There are a number of decisions you need to make when establishing a streaming audio presence on the Web, and there are several options for what you can do and how you can do it. Do you build it yourself or outsource it? What sort of revenue models are out there - how can you make money, and how much will it cost? What technologies are out there for getting you on the air; how are they different and what makes sense for you? The answer to these questions will depend on what your expectations are and what you want to accomplish. Fortunately, even though there are more decisions to make, the process of getting online is getting easier as the number of more complete services increases.
Different types of streaming media
While there are many vendors who provide streaming media networks and systems, they often refer to drastically different ideas, such as on-demand file delivery, which may or may not be appropriate for broadcasters looking to establish a streaming audio presence. For our purposes, we'll describe streaming media as anything that provides continuous audio programming to listeners over the Internet. In practical terms, all streaming media technologies can be broken down into two basic approaches and a third approach that is a combination of those two. The two basic approaches can be described as live real-time streaming, where the audio is encoded in real-time from a live source and delivered directly to the listener's player, and file streaming, where listeners are sent a series of pre-encoded files back to back, allowing the listener to hear continuous programming.
Live real-time streaming is an approach that makes sense for radio broadcasters, as the on-air signal for your radio station is typically the content that you want to make available in the first place. Because you are streaming your on-air signal, the benefits of this approach are that you have your signature station sound and identity, and no additional effort in programming or production work is required for your Internet presence. The drawbacks of this approach are that your local advertisements will be sent to a potentially global audience, a situation that may be of limited interest to both your advertisers and your Internet listeners. This approach also fails to take advantage of the additional degree of interactivity that is possible on the Internet, where information about your individual listeners and their preferences is available to you and can be used to tailor content or advertisements.
The next level in complexity includes systems that use a mixed approach, where you have a live real-time stream that can be interrupted at key points to insert audio ads or special Internet-only content. The main motivation behind the development of mixed systems has been for ad insertion into live radio broadcasts, which opens up new opportunities for selling ads on your Internet presence. There are two ways in which this sort of ad insertion is accomplished. Some systems allow you to do ad insertion at the encoder, which means that an ad server works in conjunction with your station automation software to insert Internet-only ads on the fly into the encoded bitstream of your station's audio. At the same time a local ad is playing over the air, your Internet feed plays an Internet-only ad. Since the ad insertion is done at the encoder, all the Internet listeners will receive the same Internet-only ad. Two examples of this are Real Broadcast Network's live ad insertion system and RCS' InSert, which both provide a mechanism for substituting Internet-only ads.
Other mixed systems use an approach where the system uses knowledge about the listener's geographical location and other demographic information to drop in audio ads that are specifically targeted for each listener. These systems require additional software on the network side to determine information about the listener and select the appropriate ads. To play the ads, the player temporarily disconnects from the live audio stream to download and play the inserted ad files. It then reconnects to the live stream when the ads are done. From the listener's perspective, the audio for the inserted ads is continuous with the live stream. The biggest technical hurdle to overcome with this type of solution is synchronization, as different network latencies make it difficult to predict exactly where you will drop back in to the live stream when you reconnect. Because some of the major server and player architectures weren't designed with this type of synchronization in mind, companies like SurferNet have worked around these limitations by coming up with a custom network and their own player.
Since they don't make use of live content from your on-air signal, systems that use file streaming are probably the most radical departure for radio broadcasters. File streaming systems use pre-encoded files for all of the program material, and typically make use of playlists to instruct the player on which files to download and play and in what order. Because the files are played back to back, the listener experiences essentially continuous audio. Listeners can drop in at arbitrary times and have the same experience and can potentially skip material they don't like. Programming can still be tailored for different regions and times of the day based on how the playlists are generated.
One advantage to file streaming systems is that they make it easier for you to support multiple channels or program formats. Adding a new channel is as simple as encoding additional audio files and coming up with new playlists, potentially using some of the same stored files in common with other channels. Contrast this with real-time streaming, which requires a separate encoder with continuous audio streams fed to your servers for each audio source you wish to provide. It is interesting to note that an increasing number of Web-only broadcasters are using the file streaming approach because it can be done more in the computer domain and requires less in the way of on-air studio facilities and outboard audio gear.
On the network delivery side, file-streaming systems have one significant advantage over live real-time streaming - they can take advantage of networks built for streaming multimedia files. Because a live stream is continuously generated in real-time, you can't take advantage of caching at edge servers and some of the other tricks and distribution schemes that multimedia networks use to ensure fast delivery of files. On the downside, the user experience with file streaming systems is less refined than a typical radio broadcast, as there may be gaps between audio files, and you loose the ability for different types of segues.
Build it yourself, or outsource?
It wasn't long ago that to get your station's audio on the Internet you had to purchase and set up encoders, pay for bandwidth at an ISP, and set up your own servers. Now there are a growing number of providers who will take care of all of that for you, typically offering complete turnkey solutions. Because there are so many different types of services available now, some of which are cheaper than what it would cost to build it yourself or available for free, you need to look at several factors in deciding which one might make sense for you. The main areas you need to take into account when comparing services are the revenue model or cost of service, number of simultaneous listeners supported, coverage of publishing fees, and specialized features or services offered that affect the listener experience.
The cost for a complete system can range from expensive monthly fees to zero cost (in exchange for airtime ad bartering, revenue share on the Internet audio and/or banner ad insertion). Some multimedia service providers charge based on the total number of bytes moved to the end user, which might not make sense for continuous streaming radio stations. It is worth shopping around, as the model in use in the industry is moving towards cheaper or zero-cost solutions where you are expected to make money from sharing profits on Internet ads.
It is important to find out up front the maximum number of listeners supported, as this can vary wildly from provider to provider. In some cases, there is an additional cost if you wish to support a larger number of listeners. Companies like Real Networks charge more for server packages as the number of listeners supported increases. Because of the economies of scale that come with the size of the larger providers, often they can provide access to bandwidth that would be difficult or completely cost prohibitive for you to build yourself. Some providers have such a large network of servers and such a massive bandwidth available that they promise a virtually unlimited number of listeners.
An area that is still being worked out in the courts is publishing rights and fees for Internet audio broadcasts. It is worthwhile to find out exactly what relationship or arrangements a solution provider has made with the RIAA, ASCAP, and BMI, if any. The last thing you want to happen is to find out that you retroactively owe expensive fees for the material that you've been broadcasting on the Internet.
Another reason that it may be worthwhile to go with a solution provider is that they can provide you with technologies and special features that don't exist off the shelf right now. Ad insertion technologies are a new area, where providers can not only supply custom technologies, but can also give you the benefit of a large sales force and a national network of advertisers. Other specialized features include player branding for your station, where the listener experience is customized for your station with a distinctive look. This is typically done without requiring the listener to download a custom application.
On the farthest extreme away from broadcasting your on-air signal, there are companies that offer pure file-streaming services that essentially allow you to create a customized Internet radio station without the need for a live audio feed. Companies like Live365 offer file streaming services where all you have to do is encode and upload audio files to set up your own streaming Internet radio station. Other solution providers take care of everything including content, programming and ads. Everstream, a company that is marketing its services to newspaper and print media websites, offers continuous audio programming in multiple formats with customized content so that a particular site can offer several different channels of music and news, each featuring its own IDs and ads. While these types of services may not be directly applicable to radio broadcasters interested in getting their on-air signal on the Internet, they are worth noting as websites become more multimedia enabled and compete for the Internet listener's attention.
There are many issues to think about beyond getting your station's audio on the Internet. Are you creating a marketing presence for your station, an alternate revenue source, or a fundamentally new broadcast service with its own revenue model? Your expectations will have a major effect on what kind of Web presence you will need. Depending upon your goals, you may benefit from new services and technologies like targeted ad insertion and alternate content, or by offering multiple channels of content. As many of these technologies are relatively new and still in the development stages, it's worth doing some comparison shopping to see what is out there and how well it works for you.