Lossless digital audio formats
Uncompressed formats
The first digital audio formats encoded information, as they say, “as is”.
This format is, first of all, convenient for the computer itself: there is a simple correspondence between a sequence of digital signals and a sound waveform. Not surprisingly, despite the further development of compressed audio formats, uncompressed formats continue to be actively used in the field of professional (and amateur) sound work. What are these formats?
First, it is the audio CD format itself. For him everything is simple: this is the format in which most audio discs are recorded (unless, of course, several albums have been recorded on them at the same time – in this case, we are talking about a format obviously compressed). The Audio CD format uses Pulse Code Modulation (PCM) to convert “digital” to “analog.” This is the type of conversion that forms the basis of most other audio formats.
Everything is clear with CDs, but what about sound presentation formats on computers? Here, as you can imagine, the biggest computer and software manufacturers were the first to rush in, developing formats like AIFF (Apple) and WAV (Microsoft and IBM), known to computer and operating system users of these developers to this day. from today. Furthermore, Apple, being the undisputed leader in the field of computer multimedia at the time, introduced its format in 1988, ahead of the developers of the WAV format by three years. If you let the computer read the data on the audio CD and translate it into any of these formats, you will get an exact copy, which will take up the same amount of space on your computer’s hard drive as the original on the CD. The only difference will be in the format in which this data will be packed.
It is important to note that the capabilities of these formats exceed those of an audio disc. They allow much greater audio depth (up to 32 bits) and sample rates (AIFF up to 196 kHz, and in WAV, any). But even these parameters were not satisfied with the developers of ultra-high fidelity audio formats. These formats were first used on Super Audio CD (SACD) developed in 1999 by Sony and Philips, and were eventually installed on personal computers for lovers of high-quality sound.
The digital audio format used in SACD is called DSD (Direct Stream Digital). It uses a fundamentally different way of analog-to-digital and digital-to-analog conversion – the waveform is encoded with just one bit at a time. In other words, the “depth” of the sound is only one bit, but this information is updated much more frequently than in most other formats, that is, the sample rate is increased several times. This allows the algorithm to recreate the shape of the sound wave by estimating the density of these pulses in a given time interval.
This way of representing the signal made it possible to avoid non-linear quantization noise. It is true that the higher the frequency of the sound, the lower the number of pulses that encodes it, the precision of the encoding decreases, which means that the constant noise level in relation to the signal, that is, the hiss, increases . But at a high sampling frequency, the level of this hiss becomes critical already in the ultrasound region, which is imperceptible to the ear. At the same time, a much more important task has been accomplished: there is no quantization noise, a kind of “curse” of digital sound. They all “left” with the same inaudible noise. It is not surprising that this format fell in love with critics of “digital”, it seemed to bring it closer to “analog”, eliminating traces of discretion in its sound.
However, it turned out that DSD files are completely unsuitable for editing: they are edited only by converting to PCM (and vice versa when saved), and each such conversion multiplies the amount of noise that soon becomes critical for hearing. . It’s one thing when you record ready-to-use analog material in DSD, which has passed analog mastering. But what should you do when recording and editing new audio material? So, there were formats that went back to pulse code modulation with additional adjustments against quantization noise. Notable among these is DXD (Digital eXtreme Definition), which has found application not only for mastering source data for SACD, but also for listening to high-quality audio on a computer.
