True Audio Checker - Project Description

From Tau Projects
(Redirected from aucdtect)
Jump to: navigation,
True Audio Checker Logo

The Aim of the Project

The aims of the CD authenticity detector project are authenticity control of the music records when purchasing musical discs and automatical determination whether the record is an authentic one or is recovered from the lossy encoded data (for example, MP3). The necessity of such a program and possible solutions are widely known, correct solution is original one.

Dear friends! This project needs your donation, even if it's will be a very small amount of money. Please support our project! Feel free to request new functionality with your donation. The current objectives of the project are listed below. The development team thanks you in advance for your support!

Project Overview

It is no secret that the Internet and file-sharing networks are literally overloaded with music. Lossy formats - like MP3 - offer the highest degree of compression, and are easily the most popular means of file-sharing. When played through relatively low-fidelity audio systems, the sound offered by such files is often indistinguishable from CD quality audio - to the average human ear. However, lossy-encoded audio tracks fare poorly when played on Hi-Fi systems, or when enjoyed through a pair of quality stereo headphones.

Example audio collection

Since the untrained ear is an unreliable detector, the purchase of traditional audio CDs has given rise to a new concern - the authenticity of the audio data they contain. Here is a sample collection of audio CDs, purchased between 2000 and 2004 in various music stores. As you can see, approximately 25% turned out to be so-called - fake-audio CDs, produced from lossy audio sources - most likely from low quality MP3s.

One method for auditing such CDs is MPEG artifact searching. The most useful and simple artifact to address is the frequency cut-off, which signal is scarcely affected.or even about 18kHz. The frequency cut-off (produced by an ear-model of the sound, used by the MPEG algorithm) is usually quite sharp at one or more high frequencies (16-20kHz). Music produced from such files is less dynamic, exhibiting fewer sound distinctions (as with drums and other sharp sounds).

Other artifacts of MPEG-coding add a specific type of noise, corresponding to MPEG-coding errors (numerical noise) and Fourier transformations, and a decreased correlation between channels (known as - sound center fluctuations.).

Some studio companies remove live recording noise very inaccurately, by simple cutting-off all the frequencies above 16-18kHz, using digital filtration. The resulting music suffers from similar sound artifacts to those exhibited by MPEG-coded audio recordings.

MPEG-coding errors could be decreased using a dithering technique (a smart-smoothing of the audio). Sounds can be made to seem more distinct through the addition of noise and smoothing of the signal spectrum (a technique known as noise-shaping). Still, despite such techniques, data from the original music is irretrievably lost; instead, one is left with music of inferior quality, after a few processing tricks and the addition of old and new artifacts.

The Aucdtect algorithm was developed to detect whether MPEG artifacts are present in recordings. The program and could be used, for example, to detect inaccurate numerical processing of a music recording that results in the loss of sound quality.