18
| Section 3
The codec represents the manner in which the 1.4Mbps stream of un-compressed audio data will be
converted into a smaller data stream by means of digital compression. ProSTREAM applies lossy compression
algorithms
to the processed audio in order to create a live data stream for end users via the Internet. A lossy
compression technique produces a data stream small enough to be delivered to the client via DSL, EDGE
wireless, or any common Internet network infrastructure.
Telos Systems licenses Motion Picture Experts Group [MPEG] compression codecs from Fraunhofer IIS
[http://www.iis.fraunhofer.de/en/bf/amm/produkte/audiocodec/] and includes the codecs as a source
component of ProSTREAM. Lossless encoding, like a ZIP or DMG file on a computer allow an end user to
un-compress the data and restore all of the original information. In order to gain much greater compression
ratios, audio compression employs lossy compression or encoding by making certain assumptions about
human hearing. For instance, a C# note played by a guitar at 85 db of power overpowers a C harmonic of a
keyboard at 30 db of power, so the 30db signal is minimized or tossed.
Lossy compression relies upon subjective process, thus there are many techniques to digitally compress
audio and there are many more opinions as to which are best. In anticipation of the need for compression
algorithms, the Motion Picture Experts Group (MPEG) [http://mpeg.chiariglione.org/] was formed by
the International Standards Organization (ISO) [http://www.iso.org/] and the International Electro-
mechanical Commission (IEC) [http://www.iec.ch/] to address the needs of digital compression. Their
research has yielded the codecs licensed and used by ProSTREAM.
MPEG has released different standards as technology has required, and each successive standard has built
upon the previous yielding modern codecs which can deliver extremely clear audio at incomprehensibly
high ratios of compression. When selecting a codec, there is a trade-off between more modern compression
standards which sound much better than older standards, but may risk incompatibility with older client
software and devices. The core of this choice will likely be between .MP3 (MPEG Audio Layer III) and the
newer and aptly named Advanced Audio Coding (AAC) (MPEG-4 with MPEG-2 components).
Many factors may be considered when choosing a codec to employ for the output stream.
♦
Is content mostly vocal or musical?
♦
How is the processor tuned?
♦
What types of clients are connecting to the audio stream?
Compression Codec
It is generally accepted that AAC performs better and is the codec of choice. It is rare to find clients which
are MP3 only a decade after the AAC codecs became standard, but they do still exist. MP3 was released
in 1993 and as the first digital compression technique to find widespread acceptance is in most all digital
stream clients that exist.
AAC benefited greatly from being the second go at audio compression as it was able to be designed with all
that was learned with MP3. AAC also boasts multiple “profiles” to fit the need of the transmission.
MP3 “The standard for digital audio”:
MP3 is an all encompassing audio compression codec. MP3 is the safest choice for decoding compatibility,
but creates the least clear audio of all codecs available.
