Abstract

Due to effective compression algorithms and the increases in readily available and affordable bandwidth, audio tracks are no longer protected by their once considerable data size. As a result, illegal copying and distribution of music via the Internet must be considered a new form of piracy that was made possible primarily due to the reduced size of the digital audio data and the increased bandwidth available to both consumers and organized crime. This massive distribution potential for pirated copies already appears to have severely curtailed the market volume of physical CDs, particularly that of CD singles.

Contact

Christoph Busch
Email: christoph.buschh-da.de

Michael Arnold

ACWA

Audio Protection

Besides audio files, music scores are the source for playing conventional music. They are mostly sold in the traditional form: as printed music scores. But the benefits of the Internet also affect this area. More and more publishers are setting up Internet por- tals for distributing music scores in digital format. But they are also afraid of copy- right infringement. Copyright infringement happens often in the analog world through the use of traditional copying machines. Unfortunately, the benefits of the digital representation can be easily misused. Two major reasons can be identified so far: First, digital representations can be copied with loss of information. Once a digital format is obtained from a printed music score, it can be copied again and again without any quality degeneration. Second, distribution of digital representation via the Internet is fast and without spatial limitations.

The negative effects are currently stressed by music labels, which are endangered by the massive exchange of copyrighted audio files using Peer-To-Peer architectures.
These possible effects had never been con- sidered before, since audio files had been judged as very difficult to distribute due to the huge amount of data. But modern compression schemes and permanent fast Internet access, especially for private households, have taught music industry a severe and costly lesson.

A©WA - Audio @opyright Protection by WAtermarking

The incessant increase in hardware and network capacities and improvement of software and compressing algorithms like MPEG 1 Layer III (MP3) generate a rapidly growing market for dealing in audio data. Easy to use peer-to-peer services are cur- rently used to exchange the data stored locally on the systems of participating clients. Therefore, the problem of uncon- trolled distribution of copyrighted audio data looms large. A©WA embeds a water- mark into audio data to provide the means for protecting the intellectual property rights of authors and owners alike.

A©WA is based on a statistical method and an underlying psycho-acoustic model to embed a transparent and robust watermark into audio streams Using A©WA, two watermarks can be simultaneously embed- ded without interfering with one another. The default data rate of 8 bits/second can be adjusted according to the requirements regarding robustness and the minimum data size of the audio file to be protected. The authorization for reading the water- mark is driven by using a secret key.

The A©WA embedded watermark is robust among other things - against

• – MPEG 1 Layer III compression down to 128kBit/s
• – EQ ± 6dB – Sample rate conversion (44.1 –> 22.05 kHz)
• – Stereo-Mono conversion – Format