Webradio
See also: Radio
The webradios (or netradios ) are stations more or less similar to the radio stations but which are diffused on Internet thanks to the technology of the Streaming.
As for the traditional radio stations, there exist webradios general practitioners and others with music set of themes. This is all the more easy as the emissions are not subjected to quotas like those imposed by the SCUMS or CRTC with the Radios FM. There are thousands because, technically, no matter who can create his own radio station.
In France, the webradios also profit from a specific regime concerning the royalties. They belong to a category specific to SACEM. Thus, any French creator of a webradio must legally discharge a right at the SACEM and possibly other companies of beneficiary for the diffusion of music (indeed, the SACEM always does not have an agreement with the representatives of foreign artists in particular).
In addition, some webradios diffuse only contents under Copyleft, therefore without obligation to discharge royalties on the royalty.
Many of these webradios can propose with their listeners contents without publicity, considering the relative facility of diffusion by the vector Internet which requires much less means than the radios FM. The webradios relatively often function as an association and can be financed by a system of donations.
History
The first webradio was created by using the technology of multicast. The first webradio diffused 24:00 /24 was called Radio HK and was born in February 1995. The " technologie" used was different: audio CD turned in loop and was diffused by the software of audioconference Cu-SeeMe. Radio operator HK was then one of the first to adopt the waiters RealAudio.
Medium of communication intended up to now for relatively restricted audiences, with contents usually targeted for customers particular and characterized by the fact that the diffusion is often restricted with Internet, by amateurs although more and more of stations show structure, of professionalism and having until the same rights of diffusion as the conventional ground stations.
However, one indifferently notes since the beginning of the years a 2000 important catch of advantage of the existing radios FM for the diffusion on Internet, that it is in streaming or by podcasts, owing to the fact that the stations are public, associative or commercial. Press companies or the audio-visual one (Télérama, Release, ARTE…) also benefitted from the rise to power of Internet in the hearths and the offices to create their own online radios, following the example Arte Radio, French first webradio of public service, created in 2002.
The stations of webradio can attract few tens with several tens of thousands of listeners according to popularity of the site, the emission, the organizers, the regularity of the diffusions and the capacity of the waiters.
In 2007, part of the webradios create by small structures (ordinary person or small associations) will have to close because of new regulation on the webradios dictated by the production companies (SACEM and SCP inter alia) which ask for rights of diffusion too important compared with the receipts of these webradios…
To listen to a webradio
To be able to listen to a webradio, a Ordinateur connected to Internet, a Carte its and loudspeakers or a Casque are necessary. The limiting element is today connection Internet, and especially the Band-width available. At the software level, to reach the Web site of the webradio, a Navigateur Web is necessary. To listen to a webradio a Multimedia player is in general necessary. Concretely listening is done via the URL of diffusion which either is put in a reader on the site even, or which can be inserted manually in a reader. The band-width is a factor limiting the possible quality of listening, some webradios propose several qualities possible of listening, even sometimes a very low quality for the cellphone. In order to mitigate instabilities of flow, the multimedia player used uses the principle of buffer memory, principle also used for example for the CD engraving. To note finally that there exist apparatuses which approaches a radio operator FM making it possible to listen to webradios.
Diffusion
Technical models of diffusion
Model Customer-server
In this model, a webradio generates an audio stream (voice of the organizers, songs, jingles…) towards a waiter of streaming which is given the responsability to diffuse it with the customers who connect themselves to it. In the case of a traditional station, it prepares flow in studio (with platinums, computers, mixing, microphones, etc) which is then encodé: the signal analogical is coded in a numerical format using a codec. It sends this audionumerical flow towards the waiter of streaming, which is given the responsability to redistribute it towards the listeners of the webradio. Very often, the webradios do not have with frankly speaking about studio, and function in a way much more simplistic. For example, in the case of a radio which diffuses music, all can be done directly numerically: a multimedia player connects himself to the waiter of streaming and plays a list of titles more or less by chance. A webradio can technically diffuse numerical flows of other kinds, for example of the video, using same software bricks.
It is also necessary to underline the principal difference with the hertzian radios: the waiter of streaming diffuses flow only with the customers who connect themselves, where the traditional radios diffuse waves uninterrupted. It is this characteristic which justifies customer-server name. When a listener (a " client") connects itself to the radio (with the " serveur") via a Web page, a multimedia player, his computer creates a connection with the waiter of streaming. This last then sends the binary data flow in this connection. Binary flow can be coded in Mp3, Ogg, Wma, RealAudio etc connection can be done via a protocol open standard (HTTP for example) or closed (of which the use is protected by patent: for example Rtsp). The multimedia player of the listener decodes flow then and converts the numerical data so that the computer can produce the audio signal which makes it possible the listener to listen to it (for example via a chart its).
Technically speaking, the advantage of the use of a waiter of streaming lies mainly in the fact that this one is conceived for this activity. One could indeed listen to audio streams with a traditional Web server. A waiter of streaming has in more generally a rather broad Tampon report satisfying part of the flow to be diffused, and which makes it possible on the one hand not to be too dependant on the load of the machine as of the access discs (which would slow down the diffusion and could cause " blancs" with listening), but to also perhaps be able to return quickly packages in the case of missing a deterioration of the quality of connection. Moreover, it is in general capable of réencoder to stolen flows, in another format, with another quality, to even proceed to operations such as the equalization of the noise level (in order to keep about the same noise level, whatever the level of the source).
The technical limitation which appears quickly with the webradios relates to the Band-width, mainly with the customer-server model. Thus, hundred flows with 128 kilobits/second take 12.800 kilobits/second in band-width, that is to say largely more than what is currently available for a domestic connection Internet, but which is accessible to a waiter connected to connections 100 megabits/second or 1 gigabit/second (professional connections or academics in general). For such webradios, it is current to go down quality from the diffusion to 64 even 32 kbps (with the detriment of the quality of the sound), like diffusing into mono. In the same order of idea, the waiters of streaming are often able to redirect the users whom they cannot serve towards additional waiters.
This model is the most widespread model. Many shelterers propose offers of streaming, and one can quote a certain number of very widespread server softwares: SHOUTcast (commercial), Icecast (free), Windows Media Server (commercial), Real Server (commercial). The development of ADSL and high banc in general in France for the private individuals contributed certainly to this success, and also allowed the many appearance and growth webradios, which can from now on diffuse with better flows, and thus of better qualities.
However, a certain number of criticisms are made with regard to the customer-server model, and mainly of its consumption of band-width. Alternatives to this model (used successfully for other applications) were found: this article lists some of them.
Model peer-to-peer
The webradios can also be transmitted by peer-to-peer (), which relieves their waiters and more strongly implies the listeners.
The audio stream generated by the webradio is encodé and transmitted by a first node (often called node root) to one or more other nodes, which are the listeners. These listeners retransmettent in their turn the flow which they receive with other nodes/listeners. The listeners of flow thus become " répéteurs" , and thus allow others to listen to the webradio.
The nodes are organized using a tracker, a waiter which holds up to date a list of the pars to make it possible the new pars to find somebody with which to connect itself. These new pars, once they connected, begin with retransmettre in their turn flow for other pars. They are declared then near the tracker, which will be able to send their address to even new futures.
With this method, plus radio has listeners, plus it can have others of them: the number of possible listeners is thus adapted permanently to success of the webradio, contrary to the diagram of diffusion " classique" using a waiter, where the number of possible simultaneous listeners is determined by advance. This mode of diffusion also makes it possible to realize savings of band-width, if and only if the webradio has a big number of listeners. Indeed, with many listeners reduced, these protocols have a very restricted quality of service and are often of poor quality.
The main issue of this technology comes from the low quality of service during the unhookings created by the disconnection of nodes. A possible solution would be to make store with the listeners a certain quantity of flow received (in a plug) to be able to send it to the pars which make the request of it. It is thus possible for the pars to take again the diffusion of flow a few moments behind, in the event of disconnection for example. So that this method is effective and usable, it is necessary however that the plug has a sufficient size.
This model is implemented in particular by the programs Mercora (commercial) and PeerCast (free).
Model multicast
The multicast résoud partly the problem of band-width met by the waiters of streaming: indeed, those send a flow, which is then repeated by the routers with each intersection. To simplify, if a webradio diffuses towards 2 listeners who are behind the same router, but who enters the waiter of the webradio and the router of the 2 listeners, there are 3 other routers, the webradio will send a single flow bound for the group of listeners. This flow will be transmitted same manner as a traditional flow (or Unicast) by the 3 intermediate routers. When it will arrive to the last router, that of the two listeners, this router will duplicate the stream: a stream for each listener. The principle remains identical when one increases the number of listeners and routers.
This thus small-scale model costs of band-width, but the main issue which even slows down prevents its adoption, is the fact that the current public networks do not support the multicast! Certain private networks, like those of universities or research centres are equipped routers multicast, but the general public does not have in general access there. On the other hand, the private networks of some suppliers of access to Internet are equipped in multicast, and those benefit from it to propose services in audio and video streaming complementary to the offer of access to Internet.
Diffusion on line and diffusion with the request
A webradio, that it diffuses programs carried out on line or that it repeats preregistered programs, fact of the diffusion on line, because one cannot control flow: if one starts it at two different moments, one will not have the same audio data. All the listeners connected to same a webradio receive the same audio stream simultaneously, except for some tenths of a second.
On the other hand, the diffusion with the request (one-demand English streaming) makes it possible to keep control on flow. One can thus D-listen to already diffused contents. If one starts a flow with the request at two different moments, one will obtain the same data.
The difference between direct and with the request is similar to that which exists between audio CD of a song and the same song heard with the radio: if one missed the beginning of the song with the radio, one cannot take it again since the beginning (direct), while one keeps total control on the diffusion with CD (upon request). Besides SACEM differentiates the webradios in its tariff scales according to their capacity to be been useful of the contents in flow or with the request (expensive).
One can note the relative similarity between diffusion with the request and Podcasting (which makes it possible to recover the audio file at home to listen to it later), even if the mode of diffusion is technically rather different.
Formats of data
One of the formats audio the most used for the Streaming is MP3. The binary debits can go from 24 kilobits/second (when the emissions diffuse only voice) to 320 kilobits/second (in the case diffusion of audio recordings of very high-quality) in this format.
Other formats and extensions exist, more adapted often to connections low speed (which are limited to 56 kilobits/second in theory and to approximately 51 kilobits/second in practice) or to a modest ADSL. A good quality of sound requires approximately 128 kilobits/dry in the format MP3, which is not possible that with an intermediate bond or high-speed.
Extension of the MP3, the MP3Pro. This last double in reading the quality of a received flow (one 64 kb/sec is equal then to one 128 kb/sec). At least 2 readers are compatible with this system of compression: JetAudio (the “decoder” mp3PRO is integrated there of origin) and Winamp to which it is enough to add it plug-in or codec mp3PRO. Appeared after the MP3, AAC also allows in a less flow, to improve restored sound quality. Quality FM (quality similar to that of a Radio operator FM) is available with 32 kb/sec and quality CD (quality similar to that of CD) to 48 kb/sec.
Among the open formats being able to be streamés, one can quote Ogg Vorbis. It allows, with flows lower than the MP3, but higher than the AAC, to carry out a certain saving in band-width.
Internal bonds