To convert a so-called analog audio to digital, we use a process called: sampling. Sampling is done on a converter (or sound card). The principle is to take regular snapshots, which are the measurements of the analog signal voltage, and transform them into digital data whose language is numbers (numbers).
Here is a diagram representing the samples included in the amplitude of a sound wave. The number of samples in this wave defines the sampling frequency or sampling frequency.
Sampling frequency
The sampling rate is expressed in hertz (Hz) or (kHz). The following values are commonly found: 44,100 Hz, 48,000 Hz, 96,000 Hz, 192,000 Hz. The CD and the digital world standard are 44,100 Hz. This means that for every second, there are 44,100 samples. (samples) reproduced. The higher the sample rate (number of “snapshots” of the audio taken per second), the more accurate the analysis and coding of the music in digital data. The sampling rate affects the audio frequency range from the lowest to the highest pitch that can be stored.
Sampling frequency
16-bit / 44.1 kHz coding was the best quality available when the CD was released in the early 1980s, but things have changed, and we can now record and distribute music at higher bit-depth levels and sample rates. These formats have been used in studio recordings and for mastering for many years.
High-resolution audio (HRA) matches any recording format above the 16-bit / 44.1 kHz standard for CDs, and HRA recordings usually use 24-bit encoding, providing a greater dynamic range than CD and sampling rates up to 192 kHz . This is the pinnacle of HRA business records. First and foremost, it’s about getting as close as possible to the sound heard in the studio.
Which sampling rates should you choose?
In order to capture the smallest details at high frequencies, we need to try more frequently. The way it works is that a given sampling rate can accurately detect audio frequencies down to just under half its value. For example, a sample rate of 48 kHz can accurately detect audio frequencies as low as just below 24 kHz. This limit for half the sampling frequency is called the Nyquist frequency and is named after one of the engineers who developed the calculation behind the sampling principle.
The human ear can generally hear in the following spectrum: 20 Hz – 20,000 Hz. As we have just seen, for no obvious loss, the sampling rate must be at least twice as high as the maximum frequency contained in the audio when digitizing. The sampling rate must be at least 40,000 Hz for a correct result for our ears.
This is why 44 100 Hz resolution is the most widely used because it allows us to cover the spectrum up to 22 050 Hz. We even benefit from a small margin because we could have rounded up to 40,000 Hz, but it also means that if you export your music at a sampling rate higher than 44,100 Hz, your ear can’t hear the difference.
Anti alias filters
The first thing an analogue to digital converter does to analogue audio before sampling is to filter all frequencies above the Nyquist limit of the desired sampling frequency. If not filtered, all frequencies above Nyquist are injected again into the sample. This is called an alias effect.
Fortunately, almost all converters on the market today have implemented high-quality anti-aliasing filters. As a result, it seems undesirable aliasing effects are not, and all frequencies below the Nyquist recorded accurately. In most cases, as long as you use a good quality converter and a sampling rate of at least 44.1 Khz, you can record all frequencies in the area of human hearing in an orderly manner. Since the analog to digital converter measures each sample, you have to assign a number to that sample, as that is what makes it digital instead of analog.
How about sound cards up to 192,000 Hz?
There are two benefits to working at a very high frequency:
The first is that the drivers for your sound card (especially professional converters) will be optimized for a given sampling rate. In general, the ASIO drivers for your drives are optimized to the maximum sample rate it offers: 96,000 Hz and 192,000 Hz in most cases. This may be surprising, but it will have less delay and more relief for the microprocessor with a higher sample rate.
