Music Creation. Mastering - part 2

I wrote about the fact that mastering in the process of music production is the last step on the way from the idea of ​​music to its delivery to the recipient in the previous issue. We have also taken a close look at digitally recorded audio, but I have not yet discussed how this audio, converted to AC voltage converters, is converted to binary form.

1. Every complex sound, even a very high degree of complexity, actually consists of many simple sinusoidal sounds.

I ended the previous article with the question, how is it possible that in such an undulating wave (1) all musical content is encoded, even if we are talking about many instruments playing polyphonic parts? Here is the answer: this is due to the fact that any complex sound, even very complex, is really it consists of many simple sinusoidal sounds.

The sinusoidal nature of these simple waveforms varies with both time and amplitude, these waveforms overlapping, adding, subtracting, modulating each other, and so first the sounds of individual instruments are created, and then full mixes and recordings.

What we see in figure 2 are certain atoms, molecules that make up our sound matter, but in the case of an analog signal there are no such atoms - there is one even line, without dots marking subsequent readings (the difference can be seen in the figure in as steps, which are graphically approximated to obtain the corresponding visual effect).

However, since playback of recorded music from analog or digital sources must be done using a mechanical electromagnetic transducer such as a loudspeaker or headphone transducer, the difference between pure analog audio and digitally processed audio blurs is overwhelming in most cases. At the final stage, i.e. when listening, the music reaches us in the same way as the vibrations of air particles caused by the movement of the diaphragm in the transducer.

2. Molecules that make up our sound matter

analog digit

Are there any audible differences between pure analog audio (i.e. recorded analog on an analog tape recorder, mixed on an analog console, compressed on an analog disc, played back on an analog player and amplified analog amplifier) ​​and digital audio - converted from analog to digital, processed and mixed digitally and then processed back to analog form, is that right in front of the amp or practically in the speaker itself?

In the vast majority of cases, rather not, although if we recorded the same musical material in both ways and then played it back, the differences would certainly be audible. However, this will be due rather to the nature of the tools used in these processes, their characteristics, properties, and often limitations, than the very fact of using analog or digital technology.

At the same time, we assume that bringing the sound to a digital form, i.e. to explicitly atomized, does not significantly affect the recording and processing process itself, especially since these samples occur at a frequency that - at least theoretically - is far beyond the upper limits of the frequencies we hear, and therefore this specific graininess of the sound converted to digital form, is invisible to us. However, from the point of view of mastering the sound material, it is very important, and we will talk about it later.

Now let's figure out how the analog signal is converted to digital form, namely zero-one, i.e. one where the voltage can have only two levels: the digital one level, which means voltage, and the digital zero level, i.e. this tension is practically non-existent. Everything in the digital world is either one or zero, there are no intermediate values. Of course, there is also the so-called fuzzy logic, where there are still intermediate states between the “on” or “off” states, but it is not applicable to digital audio systems.

3. Vibrations of air particles caused by a sound source set in motion a very light structure of the membrane.

Transformations Part One

Any acoustic signal, be it vocals, acoustic guitar or drums, is sent to the computer in digital form, it must first be converted into an alternating electrical signal. This is usually done with microphones in which vibrations of air particles caused by the sound source drive a very light diaphragm structure (3). This can be the diaphragm included in a condenser capsule, a metal foil band in a ribbon microphone, or a diaphragm with a coil attached to it in a dynamic microphone.

In each of these cases a very weak, oscillating electrical signal appears at the output of the microphonewhich to a greater or lesser extent preserves the proportions of frequency and level corresponding to the same parameters of oscillating air particles. Thus, this is a kind of electrical analogue of it, which can be further processed in devices that process an alternating electrical signal.

At first microphone signal must be amplifiedbecause it is too weak to be used in any way. A typical microphone output voltage is in the order of thousandths of a volt, expressed in millivolts, and often in microvolts or millionths of a volt. For comparison, let's add that a conventional finger-type battery produces a voltage of 1,5 V, and this is a constant voltage that is not subject to modulation, which means that it does not transmit any sound information.

However, DC voltage is needed in any electronic system to be the source of energy, which will then modulate the AC signal. The cleaner and more efficient this energy is, the less it is subject to current loads and disturbances, the cleaner the AC signal processed by the electronic components will be. That is why the power supply, namely the power supply, is so important in any analog audio system.

4. Microphone amplifier, also known as preamplifier or preamplifier

Microphone amplifiers, also known as preamplifiers or preamplifiers, are designed to amplify the signal from microphones (4). Their task is to amplify the signal, often even by several tens of decibels, which means to increase their level by hundreds or more. Thus, at the output of the preamplifier, we get an alternating voltage that is directly proportional to the input voltage, but exceeding it by hundreds of times, i.e. at a level from fractions to units of volts. This signal level is determined line level and this is the standard operating level in audio devices.

Transformation part two

An analog signal of this level can already be passed digitization process. This is done using tools called analog-to-digital converters or transducers (5). The conversion process in classic PCM mode, i.e. Pulse Width Modulation, currently the most popular processing mode, is defined by two parameters: sampling rate and bit depth. As you rightly suspect, the higher these parameters, the better the conversion and the more accurate the signal will be fed to the computer in digital form.

5. Converter or analog-to-digital converter.

General rule for this type of conversion sample, that is, taking samples of analog material and creating a digital representation of it. Here, the instantaneous value of the voltage in the analog signal is interpreted and its level is represented digitally in binary system (6).

Here, however, it is necessary to briefly recall the basics of mathematics, according to which any numerical value can be represented in any number system. Throughout the history of mankind, various number systems have been and are still used. For example, concepts such as a dozen (12 pieces) or a penny (12 dozen, 144 pieces) are based on the duodecimal system.

6. Voltage values ​​in an analog signal and representation of its level in digital form in a binary system

For time, we use mixed systems - sexagesimal for seconds, minutes and hours, duodecimal derivative for days and days, seventh system for days of the week, quad system (also related to duodecimal and sexagesimal system) for weeks in a month, duodecimal system to indicate the months of the year, and then we move to the decimal system, where decades, centuries and millennia appear. I think that the example of using different systems to express the passage of time very well shows the nature of number systems and will allow you to more effectively navigate issues related to conversion.

In the case of analog to digital conversion, we will be the most common convert decimal values ​​to binary values. Decimal because the measurement for each sample is usually expressed in microvolts, millivolts and volts. Then this value will be expressed in the binary system, i.e. using two bits functioning in it - 0 and 1, which denote two states: no voltage or its presence, off or on, current or not, etc. Thus, we avoid distortion, and all actions become much simpler in implementation through the application of the so-called change of algorithms with which we are dealing, for example, in relation to connectors or other digital processors.

You are zero; or one

With these two digits, zeros and ones, you can express every numeric valueregardless of its size. As an example, consider the number 10. The key to understanding decimal-to-binary conversion is that the number 1 in binary, just like in decimal, depends on its position in the number string.

If 1 is at the end of the binary string, then 1, if in the second from the end - then 2, in the third position - 4, and in the fourth position - 8 - all in decimal. In the decimal system, the same 1 at the end is 10, the penultimate 100, the third 1000, the fourth XNUMX is an example to understand the analogy.

So, if we want to represent 10 in binary form, we will need to represent an 1 and a 1, so like I said, it would be 1010 in fourth place and XNUMX in second, which is XNUMX.

If we needed to convert voltages from 1 to 10 volts without fractional values, i.e. using only integers, a converter that can represent 4-bit sequences in binary is sufficient. 4-bit because this conversion of a binary number will require up to four digits. In practice it will look like this:

0 0000

1 0001

2 0010

3 0011

4 0100

5 0101

6 0110

7 0111

8 1000

9 1001

10 1010

Those leading zeros for the numbers 1 to 7 simply pad the string to the full four bits so that each binary number has the same syntax and takes up the same amount of space. In graphical form, such a translation of integers from the decimal system to binary is shown in Figure 7.

7. Convert Integer Numbers in Decimal System to Binary System

Both the upper and lower waveforms represent the same values, except that the former is understandable, for example, for analog devices, such as linear voltage level meters, and the second for digital devices, including computers that process data on such language. This bottom waveform looks like a variable-fill square wave, i.e. different ratio of maximum values ​​to minimum values ​​over time. This variable content encodes the binary value of the signal to be converted, hence the name "pulse code modulation" - PCM.

Now back to converting a real analog signal. We already know that it can be described by a line depicting smoothly changing levels, and there is no such thing as a jumping representation of these levels. However, for the needs of analog to digital conversion, we must introduce such a process to be able to measure the level of an analog signal from time to time and represent each such measured sample in digital form.

It was assumed that the frequency at which these measurements would be made should be at least twice the highest frequency that a person can hear, and since it is approximately 20 kHz, therefore, the most 44,1kHz remains a popular sample rate. The calculation of the sampling rate is associated with rather complex mathematical operations, which, at this stage of our knowledge of the conversion methods, does not make sense.

More is it better?

Everything that I mentioned above may indicate that the higher the sampling frequency, i.e. measuring the level of an analog signal at regular intervals, the higher the quality of the conversion, because it is - at least in an intuitive sense - more accurate. Is it really true? We will know about this in a month.