Metadata is a set of data that describes and gives information about other data. In this case, the metadata that comes from the Radio Automation System (RAS) contains information about every audio file played, such as the title, artist, duration, and category or group. It also contains information about ad breaks; when they start and when they end.
This information is used to:
Feed the player's Now Playing display.
Record data for digital performance royalty reporting, SoundExchange, for U.S.-based publishers.
Record data for the Triton Digital 12 Hour History service.
Signal ad break starts and finishes, to enable digital audio ad insertion from Triton Digital TAP (direct sales and ad network) or TAP Programmatic (programmatic).
It is necessary to tailor your RAS metadata output to enable those features on your audio stream. This document includes process and integration details that should help you achieve the proper matching of your metadata output with Triton Digital's requirements.
Metadata Transmission Protocols
Triton Digital accepts the following transmission protocols for receiving metadata:
TCP Client (IP and Port number required to connect to automation system, also known as a pull).
TCP Server (Only a Port number required as the data is sent to the streaming computer, also known as a push).
UDP (Port number only required as the data is pushed to the streaming computer).
Serial (Baud Rate of 2400 and above only).
HTTP.
Physical File (*.txt, *.xml, *.bin, etc.).
Parsing
Parsing the Metadata is key to being able to tell the Triton Digital system what function to perform when it receives the data. This is done by separating the data into individual fields and identifying each element. Usually, these elements are the following:
Artist
Title
Length or Duration
Category, Group, Audio Number (the value that identifies what that element is, such as a spot or a song)
When we receive the data, we program our system to perform functions based on the values that are found, specifically in the "category\group\audio number\etc" field.
For example, if we received the following data:
^Leona Lewis~Better In Time~00:03:51~Song|
...we first assign each field with an order number:
^ 1 ~ 2 ~ 3 ~ 4 |
We then give each field a name that matches the data, such as:
^Artist~Title~Duration~Category|
We then program our system to read and analyze the fourth field in from the beginning, which is the category field. For example we will use the possible values of (SPOT), (SONG), and (SWP,BMP,PRO), but these values can be, and usually are, different from one automation system to another.
SONG - Contains Title, Artist, and possibly Album, Label, and any other information needed, of the Song.
SPOT - A commercial.
SWP,BMP,PRO - Bumper, Sweeper, Promo, etc.
After programming our system with the above values so it knows which is a Song, Commercial, and other events, our system is ready to execute one of the three different functions.
Within milliseconds of the audio being played on the Automation System, we receive the metadata for that audio event and parse it. Based on which of the three above values it is, we execute one of the three following functions. This all happens within a few milliseconds:
NOW PLAYING: Send the data to the player or apps.
BREAK: Begin In-stream Ad Insertion.
ENDBREAK: Eend In-stream Ad Insertion, buffer the remaining ad, and rejoin live stream.
After the execution, and the data is sent to our system, it is then translated into a format we call STWCUE and sent to our servers which identify it as a cue point. We then use custom tools to read and validate these cue points to ensure they are performing correctly.
In-stream Ad Insertion Transition Best Practices
Transitions are the points at which the on-air live audio commercial is cut off (muted) and replaced with a different scheduled commercial from our system as well as when the last ad has played, and the system goes back to the live audio. These transitions going into and coming out of a break can vary greatly, depending on a number of factors.
In order to achieve smooth transitions into- and out of break, Triton Digital recommends you follow best practices for transitions, such as:
Feed the stream post-delay audio, not pre-delay audio.
Optimize the path that the metadata follows. Routing through multiple computers, networks, firewalls, and metadata handlers (Arctic Palm, BSI, Data Repeater, etc.) can result in bumpy transitions.
Use speedy transmission protocols if possible (ideally TCP/IP); using slow transmission protocols such as serial inputs will work but is not recommended.
Allow for tiny gaps between elements, such as 300 milliseconds. Some stations use extremely tight transitions (zero gap) or even overlapping elements in order to achieve what they feel is a "tight" sound. However, our system cannot mute the audio when the on-air audio is overlapping.
Metadata Formats
There are many different automation systems, and therefore many different metadata formats. Fortunately, Triton Digital can work with almost any automation system. Most systems have either a pre-configured metadata format, or the ability to customize one. Most pre-configured formats work with our system. However, if you use your system in an unconventional or customized way, we will most likely be able to modify our system to accommodate yours.
In most cases, Triton Digital has a "preferred format" for working with each kind of automation system. For details and integration examples for a variety of automation systems, see Integration Details.