]
inheritconfig: Specify the path of the sc_serv2 configuration file for the instance you want to use the transcoder with [Default = <no value>]
The aim of this option is to make it easier to get an instance of sc_trans to run with sc_serv by ensuring that any passwords, etc match up. By using this, option any of the configuration options already set in the sc_trans config file (see below) will be replaced by those found in the sc_serv configuration file.
The configuration options read from the sc_serv2 configuration file including what they are mapped from in the sc_serv2 value to the sc_trans value are:
- portbase -> serverport
- password -> password - Master source password overriden by streampassword
- streampassword -> password
- uvoxcipherkey -> djcipher
- srcip -> serverip - Will use the ip set unless is empty or srcip=any
- yp2 -> outprotocol - Sets outprotocol=1 if with yp2=0 otherwise it sets outprotocol=3 as a v2 DNAS is assumed to be used and if found to be v1 configuration then outprotocol=1
Additionally this mode will attempt to read configurations from a v1 DNAS file.
include : Specify an additional include file containing settings to be processed from the current point in the main configuration file [Default = <no value>]
Note: You can do multiple calls of this allowing for a basic configuration file with then 'specific' encoder configurations set in individual conf files though you need to ensure not to include a reference to the same file in itself.
serverbackupfile : Set the server side file for broken connection playback [Default = <no value>]
serverintrofile : Set the server side file that is played when a user connects [Default = <no value>]
Note: With both of these options, they will override any files specified in the sc_serv's configuration file and also requires 'outprotocol=3' for it to be supported (see section 3.11 for the options 'outprotocol' supports).
Additionally when connecting to sc_serv via sc_trans, the server intro and backup files as specified in sc_servs configuration file (see sc_serv2.txt - section 4.5) will not be played so you need to specify them in the sc_trans configuration if required for the stream (this is a quirk of how sc_serv currently works).
unlockkeyname : Specify the name to use to unlock mp3 encoding (case sensitive)
unlockkeycode : Specify the code associated with name to unlock mp3 encoding
Due to licensing requirements for MP3 encoding see section 2.5 for how to obtain the key required for unlocking MP3 encoding in sc_trans.
<MULTI> (one set for each connection):
endpointname : Specify the name of the endpoint [Default - <no value>] Note: This is used to identify the encoder such as when using the AJAX api interfaces and if not specified then sc_trans creates a default name as 'endpointX' where 'X' is the encoder index starting at one. See sc_trans_ajax_api.txt for more information on the use of this.
outprotocol : Specify the server protocol to use [Default = 1] This can be:
1 = SHOUTcast 1 (Legacy) 2 = Ultravox (Ultravox 2.0) 3 = SHOUTcast 2 (Ultravox 2.1)
serverip : Specify the IP address of server to connect to [Default - <no value>] Note: If 'srcip' has been set to a value other than 'any' in the sc_serv configuration then you need to set it for you sc_trans configuration otherwise sc_serv will not see the sc_trans connection attempts (see sc_serv2.txt - section 4.2).
Additionally, you will need to make sure that any IP added is valid and can be fully resolved otherwise the connection to the sc_serv instance is very likely to fail.
serverport : Specify the server port used to connect to [Default = 8000]
This should relate to the 'portbase' value set in sc_serv's configuration (see sc_serv2.txt - section 4.8).
password : Specify the password used to connect to a DNAS server [Default = foobar]
This matches the 'password' value in the sc_serv configuration file (see sc_serv2.txt - section 4.8) or 'streampassword' (see sc_serv2.txt - section 4.12 - this only applies for v2 DNAS servers).
streamid : Specify the stream identifier for the SHOUTcast 2 connection [Default = 1]
This relates to the 'streamid' value in the sc_serv configuration (see sc_serv2.txt - section 4.8).
uvoxuserid : Specify the user ID for accessing SHOUTcast 2 (Ultravox 2.1) [Default - <no value>]
This is currently not used and is provided for future usage.
Note: If 'requirestreamconfigs' is enabled in the sc_serv configuration (see sc_serv2.txt - section 4.12) then you need to ensure that the 'streamid' value used in your sc_trans configuration exists in the sc_serv configuration along with any endpoint, password and other values appropriate to the 'streamid' matched with the equivalent sc_serv configuration values.
uvoxradiometadata : Enable to send the older AOL Radio style metadata to the server [Default = 0]
uvoxnewmetadata : Enable to send SHOUTcast 2 style metadata to the server [Default = 1]
uvoxmetadatafudgefactor : Specify the delaying factor for Ultravox metadata in seconds [Default = 0.0]
shoutcastmetadatafudgefactor : Specify the delaying factor for SHOUTcast metadata in seconds [Default = 0.0]
playlistfile : Specify the name of playlist file to use [Default = <no value>]
shuffle : Enable to allow shuffling of the specified playlist [Default = 1]
xfade : Specify the number of seconds to do for a crossfade [Default = 1]
Specify xfade=0 to disable crossfade
xfadethreshold : Specify the minimum duration in seconds a file must be to allow crossfading in and out on it [Default = 10]
Priority playlists are those used to override the playing playlist once the current song has stopped playing. This can be useful if a sudden playlist / announcement is required.
playlists : Specify the folder used for priority playlists [Default = priority\ or priority/ - relative to sc_trans[.exe]]
archive : Specify the folder used for archiving playlists [Default = priority\archived\ or priority/archived/ - relative to sc_trans[.exe]]
<MULTI> (one set for each additional playlist):
playlistfilename - Used to build named collections of playlists [Default = <no value>]
playlistfilepath - Used to build named collections of playlists [Default = <no value>]
Note: These playlists are used as part of the <playlist> events in calendar.xml where the name set with 'playlistfilename_X' is used in the loading done. See calendarxml.txt and section 3.2 for more information on these events.
Replay Gain Control
applyreplaygain : Enable to honour the replay gain values stored in the file [Default = 0]
defaultreplaygain : If no replay gain is in the file then apply this adjustment [Default = 0.0]
djreplaygain : Specify the replay gain to be applied to DJ streams [Default = 0.0]
capturereplaygain : Specify the replay gain to apply to the live capture input [Default = 0.0]
calculatereplaygain : Enable to calculate the replay gain on the handled files [Default = 0]
replaygaintmpdir : Specify the temporary directory for replay gain calculator to work in [Default = (Windows = %TEMP FOLDER% Linux = /tmp/)]
replaygainrunahead : Specify the number of tracks head start to give the replay gain calculator [Default = 2]
replaygaindontwrite : Specify if the replay gain values calculated should NOT be written to the file [Default = 0]
enhancereplaygain : Specify if the file has replay gain and it is being used, an additional amount of gain to add [Default = 6.0]
Winamp ships with a default value to add -6dB to file which do not have replay gain.
vuimagedirectory : Specify the folder in which to look for vu images to use [Default = ""]
The image files are named vuimage_XX.<vuimagesuffix> where 'XX' is from 0 to 100 and the path specified will need to be properly terminated to work correctly.
vuimagesuffix : Specify the suffix for vu images to be used from the image directory [Default = png]
vuimagemimetype : Specify the mime type for vu images to be used [Default = image/png]
Note: This controls the images available when using the 'vumeterleft' or 'vumeterright' webcommands (see section 4.1). Additionally sc_trans only allows for one active use of each webcommand from the current sc_trans instance with the most current attempt made being the active instance which will be updated.
In all cases, the default value is for no debug logging for the options listed below.
shuffledebug : Enable to activate debug logging of playlist shuffling [Default = 0]
shoutcastdebug : Enable to activate debug logging for SHOUTcast transmission [Default = 0]
uvoxdebug : Enable to activate debug logging for SHOUTcast 2 (Ultravox 2.1) transmissions [Default = 0]
gaindebug : Enable to activate debug logging for replay gain on playback [Default = 0]
playlistdebug : Enable to activate debug logging for playlists [Default = 0]
mp3encdebug : Enable to activate debug logging for MP3 encoding [Default = 0]
mp3decdebug : Enable to activate debug logging for MP3 decoder [Default = 0]
resamplerdebug : Enable to activate debug logging for the resampler [Default = 0]
rgcalcdebug : Enable to activate debug logging for the replay gain calculator [Default = 0]
apidebug : Enable to activate debug logging for the AJAX api [Default = 0]
calendardebug : Enable to activate debug logging for the calendar events [Default = 0]
capturedebug : Enable to activate debug logging for live capture [Default = 0]
djdebug : Enable to activate debug logging for DJ management [Default = 0]
flashpolicyserverdebug : Enable to activate debug logging for the flash policy server [Default = 0]
fileconverterdebug : Enable to activate debug logging for the server side file converter [Default = 0]
sourcerelaydebug : Enable to active debug logging for relayed sources [Default = 0]
sourceandendpointmanagerdebug : Enable to activate debug logging for endpoint management [Default = 0]
Administration Weblet Commands
The transcoder supports web based control of the features it provides allowing for the remote control of sc_trans much like the DNAS (sc_serv2) supports. To enable the use of the administration interface you need to specify a port and related password and username as detailed in the Administration section (see 3.1).
Once enabled then options detailed in the following sections will become available though the main interface and controlling aspects are available via the 'api' command with more information provided on how to use it in sc_trans_ajax_api.txt.
Supported Weblet Commands
The following are the supported commands for the administration weblet the transcoder supports and are used in the following manner:
quit : Will make sc_trans quit (equivalent of SIGKILL on non-Windows).
rotatelogs : Will rotate the logs (equivalent of SIGHUP). The result of this is that the current log file contents will be moved into <logfile>_1 e.g. sc_trans_1.log, <logfile>_1 will be moved into <logfile>_2 e.g. sc_trans_2.log and so on for all log files which can be found which match the current log file's name. This is useful if timed to have it create day specific log files.
nextsong : Will advance to the next song (equivalent of SIGWINCH).
loadplaylist : This will re-load the current playlist (equivalent of SIGUSR1).
toggleshuffle : Use to toggle the current playlist shuffle state (equivalent of SIGUSR2).
kickdj?duration=hh:mm:ss : You can specify an amount of time for how long to kick the current DJ for. If a duration is not set i.e. just using 'kickdj', then this will defaults to 15 minutes (00:15:00). If a duration is specified then all three fields of 'hh:mm:ss' must be specified even if their value is zero (ie 01:30:30).
restart : Restart the current sc_trans instance.
vumeterleft or vumeterright : This will initiate a server side image push that represents the vu levels as long as there are images are available to sc_trans to use for this (see section 3.14).
vu.bin : This will return a continuous stream of type 'misc/sc-vu' as long as sc_trans is transcoding. The format of the stream is:
0xa5 leftVal rightVal 0xa5 leftVal ..... etc
If the audio signal is mono there is no right val and you get:
0xa5 leftVal 0xa5 leftVal .... etc
writeconfig : This will write out the currently stored configuration and event details to the respective configuration and calendar.xml file. You will need to ensure that you have specified valid files in the configuration options for these files otherwise they won't be saved correctly and any changes may be lost.
crossdomain.xml : This will return the current copy of crossdomain.xml as set with 'flashPolicyFile' (see section 3.5).
timeshift?t=# : [EXPERIMENTAL] This applies a time shift to the audio though the default is to have no shift (i.e. t=0). This should not be used unless you know what it does (added for specific Aol use).
api : This is the point from where you can make use of the AJAX api which sc_trans offers and is used as a POST method. For more information on the parameters and options supported see sc_trans_ajax_api.txt.
As detailed in section 3.8 there are a number of ways to allow for the use of metadata from the files being played though there are certain limits to this support for different file formats currently supported by sc_trans as detailed in the following sections.
There is support for ID3v1 and ID3v2 tags. With ID3v2 there is support of version 3 and 4 of the specification though there are the following limitations:
- Unsynch data is not supported
- Compressed frames are not supported
- Encrypted frames are not supported
- Appended v4 frames are not supported
FLAC and OGG Files
There is support of the native metadata (Vorbis comment tags). Metadata read from these files are mapped to expected values which SHOUTcast expects to get. For more information see sc2_xml_metadata.txt for details on the various field mappings applied.
Filename Metadata Extraction
The transcoder along with being able to make use of the metadata from files is also able to attempt to extract metadata from the filepath of the files by use of a template.
Guessing of the metadata from the filename will happen if 'usemetadata=0' is set in your configuration file (default behaviour is to use the file tag) or if no tag is detected in the file. For the best experience it is always best to ensure any files you use have tags and the required metadata in them.
Filename Metadata Extraction Pattern
The 'metadataPattern' setting is used to specify the pattern used to extract metadata from a filename if there is no detected metadata or 'usemetadata=0'.
The default pattern used is *\%N.* (Windows) and */%N.* (all other OSes). The following are the formatting options which are supported in the pattern:
%N = Song name %G = Genre %A = Album %R = Artist %Y = Four digit year %# = Sequence of digits %% = % character  = Brackets optional elements (no nesting) * = String Other characters are matched exactly as found.
The pattern matching is done working from the right to the left.
Filename Metadata Extraction Examples
The following examples show the metadata pattern matching working.
metadataPattern = */%R/%Y%A/Audio/[%#_]%N.* filepath = /home/songs/Mask and Wig/1969 Irreverance Of Things Past/Audio/03_The Royal Blues.mp3
Extracted metadata result:
ALBUM = Irreverance Of Things Past ARTIST = Mask and Wig SONG = The Royal Blues YEAR = 1969
metadatapattern = */%A/[%R_]%N filepath = /home/album/song
Extracted metadata result:
ALBUM = album SONG = song
6.2.3 Example #3
metadatapattern = */%A/%R_%N filepath = /home/album/artist_song
Extracted metadata result:
ALBUM = album ARTIST = artist SONG = song
File Metadata Construction For SHOUTcast 1
You can specify how metadata elements are concatenated together to from the SHOUTcast 1 metadata string by specifying the 'displaymetadatapattern' option.
In a standard Winamp client install then it will expect the data in the equivalent format of '%R[ - ]%A[ - ]%N' and uses that format to parse the data. If you deviate from this or if the user has changed settings in their Winamp client then the results may not display properly with what is expected to be shown.
The following token substitutions are made with 'displaymetadatapattern':
%N = Song name %A = Album name %R = Artist %G = Song genre %Y = song year %t = Station name (a.k.a. stream title a.k.a. icy-name) %u = Station url (a.k.a. icy-url) %g = Station genre (a.k.a. icy-genre)
Characters in brackets are only displayed if the token substitution to the left of them did not result in the insertion of an empty string.
The transcoder supports a playlist file to specify the file(s) to use when it is working and is a file with the .lst extension with one file per line. The file may be encoded in UTF-8 to allow for the support of non-english filenames, etc.
Lines in the playlist beginning with a # will be ignored which allows you to disable a specific file or script in the playlist or comments on what is done.
The playlist file will be processed until the end of the file is reached or until the parser comes across a line containing ###!EOF is detected.
With versions of sc_trans prior to those from the end of 2010 there was a need to have a line break after the last entry in the playlist for them to be correctly recognised. This is no longer a requirement of the playlist format.
Wildcards are supported in the processing of the folder contents though this will not do a recursion of any sub-folders within the specified folder e.g. c:\music\*.mp3
This will only find mp3 files in the 'music' folder and will not look in any sub-folders which may be inside the 'music' folder. For sub-folders to be included then you will need to explicitly specify the sub-folder for it to be recognised by the playlist loader.
Additionally these entries can be intermixed with regular explicit entries e.g.
c:\music\foo.mp3 c:\music\otherstuff\*.mp3 c:\music\bar.mp3
Any entries which relate to file formats not supported (determined by the file extension) will be ignored from the playlist stored e.g. test.aac would not be used if read.
Supported Playlist Formats
The transcoder supports the following playlist file formats though is limited to use only the filepaths from the playlist formats which also provide additional title / metadata information in them:
Plain text file - this is the equivalent of a commentless M3U / M3U8 file PLS M3U M3U8 (UTF-8 encoded version of the M3U format)
If a loaded playlist contains any entries with wildcards in them then the playlist loader will only attempt to find the matching files in the base folder specified and will not recurse into any sub-folders found. This is different to how Winamp for example (and any other programs) handles the loading of such entries where it would then recurse through any sub-folders found. This behaviour may change at a future time but is not guaranteed.
An additional feature of the transcoders playlist support is the ability to run remote applications in place of a fixed playlist line. This is done by starting a line with #! e.g.
/music/foo.mp3 #!/bin/perl /home/user/songfromDB.pl /music/bar.mp3
The script run in this case should return the entry in the playlist to run at this point. The script is also passed an integer parameter indicating how many tracks ahead must also be returned by the script.
By returning these extra tracks as well as the first one expected then this information will be used by YP2 to refine searching and also to establish a "coming soon" track which will be shown to clients or on the SHOUTcast site.
So with this style of playlist handling, it is possible to have an external app running which will provide all of the files for sc_trans to process which could be of use if you have a specific requirement for when and how files are played.
Additionally you can pair remote application actions together in such a manner:
#!php call track script #!php call commercial script
where this will alternate between the first and second scripts returning the file to be processed through sc_trans.
Every entry in the playlist is run through a wildcard expander specific to the operating system the transcoder is being run on. Under *nix systems then glob() is used and for Windows systems FindFirstFile() is used to handle the wildcard.
There are a number of scenarios to be considered for remote application entries in a playlist as detailed in the following list:
- The standard out from the command is used as the name of the file to open.
- If the standard out is empty then the entry will be skipped.
- If the command returns an error code (non zero) then the entire playlist of which the entry is a member off will be considered to be finished.
- The data written to standard out should be encoded in the UTF-8 format and if returning multiple entries needs to be separated by a newline character.
- If only a single track name is returned by the remote application then the following values in the playlist will be used for "coming soon" calculations.
- If you return the requested number of additional tracks (passed in when it is run) then these will be used for "coming soon" calculations.
- If there are [ or ] in the filepath then for them to appear instead of being used as a filter, you will need to escape them normally as \[ and \] else it is often likely that the files will fail to load. This is more of an issue on non-Windows versions due to how the shell and expansion works.
The DJ port (configuration option 'djport' or 'djport2') mimics the behaviour of sc_serv by allowing someone running another copy of sc_trans (or the Winamp DSP plug-in) to connect to your sc_trans and then to broadcast through it which effectively turns your sc_trans into a relay.
The 'djport' or 'djport2' value (see section 3.3 for more details setting this up) is the IP port you choose to have your sc_trans to listen on for an inbound DJ connection (by default this feature is not enabled).
A simple example of this feature is if you have a friend who wants to do a show on your station. At the appropriate time they would start their own instance of sc_trans which will then connect to your instance as if it was a server. When the connection happens then your sc_trans instance will stop the current playlist playback and will crossfade into your friends show. Then when the show is over and the friend disconnects then your instance of sc_trans will crossfade back into the playlist from the point when the friend had connected.
The DJ broadcast will also be copied to disk for later playback assuming that 'djcapture' has not been changed from the default settings though you can turn this feature off if required. See section 3.3 for more available options.
Setting Up DJs In The New Scheme
DJ accounts are set up by using the 'djlogin', 'djpassword' and 'djpriority' options as shown in <MULTI> options under section 3.3.
If we have two DJs who we want to allow access into our system then we would have the following in our main configuration file (obviously in real use change the names and passwords to what you need):
djlogin_1=peter djpassword_1=foobar djpriority_1=4 djlogin_2=paul djpassword_2=goaway djpriority_2=2
The priority value is used if the scheduling allows for two different DJs to connect into the instance of sc_trans at the same time. If this did happen then the DJ with the higher priority will take precedence on the access made.
Now that we have setup the DJs in our main configuration, we will need to setup access for them via the calendar.xml file scheduler (see calendarxml.txt). In this example Peter is allowed access to sc_trans at any time whereas Paul is only allowed a connection to sc_trans on Tuesdays at 2am for an hour. Finally the shows Peter makes will not be saved whilst Paul's ones will be archived.
<?xml version="1.0" encoding="UTF-8" ?> <eventlist> <event type="dj"> <dj archive="0">peter</dj> <calendar/> </event> <event type="dj"> <dj archive="1">paul</dj> <calendar starttime="02:00:00" duration="01:00:00" repeat="4"/> </event> </eventlist>
Now everything is setup, all that is needed is to run sc_trans and for Peter and Paul to setup their setups (either via another sc_trans or Winamp + Source DSP for example) with the required login details and then they will be able to connect as set in the schedule.
Note: When connecting as a SHOUTcast 1 source, the DJ must specify the login password in the form of <djlogin>:<djpassword> e.g. paul:goaway If not done correctly then the DJ will not be able to connect. Section 3.3 provides more information on the configuration values required.
Virtual Memory Footprint (Linux / Mac OS X / BSD)
Due to how sc_trans works it will create a large number of threads which may become an issue due to the default virtual memory stack allocation per thread potentially being large (typically 10MB per thread). For most people this should not be an issue as the memory is not really allocated unless it is used. If however if you have a resource constrained system or if you plan on running many copies of sc_trans then you could potentially run out of virtual memory.
The default thread allocation can be reduced by issuing a "ulimit" command before running sc_trans and in the example reduces it to 512KB per-thread e.g. ulimit -s 512
Locale Error (Linux / Mac OS X / BSD)
If you receive the following error then you have a locale related issues:
terminate called after throwing an instance of 'std::runtime_error' what(): locale::facet::_S_create_c_locale name not valid Abort trap
If this happens then you need to set the 'POSIX' locale in your terminal's environment values. If this does not happen then everything is fine or you are using a newer version of sc_trans from the end of 2010 which should no longer exhibit this issue.
setenv LC_ALL POSIX
Mac OS X:
You can test to see if this has worked by checking the shell command "locale":
LANG= LC_COLLATE="POSIX" LC_CTYPE="POSIX" LC_MESSAGES="POSIX" LC_MONETARY="POSIX" LC_NUMERIC="POSIX" LC_TIME="POSIX" LC_ALL="POSIX"
Include as part of the sc_serv installation are a number of example configurations to get things started with using some of the features of the transcoder. The provided examples are:
sc_trans_basic.conf sc_trans_capture.conf sc_trans_dj.conf sc_trans_playlist.conf sc_trans_simple.conf
All of the configuration examples are documented and will relate back to details in this file as appropriate. You will need to change some details in these example files or to obtain any applicable license keys for the setup you are making (see section 2.5).
In all cases the examples are designed to work from the same install folder as sc_trans.
Remember these examples are a way to get you up and running faster though you are still likely to have to adjust things depending upon the systems you are using and how you want to manage your sources and server connections which are covered in previous sections.
Additionally if you are not happy with editing the files or find it still too confusing then you can use the configuration builder which (if installed) can be found in the 'config_builder'. Open 'config_builder.html' in your browser and then use this graphical tool to setup a configuration file for the transcoder (and additionally the server).
This is the base configuration from which the other configuration examples are based and this will get a sc_trans instance running and making a connection to a local instance of sc_serv (see sc_serv2.txt - section 9.1 about the sc_serv_basic configuration file).
This configuration file changes the required options in the sc_trans_basic configuration to get a sc_trans instance running as a capture device e.g. from a microphone. This shows the use of the 'include' option (see section 3.9).
This configuration file changes the required options in the sc_trans_basic configuration to get a sc_trans instance running with the ability to accept DJ connections using both a SHOUTcast 2 source and a legacy SHOUTcast 1 source in addition to how calendar.xml files are used to control DJ access.
This configuration file changes the required options in the sc_trans_basic configuration to get a sc_trans instance running with the ability to schedule a jingle to play around 20 minutes and 40 minutes past the hour as well as to schedule a different playlists to be played in the evening and also on the weekend. This example shows of some of the features of the event system and how it is possible to automate playlist changes.
This configuration file can be included in any other configuration files to allow you to get debug logging output when trying to investigate any issues you may be experiencing.
Use this if you just need to get a very basic transcoder running or are impatient or are struggling to get it running despite the previous example configurations.
This configuration file is designed to be used just as is and is the simplest form of the configuration file you can have which will allow you to get a running transcoder to work with an instance of the server on the same machine.
This works by assuming default settings of the transcoder and that the server has been properly configured to work as this configuration uses the 'inheritconfig' option to get some of the passwords and other configuration details directly needed for the transcoder to work with the server instance by using sc_serv_simple.conf (see sc_serv2.txt - section 9.5) or change this to reference a known server configuration you want to use this with.
All you need to do when using this configuration file is to:
- Edit the playlist file to be used (as referenced in the playlistfile entry)
- Setup the required encoder options including entering the MP3 unlock code by filling in 'unlockkeyname' and 'unlockkeycode' entries if using MP3 decoding (section 2.5 has details on why this is required and how to obtain the unlock key and code)
- Change the stream information details as appropriate