Music recorded through a microphone is a complex mixture of vibrations with varying frequencies and amplitudes. During digitization, an analog-to-digital converter (ADC) chip measures the signal strength and stores this measurement. This process is repeated at high speed (and therefore high frequency), with data being generated during each measurement and sequentially written to a file.
The higher the sampling rate and the finer the resolution of the A/D converter, the better. If, as with CDs, a sampling rate of 44.1 kHz is used, this means that 44,100 individual numbers per second must be recorded in a file. According to the Nyquist-Shannon theorem, this allows for a maximum frequency of Fa/2 = 22 kHz. The accuracy of the digitization is 16 bits per sample. Multiplying this by 2 for both audio channels results in a data volume of approximately 255 Mbit for a three-minute tango. In reality, the data volume is larger because additional error correction and management data are written. Calculating the amount of data per second from this yields the so-called bit rate (2 * 44,100 * 16) of 1411.2 kbit/s (or kBps).
The amount of data on a CD is therefore very large and was very difficult for PCs to handle until the late 1990s. Incidentally, some music tracks, for example for audio DVDs, are digitized with even higher sampling frequencies (96 and 192 kHz).
The size of music files is inconvenient. However, one can simply view the file as a collection of arbitrary data, using concepts familiar from computer science. lossless compression methods To apply this, for example, to compress the bit sequence (1000000001111001), one can write (1, 8*0, 4*1, 001). Such methods significantly reduce the file size and allow it to be correctly restored. Compression rates between 40% and 70% of the original size are possible [4].
One Lossy compression method Developed from 1982 onwards under the direction of Hans-Georg Musmann and Karlheinz Brandenburg at the Fraunhofer Institute for Integrated Circuits in Erlangen, MP3 was incorporated into the MPEG-1 standard and finally released in 1995 as MP3 (ISO MPEG Audio Layer 3). The method uses psychoacoustic information and assumptions, similarities between channels, and statistical predictions to achieve better compression rates. [An in-depth introduction is provided.[3,4]. This allows for compression factors of up to 7x compared to the original file. A measure of quality here is the bitrate. With the classic MP3 method, it can range between 8 and 320 kbps. Generally, bitrates above 96 kbps deliver good sound quality.
The program used for compression and decompression (the so-called codec) determines the quality of music playback, along with other factors such as bitrate, complexity of the music signal, quality of the converter for the raw signal, and, last but not least, the playback technology used. The quality of codecs has improved significantly since the late 1990s [3]. The freely available LAME codec is commonly used for MP3 files today.
Audio files nowadays contain metadata in addition to the music itself. Various information such as title, artist, and year of recording can be stored here to make managing the music tracks easier [5].
The following workflow is therefore required for handling music files:
a) Reading the data stream from CD (so-called grabbing), or alternatively, digitizing an analog recording.
b) Compression with the desired CODEC
c) Inputting metadata (tagging)
d) Playback after simultaneous decompression by the associated CODEC
The following options are available when choosing the CODECS to use.
1. Lossless CODECS
The following formats are common here:
– ATRAC (adaptive transforming acoustic coding) der Firma Sony
– Apple Lossless (.m4a oder .mp4)
– FLAC (Free lossless audio Codec)
– MPEG4 Audio Lossless Coding (ALS)
– WMA lossless (Windows operating system)
FLAC is somewhat widespread. However, compared to uncompressed material, the achievable compression rate is low.
2. Lossy Codes
– MP3 (.mp3)
– Ogg Vorbis (.ogg)
– WMA (Windows Media Audio, .wma)
MP3 and Ogg Vorbis are quite popular in this regard. The WMA format is also common on Windows PCs.
A table with other compression methods can be found here [6]. Some formats also include the option of digital rights management.
3. Quality discussion
While lossy audio files like MP3 with a bitrate of 128 kB/s were considered inferior in 1997, with today’s codecs, the majority of listeners already attribute the attribute “transparent” (indistinguishable) to them. This means that individuals can no longer differentiate between the original recording and the compressed version. Even if some authors describe being able to make this distinction (“golden ear”) spontaneously and flawlessly, a statistical blind test called ABX should be used. In this test, software provides blind audio samples that a listener must characterize. For current codecs, a difference at bitrates higher than 96 kB/s is practically undetectable. To perform the test yourself, you need the playback software foobar2000 with a special ABX plugin for Windows computers [7, 8], a corresponding software is also available for Apple computers.
4. Resume
This lesson was boring, and one wonders what it has to do with tango music. A tango DJ nowadays has to collect their music and store it as audio files. Considering that storage space is cheap today (2015) (standard: 2 TB external disk, 4-6 TB internal), compression is no longer the most important criterion for the approach.
For today’s tango DJ, two other criteria are more important. One is the ability to save metadata, and the other is the best possible compatibility with the software players used for DJ applications. Support for lossless codecs is still limited. It is generally recommended for DJ applications to stick with high-bitrate MP3 encoding [9]. However, I can at least logically understand the point of view of those who support lossless methods.
Abrazos,
-Richard (DJ Ricardo)
PS: In the next episode, we’ll get practical. We’ll be ripping and encoding a tango CD.
1 http://de.wikipedia.org/wiki/Abtastrate
2 http://en.wikipedia.org/wiki/Compact_Disc_Digital_Audio
3 http://de.wikipedia.org/wiki/MP3
4 http://de.wikipedia.org/wiki/Audiodatenkompression
5 http://de.wikipedia.org/wiki/ID3-Tag
6 http://de.wikipedia.org/wiki/Audioformat#Liste_g.C3.A4ngiger_Audioformate
7 http://lifehacker.com/5903625/mp3-or-lossless-see-if-you-can-hear-the-difference-with-this-test
8 http://techland.time.com/2012/03/02/can-you-hear-the-difference-between-lossless-and-lossy-audio
9 http://www.digitaldjtips.com/2011/02/dj-music-files-formats/