Compressed Audio Files Will Never Sound As Good

Audio Tips , Interviews



Even though listeners have become used to the thin sound of compressed music played through tiny computer speakers, they complain that it really becomes uncomfortable when they turn up the volume in order to hear more musical details. Loudness wars take place when we try to get attention. We speak louder and louder. Even if we do not understand the language being spoken, we pay more attention to louder voices, they seem more exciting. But with music, the excitement comes from variations in loudness, pitch, rhythm and timbre. Held constant, any one of those parameters seems monotonous. Constant loudness bores the brain and listeners skip to another song. The problem is that all the songs on their computers are compressed by definition, so they become frustrated. Listeners do not realise that compressed audio files eliminate much of the data from the original compact disc. They are created by compressing the music into a smaller file by excluding the musical data that the ear is less likely to notice, mainly at the very low and high frequencies. This leave the music sounding brittle, flat, hollow, tinny and without punch.



1. Two-dimensional flat soundstage. The lush original is reduced to a cardboard replica, having lost its ambience, delicate subtleties, depth, realism, richness, sparkle, three-dimensional space and warmth.


2. Added sizzle and mushiness. The artificial weird bubbling and phasing, unnatural swirling metallic noises that sound like someone has added twinkling chime bars to everything, or that there is a mosquito buzzing in your ear. Fuzzy sounds swirl around in a muddled mess.


3. Added distortion. Codecs (the systems used to implement compression in audio files) rarely include any headroom for the encoding process itself, so the added processing pushes the music even further over the limits, adding distortion by generating inter-sample peaks.




High quality audio means very large audio files. Audio file compression reduces the audio file size but almost all compression codecs are lossy, which means that audio quality is lost every single time the file is saved. The most common lossy compression codecs are:


1. AAC (Advanced Audio Coding). This is based on the MPEG4 standard owned by Dolby. The Apple iTunes Music Store uses a copy-protected version of AAC. Note: Most music downloaders do not bother checking the iTunes preferences and go for the default settings (160 kbps). Even for those who change the preferences to use 320 kbps (60% space saving), the sound is still worse than the source.


2. MPEG Layer-3. This is the most popular format for music downloading. By eliminating portions of the audio file that are essentially inaudible, MPEG Layer-3 files are compressed to roughly 10% of the size of an equivalent standard, uncompressed, CD-quality, PCM audio file.


Do not confuse the audio file type with its codec. The type is determined by the file extension (what comes after the “.” in the file name) whereas the codec is the way the audio file is compressed and stored (which, in turn, determines its size). The most common audio file types are AIF and WAV, the standard, uncompressed, CD-quality, PCM audio file formats used by Apple and PCs respectively. The most common lossless compression codec is FLAC. An audio file compressed with FLAC and restored later will be a perfect copy of the original.


Written by Daniel Fournier – January 2011 – CPARIS®