In an article by Yu. Kuznetsov, M. Morozov and A. Shityakov under this title (“Radio”, 1985. No. 1, pp. 27h; 28) a description was given of a device that, despite its relative simplicity, could perform two functions at once :

• reducing the level of rumble;
• Improved stereo channel separation.

At the same time, as the analysis of circuit solutions has shown, a similar device can be built using a significantly smaller number of passive and active elements. A schematic diagram of such a regulator, developed by the author, is shown in Fig. 1.

Main technical characteristics:

• Rated input voltage, V.....0.5;
• Input resistance, kOhm, not less.... 70;
• Voltage transfer coefficient.... 1;
• Maximum stereo expansion, times.... 2;
• Level of suppression of low-frequency anti-phase rumble components in the “Stereo” mode (maximum extended stereo base), dB, at frequencies, Hz: 8. . .26(20) 20 . .18(12); 50....... 10(4);
• Harmonic coefficient at rated input voltage, %, no more. 0.05;
• Signal to noise ratio. dB, not less... 80;
• Overload capacity, dB, no less. 20;
• Current consumption, mA, no more. 7.

The device operates on the principle of sequential sum-difference conversion of a stereo signal. It consists of a node for summing the signals of the left and right channels on a dual variable resistor R1 and a node for difference signal conversion on two op-amps DA1 and DA2. In the first, the width of the stereo base is adjusted from the nominal value to zero, the second expands the stereo base by a fixed value.

The serial connection of these nodes made it possible to obtain a device that regulates the stereo width from zero to the maximum value determined by the difference transformation node.

The functions of the rumble filter in the described design are performed by transition capacitors C1 and C2 together with stereo width regulators R1.1 and R1.2.

In the right (according to the diagram) position of the resistor R1 sliders, the non-inverting inputs of both op-amps are combined, and, therefore, the signals at their outputs are equal in value to the half-sum of the signals of both channels (Mono mode).

In the left position of the sliders of this resistor, the signals at the inputs of the op-amp DA1 and DA2 are equal to the signals of the left and right channels, respectively, arriving at the input of the entire device, and the difference conversion unit expands the stereo base to the maximum value. In the intermediate position of the resistor R1 sliders, the voltages at the non-inverting inputs of the op-amp are determined by the expressions: Ul1 = Ul0 (1 - a/2) + Ul0 * a / 2, Up1 = Ul0 * a / 2 + Up0 (1 - a/2) for R2> =R1, and the output voltages of the entire device:

• Ul2 = Ul1(1+R/R4) - Up1 * R/R4;
• Ul2 = - Ul1R/R4 + Up1(1 + R/R4);

where R=R3=R5; Ul0, Ul1, Ul2 (Up0, Up1, Up2) are the voltages of the left (right) channel acting respectively at the input of the regulator, the non-inverting inputs of the op-amp DA1 and DA2 and at the output of the device, and is the relative movement of the sliders of the variable resistor R1 (a=0. ..1). In particular, when choosing the ratio R/R4 = 0.5 (which corresponds to the maximum expansion of the stereo base by a factor of two), in the middle position of the resistor R1 sliders (a = 0.5), a stereo signal Ul2 = = Ul0, Up2 = Up0 is formed at the op-amp outputs and the stereo base has a nominal value.

When deriving relationships for the voltages acting at the outputs of the regulator, it was assumed that the output resistances of the signal sources of the left and right channels are significantly less than the resistance of resistor R1. If this condition is not met, the influence of the output resistances of the signal sources Rg on the operation of the device can be taken into account by formally replacing in the expressions to determine Ul2, Up2 variable a to a new variable a1 = (a+RgR1) / (1+Rg/R1).

In practice, the influence of the output impedances of signal sources comes down to a decrease in the range of stereo adjustment in the direction of its expansion, and this influence can be practically eliminated by a corresponding increase in the ratio R/R4.

Rice. 1. Schematic diagram of the stereo width control.

Rice. 2. Printed circuit board for stereo width control.

In the low frequency region (f<= 1/2пR1C1 = 1/2пR1C2 = ~ 160 Гц) зависимость напряжения гармонического сигнала, лежащих в области частот значительно ниже 160 Гц, напряжения на входах ОУ приближаются к полусумме сигналов левого и правого каналов.

For antiphase rumble components (Ul0=-Un0), the device’s transmission coefficient is described by the following expression: K (jw) = Ul2 / Ul0 = Up2 / Up0 = 2jwR1C(1-a) / (1 + jwR1C), where C=C1=C2; w - circular rumble frequency.

Resistor R2 sets the operating mode of the op-amp for direct current and, together with capacitors C1 and C2, determines the lower cutoff frequency of the device’s frequency response. With the element ratings indicated on the diagram; Moreover, its technical characteristics such as harmonic distortion, signal-to-noise ratio, and current consumption are determined only by the op-amp used.

Dual variable resistor R1 - any, with the adjustment characteristic of group A. The resistance of resistor R1 may be different from the value indicated on the diagram. The capacitance of capacitors C1 and C2 in this case must be changed in inverse proportion to the magnitude of the change in the resistance of resistor R1.

Drawing of the printed circuit board of the device, designed for the installation of MLT resistors (R2 - R5). SPZ-23b (R1), KM capacitors (C1, C2)) is shown in Fig. 2.

A properly assembled device does not require any adjustment.

It should be noted that due to the low output impedance, the regulator interfaces well with other functional units of the stereo circuit.

M. Starostenko. Miass, Chelyabinsk region. Radio 1989, 11.

To begin with, it should be noted what a stereo base is, and why is it needed? For me, this is one of the components of high-quality sound. This is the volume of your soundtrack. Any composition mixed in a studio by a professional sound engineer has sufficient stereo amplitude, which gives it liveliness and freedom to fly. The sound becomes more transparent and velvety. This is when a sound wave does not propagate in a straight line, but seems to cover the entire space. And so, we have decided on the concept of a stereo base. But how can you achieve such a sound? Is this quality really only possible when mixing in a studio? No, this can be achieved at home, and in two ways. The first method is software, that is, using plug-ins. The second method is manual, that is, using a panorama and a little trick.

Method one.

Currently, there are many plug-ins and VST instruments capable of expanding the stereo base of a phonogram in one way or another. There is such a wonderful tool called Virtual Guitarist (Fig. 1), which simulates the playing of a live guitarist, in which this function is built-in, and naturally applies only to your own sound. Using the stereo knob, you can either expand the sound or reduce it to mono. If the guitar part in the mix serves as a melody, then it is advisable to make the sound<<шире>>, for more volume and beauty. If the guitar part keeps the rhythm, then the sound needs to be made clearer, and try not to resort to this function.

Fig.1

Fig.2

The CuBaseSX3 sequencer has a built-in synthesizer called Embracer (Figure 2), in which the stereo base can be expanded in parallel with the audio tone. Of all the plugins with which you can expand the stereo base of a phonogram, you can especially highlight the Izotope Ozone mastering package, which copes with its responsibilities in the best possible way. He does this by dividing the individual frequencies of the sound wave across the panorama. It is also possible to equalize the sound using a four-band equalizer. To better understand what we are talking about, let’s look at everything using examples:

Example 1.
This is what a sound wave looks like on the Izotope Ozone diagram when playing pure, unprocessed sound. As you can see in the figure, the sound propagates strictly in the center, and there will be no volume to the ear. If you separate individual frequencies across the panorama, and not their entire spectrum, then the difference will be noticeable even to an inexperienced listener.

Example 2.
This is what a sound wave looks like on the Izotope Ozone diagram when the low and mid frequencies are spread across the panorama by 10% into the right and left channels. In this case, low and mid frequencies in relation to the main sound actually play at different times, but since they are separated into channels, our ears hear them separately, and the brain already collects them into a single stereo picture. The main thing is not to overdo it. If the stereo base of a mix is ​​too wide, it can lose its clarity and dynamics.

Method two.

If you don’t have a suitable plugin, you can expand the stereo base manually, using any audio editor with a multi-track recording mode, or you can use any sequencer. The example uses CuBase SX.
First, we need the sound that we will process. Import this track into CuBase (File -> Import -> Audio File). Duplicate the original track, after which we should have it, as in Figure 3.

Fig.3

Next, we take the upper path to the left in the panorama, and the lower one to the right. At the same time, we impose a delay of 10 units on the lower track. But you can deviate from this figure, but the main thing is not to overdo it. That’s it, let’s combine these tracks into one file (File -> Export -> Audio Mixdown), and only in stereo, otherwise the work done will be of no use.

It should be noted that there are basic instruments that need this and those that absolutely do not need stereo phase extension.
Guitar.
The guitar is the instrument that most often undergoes this treatment. This does not depend on the type of guitar, be it acoustic or electric, nor on the style of performance. In rock music, guitar rhythm parts are the basis of rigidity and dynamics, and should cover almost the entire sound area; if it is a lyrical arpeggio part, then the width of the stereo phase emphasizes the beauty and elegance of the melody.
Piano.
Expanding the stereo range of the piano won't hurt, but you can do without it. It all depends on the instrument's place in the mix, and the style of music being played.
Violin players.
It is desirable that the violin sounds<<широко>>, but here everything again depends on the violin part and the style of the musical composition.
Vocals
Backing vocals are subject to mandatory stereo widening, otherwise they will sound like the main voice; by expanding them, we leave room for the lead vocals. But you can hear this treatment on solo vocals as well.
Bass.
It is strictly forbidden to expand the stereo base of the bass guitar, since it is the foundation of the composition and must be strictly in the center, this applies to live styles. In electronic music, everything is possible, for example, expanding the stereo base of the bass is often found in the Trance style, since the drum rhythm section is the basis there.
Drums.
Again, in live music they should be strictly in the center; you should only separate the individual components across the panorama.

For stereo playback systems with closely spaced pairs of loudspeakers, it is advisable to use a stereo expansion device. In this case, the subjective perception of the sound picture changes and a spatial impression is created about the localization of sound sources outside the acoustic systems. The sound of cassette and portable radios, tape recorders and tape recorders is improving. Recently, stereo expansion devices have begun to be used even in televisions with stereo playback. The spread of such devices is also facilitated by the fact that Philips has developed a special sound processor to expand the stereo base and create a pseudo-stereophony effect - LSI TDA3810. The devices discussed below use the principles inherent in this successful microcircuit.

The schematic diagram of the stereo base expander is shown in Fig. 10.3. The effect of expanding the stereophonic base (the width of the stereophonic sound picture) is achieved by

Rice. 10.3. Stereo expander

cross-add the filtered left and right channels. The cutoff frequency of the filters of the left and second channels is

is expressed by the relation

As a result, the mid signal is attenuated, which emphasizes the mono sound effect. Input payout coefficient

left channel before its output is

, and the transmission coefficient from the left channel input to the right channel output is negative and equal to

Therefore, below 4 kHz the stereo base

is expanding. Dividers R1R3 (R2R4 in another channel) are necessary so that the signal levels at the input and output of the device are the same.

Rice. 10.4. Printed circuit board and placement of stereo expander elements

All parts of the stereo base expander are placed on a printed circuit board measuring 44 x 41 mm made of fiberglass 2 mm thick (Fig. 10.4). It is best to start installing parts on it by installing two jumpers made of tinned copper wire with a thickness of 0.6...0.7 mm. To avoid problems during setup, before installation, select the expander elements with an accuracy of 3...5%: R1 = R2, R3 = R4, R6 = R7, C2 = NW.

Stereo width control

STEREO BASE WIDTH REGULATOR – ROCKET FILTER

According to research, the elimination of rumble (out-of-phase low-frequency components) together with the expansion of the stereo base (amplification of the difference signal in the midrange and high frequency) psychoacoustically improves the sound, all other things being equal, more than reducing the nonlinear distortions of the UMZCH from 0.1 to 0.01%.
Therefore, the introduction of an appropriate device into the amplification path is very justified and appropriate.

I present to your attention a diagram of such a regulator, developed back in 1989 by M. Starostenko, Miass, Chelyabinsk region. ("RADIO", 1989 No. 11, pp. 52-54). The regulator was assembled by me and a number of my friends, and has proven itself only on the good side.

Main technical characteristics:

Nominal sensitivity. . . . . . . . . . . . . . . . . . . . . . . . . . 0.5V
Input impedance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . not less than 70 kOhm
Voltage transfer coefficient. . . . . . . . . . . . . . . . . . . 1
Maximum stereo expansion. . . . . . . . . . . . . . . . . . . . up to 2 times
Expansion Rumble Suppression Level
stereo base 1 time (2 times), dB, at frequency, Hz
8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 (20)
20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 (12)
50 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 (4)
Nonlinear distortions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . no more than 0.05%
Signal to noise ratio. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . not less than 80 dB
Overload capacity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 dB
Consumption current. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . no more than 8 mA

The device operates on the principle of sequential sum-difference conversion of a stereo signal. The adder of the signals of the left and right channels is assembled on the double resistor R1, and the difference conversion unit is assembled on the op-amps DA1 and DA2. C1 and C2, together with R1.1 and R1.2, perform the functions of a rumble filter. In the extreme left (according to the diagram) position of the R1 sliders we have the maximum expansion of the stereo base (and minimal rumble suppression), in the right - the “MONO” mode with maximum rumble suppression. In the middle position of the sliders at the output of the device we have a normal stereo signal with good rumble suppression.
Variable resistor R1 - with a linear dependence. A correctly assembled product does not require adjustment.

Some tips from VEPR:

If the K544UD1A op-amp is replaced with a K544UD2B, then nonlinear distortion will decrease to 0.02-0.03%, and if with LF351(LF147) - then to 0.02%. Experience using this regulator has shown that MONO mode is not needed. Yes
and the rumble in those days came mainly from the “turntables” (EPU) - where they are now. Therefore, you can take R1 = 33 kOhm, and between its right (according to the diagram) terminals and R2 include a constant 33 kOhm resistor. Many simply put it on input constant resistors 68 kOhm - saving controls and space
on the control panel.

Danich Alexander, aka VEPR.
Nizhyn, Chernigov region.

02/06/2015 at 12:00

There are many methods for widening the stereo in a mix, all of which together produce excellent results. This includes panning, reverb, various effects, playing with phase, delay between the right and left channels, Mid/Side processing, and even saturation and compression can give us width. There is a lot of controversy about the use of these methods, as the output can be a mess or a lot of antiphase, which can make our mix unreadable in certain sources. Of course, when recording live instruments there is no such problem, as a rule, the sound is recorded with several microphones, the sound source itself and its surroundings (room) are recorded. With this recording method, there is no need to apply any effects, it is enough to set the required levels, clean and tighten up for readability of the original signal. But when working with already recorded fragments (samples), the picture is completely different. When using samples mixed with room and tails, we do not have the ability to fully control the stereo image of our mix, which gives a very bad result.

In the book " Beatmaker's Bible“I was attracted by the method that sound engineers used before; it consists of dividing the channel into 2, separated all the way to the left and right, oppositely duplicated with phase inversion. This method is called " Left minus right"and was used on analog mixers.

The huge advantage of this method is the preservation of mono compatibility, that is, when listening in mono, the sound is not distorted, but gives a good result. You can repeat it in your virtual mixer, but keep in mind that the levels should change strictly in parallel and directly proportional to the amount by which you change the levels of the left channel, by the same amount change the level of the right. (see above)

IN FL Studio There is a plugin that simulates this circuit - Fruity Stereo Shaper. « Fruity Stereo Shaper. A stereo effect with the ability to separately control the volume of the left, right channels and their inverted copies, as well as their delay and phase shift. Appeared in FL Studio 9 » - Wikipedia. The scheme of operation of this plugin is exactly the same as “ Left minus right", in addition, it has several convenient controls and 2 operating modes - PRE And POST, which imply a place in the processing chain, according to the name of the mode.

I use this plugin on a snare drum, as it is very difficult to achieve its liveliness and depth, and this is very important, important also for the reason that in the mix you should leave a small margin for vocals in the middle, otherwise they will conflict. I put it after the equalizer, in mode PRE, the way it actually processes the still “dry” sound, I twist the knob PHASE on 90-100% to the left, followed by shifting the channel in the mixer to the left by 5-7% , turning the stereo knob to 100% . It looks illogical, but when these offsets are summed up, you can’t hear it at all, however, the snare drum sounds much wider, without any antiphase. I also route the snare to a send channel with room reverb, which gives even more depth. This plugin can also be used on the group channel of all elements of the rhythm section, after all the processing, for me personally it is: saturation, filtering, compression. After all the processing, I put this plugin in the mode POST, handle PHASE all the way left or right, by ear, and I get an even wider sound, but this time for the entire rhythm section. It's important to remember that, the larger the phase shift, the less it is audible and the less its influence on the stereo field. Try, experiment, there are no strict rules in modern production.