Sample rate and bit depth
When a signal reaches the ADC from a preamplifier, compressor, console output, synthesizer, it represents electromagnetic oscillations.
That is, a certain wave with a variable voltage (very small values) reaches the input of the ADC. To save a signal to a file, it must be “digitized,” that is, encoded by ones and zeros. The result is a graph of the wave on the computer screen.
Even the best transducer has an error, because there are no intermediate values between zero and one, and the wave graph will only consist of vertical and horizontal segments, with no oblique lines. The graphical representation of the wave will be influenced by the pitch (oscillation frequency), its timbre (waveform) and the volume (amplitude). A high-quality ADC must correctly transmit all these parameters to the recording system.
So the sound enters the system discreetly, that is, divided into small segments. The precision of encoding an analog signal in a digital environment depends on the size of these segments. The smaller the horizontal and vertical discrete units, the more accurate the scan will be.
Sampling rate
Splitting the wave horizontally gives us an idea of the sample rate or sample rate. The more often the ADC detects changes in waveform values, the higher the sample rate. In reality, a sample is a discrete unit segment, the smallest unit of sound. The shorter it is, the higher the sample rate.
For example, a sample rate of 44.1 kHz indicates that there are 44,100 samples per second of recording. We can edit the wave, taking a segment with a duration of 1/44100 seconds as the minimum editing element. As the sample rate increases to 48 kHz, this section drops to 1/48000 of a second, allowing for more accurate impact.
Each sample is the same length as the previous one. For proper sound reproduction, the file and system sample rates must be identical. When an audio track with a different sample rate than the host (program) sample is added to the project, it must be converted.
If you play a file with a higher frequency on a lower system, it will sound slower than it should, and vice versa. Converting a signal from one frequency to another always produces distortion. To “reshape” the sound to the new sample rate, the system must divide the samples into smaller pieces and reassemble them into a single wave. Such a process can lead, at best, to simply blur the sound, at worst, to the appearance of clicks.
Of course, in the built-in speakers of a home laptop, the difference will not be noticeable. But when it comes to working with sound at a professional level, sample rate coordination is necessary.
It is not recommended to change the sample rate within the same project. A justification for higher sampling could be, for example, the need to process the file with algorithms or plugins that work better at high frequencies. Since a higher sample rate means dividing into smaller samples, the processing precision will be higher and the result will be of better quality. But it is also impossible to guarantee the effectiveness of this method: in each case the result will be individual. It is necessary to evaluate each time what is more important: the effect of processing at a higher resolution or the negative impact of the conversion.
If for some reason, after completing the job at 48 kHz, you need to convert the signal to 44.1 kHz, save the original file in case you need to re-manipulate the material (for example, for alternative mastering). Processing at a higher sample rate will produce a better effect than processing at a lower sample rate.
