Quality of digital audio in computer music

“Digital Audio” is not synonymous with “audio quality”, because during the creation, storage and copy the audio may deteriorate depending on the process used. Here are some basics for both musicians and audiophiles can ensure quality of their files.

Obtaining and storage

Technically speaking, digital audio is the digital encoding of a sound wave. It consists of a sequence of integers obtained from two processes: sampling (capturing samples of the electrical signal at regular intervals of time) and quantification (convert the level of the obtained sample in a finite and predetermined integer value) .

Both the sampling frequency as the number of quantization bits should be chosen properly to achieve the required quality for their use. For example, the CDs have a sampling rate of 44.1 kHz and 16 quantization bits, which is considered more than enought for the general public. In contrast, a professional studio usually works at 48 kHz and 24 bits.

Anyway, the choice of these values doesn’t ​​ensures the quality of sound, since the use of a low-quality hardware can cause frequency losses and noise due to, for example, the quality of the “antialiasing” filters during sampling or the goodness of “dithering” during quantification.

Once obtained, the digital audio is stored in differents reading and writing physical formats: magnetic (DAT, hard disk), optical (CD), magneto-optical (minidisc, CD-RW) and electronic (USB sticks, memories).The information can be stored continuously (CD-Audio, DAT) or in a random or indirect format (hard disk, memory, CD-ROM)

Digital audio files

Can copying, moving or sending cause losses in digital files?. All devices participating in the copy and digital file storage have systems for detection and elimination of errors. Similarly, when sending files over networks (intranet, internet) or RF (Bluetooth, WiFi), error correction protocols that ensure data integrity are used.

There is only one exception: during recording and playback of audio CDs, a failure called “Jitter” can occur, so do not recommend using this format when you want to preserve audio quality. Instead, a CD-ROM is foolproof against jitter, and is a valid format for storing digital audio without losing quality, in the form of data files.

File Compression

The most widely used file format for storing digital audio is Linear PCM with CD quality: wav files (Windows) or AIFF files (Mac). This type of file occupies too much space, which can be reduced by what is known in informatics as file compression.

There are two types of audio file compression: lossless (like ALAC, FLAC Lossless Audio or WindowsMedia) and lossy (such as AAC, AC3, AMR, ATRAC, MP3, Real Audio or Vorbis).

The complexity of the information contained in the sounds makes it difficult to use lossless compression, only achieving compression ratios between 50% and 60%. This type of compression can be used for professional applications because the original information can always be recovered.

The lossy compression, with ratios between 2% and 30% and based on psychoacoustics, remove information considered negligible for being difficult to grasp by our ears. It should not be used for professional applications, as supposedly insignificant sounds for listening in other areas may be necessary to successfully develop many of the processes taking place in a studio.

Internet Music

Commenting on the quality of audio on the Internet, we must distinguish between the two existing forms for listening to music through this medium: direct download and streaming.

Direct download consists on copying the digital file from a remote computer to ours. As noted earlier, the file copy is a perfect process that does not impair the audio content in it. Low quality mp3 success on the internet lies in the low speed of the connections, but in recent times, with the proliferation of faster connections, is growing the use of  FLAC format. As for the professional field, many studies accept sending audio files over the Internet for their production, mixing, mastering, etc, but always asking for a lossless file as WAV, AIFF or FLAC.

Streaming means that digital audio data is heard  as it downloads and is destroyed a moment later. It is the system used by Internet radios and has the problem that needs to maintain a constant flow of information so that the listener does not suffer “cuts”. Today this can only be ensured by reducing the amount of information as in the lossy compression, so can say that streaming cannot yet be used at professional level.

Conclusions

To obtain high-quality digital audio must start with a good hardware that has good anti-aliasing filters and a good dithering. The process has to be done at least at 44.1 kHz/16 bits for general use and 48 kHz/24 bit for professional use.

The resulting file can be saved, moved and copied without problems, but is not recommended burning in CD-Audio format. You can also compress whenever you use a lossless compression format.

Finally, audio can also be sent over the Internet without losing quality when not done through streaming.

Following these guidelines we can ensure that our audio maintain the highest quality at all times.