Expert evaluation of sound quality control. Methods for expert assessment of the sound quality of recordings. Boris Meyerzon. Disadvantages of objective methods

The subjective assessment of the sound quality is based on recommendations,

developed by the international organization of radio and television OIRT (OIRT - Organization International Radio and Television) to enable successful

international exchange of radio and television programs ( Meyerson B ., article vzh - le « sound engineer » № 8 from 1999 G .)

the following parameters:

1) spatial impression; (Eng. Spatial Impression). This setting

is evaluated by the student's (expert's) impression of the acoustic situation in the studio (hall) transmitted in the recording, the size of the studio, the number of performers and the nature of the musical work, the time and nature of the reverberation, as well as the acoustic balance, i.e. the ratio of direct and reflected sounds .

An important advantage of musical recordings in evaluating the spatial impression is the feeling of a sound perspective in the depth and width of the panorama, that is, the illusion of different distances from the listener to certain groups of performers, the feeling of the versatility of the sound picture, recreating the volume of sound, which is especially lost in monophonic recordings.

However, if multi-dimensionality is replaced by so-called multi-dimensionality, this should be considered a disadvantage of sound engineering work. The latter term is usually understood as such a feeling of the sound of various instruments, as if they were located in different rooms with different acoustic properties. Multi-space, if it is not specifically provided for by the director's plans to create the necessary mise-en-scenes, is perceived as a significant violation of the naturalness of sound transmission. The reason for a lot of surround sound can be: an unsuccessful location of microphones in the studio (with a poly-microphone recording method), as well as an unsuccessful use of artificial reverb.

2) Transparency: (eng. Transparency) is defined as a separate perception of each of the sound components of the sound picture, listening to all sound lines of the score, clarity of musical texture, speech intelligibility, clarity of diction.

The transparency of the sound largely depends on the skill of the sound engineer: the nature of the microphone reception during recording, the sound balance being set, the signal processing used (spectral, dynamic and spatial), etc.

Of no small importance is the process of editing all the phonograms that form the overall sound picture. With monophonic sound recording, as a result of the greater effect of mutual masking of signals, it is much more difficult to achieve transparency than with stereophonic sound.

3) Musical balance- this is the semantic balance of the loudness of the sound

individual groups of instruments and/or soloists in a common listening sound

picture, sound balance.

4) Timbre(eng. Sound color, Timbre) - one of the important parameters of the subjective assessment of the quality of the phonogram; a specific coloration of sound, thanks to which sounds of the same loudness and pitch can be distinguished from each other.

The quality of timbre transmission depends on the location of the performers and microphones in the studio, the nature of the studio acoustics, the frequency response of the sound transmission and sound recording path, the nature and amount of reverberation.

The timbre changes significantly in the presence of a large number of non-linear distortions in the path (for example, when the input channel of the mixing console is overloaded).

5) Stereo impression (stereo effect)- a sense of spatial distribution and sound resolution (in contrast to the spatial impression, which gives the illusion of an object moving deeper into the sound picture, this parameter characterizes a wide or narrow distribution of objects along the stereo panorama from left to right horizontally).

We localize the source of the sound. Bass width, sound volume, natural acoustic perspectives.

Together with the spatial impression, these two parameters make it possible to estimate

the acoustic atmosphere and the effect of the listener's feeling of presence in the room where the transmitted sound event takes place. An important aspect when considering this parameter is the assessment of the monocompatibility of the phonogram, that is, the presence of phase distortions during the translation of the phonogram in monophonic mode.

6)artistic quality(performance) is the total score of the artistic

quality of performance to which they are subject: art form, style, features

genre, interpretation, performing technique, intonation, articulation, etc.

Ideological and artistic assessment.

7) Sound reception technique. The overall rating of the technical sound quality.

The technical parameters for assessing sound quality are related to the characteristics of the sound transmission path used by the sound recording technology. The presence of interference, non-linear and amplitude-frequency distortions, detonations worsens the overall perception of the spatial sound picture, reduces the transparency of the sound, speech intelligibility, and distorts the timbre.

8) Instrumentation (arrangement). Overly rich or simply ill-conceived instrumentation can make a work inconvenient for sound recording, which can only be obtained in a multi-channel version or using careful acoustic and / or spectral separation of the performers. This parameter is closely related to transparency. .

Frequency range, frequency ratio.

9)Interference. This parameter evaluates the recording in terms of various interference heard during playback, namely:

Acoustic noise in and out of the studio;

Electromagnetic interference, background, amplifier noise, etc.;

Impulse interference (clicks, crackling, digital dropouts, etc.);

Strong non-linear distortion, audible detonation, clearly visible on

hearing of the installation site, etc.

10) Dynamic Range- this is a parameter of the intensity of the sound sensation within the limits depending on the technical conditions.

N. B. In general, in musical acoustics, the dynamic range called the distance on the loudness scale from the quietest to the loudest sound made by an instrument (or a group of instruments, or an orchestra, etc.).

In electroacoustics, the dynamic range - this is a technical framework, determined from below - by the threshold of the intrinsic noise of the sound transmission path, from above - by its overload capacity.

The concept of peak is closely related to the concept of dynamic range. - factor a. crest factor- this is difference between peak and RMS(English RMS - root mean square) signal values.

The most objective assessment of the quality of a sound recording can be obtained in a control room with appropriate acoustic processing, which maximally excludes the influence of room acoustics on the sound of a sound recording.

Listening should be carried out on high-end control units. The maximum listening volume in the control room should not exceed 90 dB.

All of these parameters are closely related and, changing one, it is impossible not to affect the other.

“Immediately after the rehearsal, I listened to the recording and I can say that the correspondence of the recording to the original is the highest. I consider this recording method and equipment to be very promising and promising.”

Yader Binyamini, conductor

Download ppt "A Unique Recording Technique"

Since the advent of sound recording, audio designers have sought to create equipment that would allow them to get as close as possible to real sound. However, despite significant progress in sound recording and sound reproducing technology, this has not been achieved in any way.

And how do you evaluate what sounds better and what is worse? Moreover, with the advent and development of digital audio, the applied objective evaluation methods based on measurements of distortion and frequency response do not give an accurate answer to this question. As you know, equipment even with ideal technical characteristics can sound far from ideal.

According to Justin Gordon Holt, acoustic engineer and founder of Stereophile magazine (USA), the founder of the subjective method of sound quality assessment, audio equipment is produced in order to listen to it, and not to measure its parameters. Traditional measurements of harmonic distortion, frequency response, or output power can reveal many things that a unit does poorly, but there is still no generally accepted procedure for comparing measurement results with how they affect sound quality. And it is clear that much of what people actually hear cannot be measured at all.

Therefore, we can say that the idea of ​​evaluating the sound of audio equipment arose in Wagner Audio Lab as a response to the demands of the times. After all, the problems with the sound quality of many devices produced today, both studio and household, are obvious.

First, it was proposed to use such sound quality assessment methods that allow you to look at the process of sound recording and sound reproduction as a whole - from a recording studio or a microphone in a concert hall to a home stereo system.

In the photo: Concert Hall of the Mariinsky Theatre. The theater was founded in 1783.

Secondly, the cooperation between Wagner Audio Lab with the Mariinsky Theater in St. Petersburg. In the process of sound recording of the Mariinsky Theater Orchestra, this technique developed step by step with the active participation of the artistic director-director of the theater, the famous conductor Valeria Gergieva. Moreover, it was first tested on a professional audio path to improve the sound recording of the Mariinsky Theater Orchestra.

So was born T&C sound quality evaluation method.

The main idea of ​​the technique is an expert comparison of the quality of sound reproduction of a musical work recorded through an acoustic system installed in the listening room with the sound of the orchestra in the theater hall during the recording. The time interval between these events should be as short as possible so that the quality of expert assessments is the most reliable. The comparison is carried out in the same building of the Mariinsky Theater in St. Petersburg according to several criteria, which are specially selected by the audition experts.

In the photo: During an audition in the Prokofiev Hall of the Mariinsky Theatre.

And the most important thing in this technique is which experts are involved in assessing the sound quality. When it comes to music, it is obvious that the best way to evaluate its sound is the one who creates it daily - the musician, and best of all - the conductor, who is the most objective in assessing the sound, so he hears the entire orchestra as a whole and monitors the coincidence with the voices of the singers .

It can be said that Wagner Audio Lab has a unique opportunity in this regard, since the maestro participates in assessing the veracity and purity of sound, its identity with real sound Valery Gergiev. We can proudly add that among the experts there are such well-known performers as Anna Netrebko, Yuri Bashmet, Igor Butman, Denis Matsuev.

In the photo: Sound recording of Denis Matsuev's concert at the Concert Hall of the Mariinsky Theatre.

Since January 2017, another high-class professional has joined our team - this is the Italian conductor Yader Binyami, a student of Riccardo Chaii, one of the best conductors in the world. The acquaintance took place at the Bolshoi Theater with the assistance of Anna Netrebko, who gave the highest recommendations to the work of the conductor. This collaboration is a very big contribution to the development and application of sound quality assessment techniques.

In the photo: Test listening to fragments of the opera after a rehearsal at the Bolshoi Theatre.

Of course, T&C methodology quite costly in various respects. But it gives the most accurate result you can rely on. The main thing is that the sound of the recording is not only flawless in terms of sound parameters, but that it also conveys the idea and emotions inherent in the musical work by the composer and performers. Thus, the equipment must be worthy of the level of the great masters.

We were confident that the T&C methodology could be used in other rooms and by other experts. But it is important that the quality of the acoustics of the hall and the audio system used are high, not to mention the level of expertise.

*J. Gordon Holt. Sounds Like? An Audio Glossary. Stereophile, Jul 29, 1993

^ 9.1. Relationship between instrumental measurements of technical parameters (objective tests) and acoustic auditions (subjective examinations).

Modern acoustic metrology, with all the perfection and versatility of modern measurement methods, still cannot give an absolutely accurate assessment of sound quality(KZ). This is due to the lack of a clear correlation between the objective parameters of the sound path (measured instrumentally, with the help of devices) and the subjective perception of the sound signal (auditory perception). For example, objective measurements of the main characteristics of CD players, even of a low price category (operating frequency range, non-linear distortion, signal-to-noise ratio, jigger ...) are very high in their parameters. And if we consider only them, then we can conclude that all models have exceptionally high sound quality and in this regard are practically indistinguishable from each other. However, in reality this is not the case, since the difference in the sound of such devices is very significant by ear. Therefore, the most reliable in this case is a subjective assessment. This does not mean that objective measurements should not be taken into account, but their results should only complement the results of listening. And the final assessment of the short circuit of the sound paths in the vast majority of cases remains after correctly conducted subjective tests.

Under correctness subjective expertise means:

• 
  properly selected listening room (without acoustic defects),
• 
  selection of the appropriate audio path,
• 
  selection of the most significant estimated parameters and their weighted scaling,
• 
  methodically substantiated selection of test phonograms,
• 
  selection of experts of the required qualification,
• 
  a sufficient number of expert auditions.

^ 9.2. Evaluation of speech sound quality

In accordance with Recommendation R.48 of the CCITT, the effective bandwidth of the audio path of the speech (tone) signal lies in the region of 300...3400 Hz. The main criterion that determines the quality of a speech signal is intelligibility, i.e. semantic clarity transmitted information to the listener. To conduct tests to assess legibility, tonal and articulation methods are used, in which expert assessments are made by trained listeners - experts. In this case, the fundamental the statistical validity of their results, those. the number of experts participating in the tests and the number of auditions should be sufficient.

Tone method is based on the ability of the human ear to quite accurately capture the minimum threshold of the volume level.

The speech signal is reproduced in the form of individual tonal strips. When listening, their level is reduced to the minimum perceived volume. The attenuation values ​​obtained with the help of lookup tables used in speech intelligibility calculations are recalculated and give the numerical value of this parameter.

To reproduce tone stripes, a tone generator and a special acoustic measuring tool are used - artificial mouth(Fig. 9.1.).

Structurally, this is a small loudspeaker in a combined box with a volume approximately equal to the volume

human mouth. The wall of the box, opposite from the loudspeaker, has a hole approximately equal in area to the speaking person's mouth. Since the frequency response and impulse responses of a loudspeaker have their own characteristics, the results of intelligibility assessment by the tonal method, even with a large number of listening sessions, are not highly reliable.

Rice. 9.1. Artificial mouth section
Articulation method involves the reproduction of syllables, words or phrases and their perception by ear by experts. Most often, syllables and non-existent words are used (for example, "shuts", "yt", "vus", "yang", etc.) so that the listeners do not guess what they heard.

In full form, the tables are given in GOST R 50840-95. For testing, pre-tabled syllables or words (usually 50 units each) recorded by professional speakers are played back through the audio path. Experts record what they hear in their protocols. Comparing what was heard with the source material, one can draw a conclusion about the quality of intelligibility.

Estimated characteristics for syllabic articulation tests are given in Table 9.1.

The articulation method was originally developed to assess the quality of speech intelligibility in the AM and MF FM radio transmission paths. The sounds of articulation table tests after passing through the radio path are listened to by experts located at a distance (for example, in different cities). The results of expert protocols are statistically processed. The reliability of such subjective tests is usually quite high. In addition, the articulation method is very convenient for periodic technological control of transmission quality. Changes in speech intelligibility and sound quality due to malfunctions in the radio transmitting equipment, when an additional device is introduced into the tested path, or when one of the path components is replaced, is detected very reliably.

But equally successful was the use of the articulation method in assessing the intelligibility of speech in auditoriums. One of the specific examples - conducted in 1999-2000. work on the study of the technical condition and adjustment of the existing sound amplification system of the auditorium of the State Academic Bolshoi Theater of Russia

Acoustic tests were carried out at three points in the stalls; at two points in the benoir; and at one point in the amphitheater. Working tuning measurements were carried out at the maximum permissible volume level in an empty hall (with a margin of 3 dB to the level of system self-excitation). Articulation syllabic tables were used for measurements. At each point, 6 independent measurements were made with the participation of 6 experts. The number of syllables in each dimension was 100.

Table 9.1. Evaluation characteristics for syllabic articulation tests

After fine-tuning the sound amplification system, the following results were obtained (Table 9.2):

Table 9.2. Measurement results.

It is equally convenient to use the articulation method to assess speech intelligibility in small halls without sound amplification. It uses a "live" announcer's voice. At least three different professional speakers are invited to read the articulation tables so that the specific coloring of their voices does not affect the measurement results.

With the advent of digital systems for recording and compressing speech, specific noises and distortions have been identified that degrade speech quality. To evaluate them, varieties of articulation tests have been introduced. To assess the visibility of the distortions introduced by the codec, the most accurate are paired comparisons of test phrases. The speech quality of the tested path is evaluated by comparison with the reference path, which is used as a standard telephone path (as per CCITT Recommendation R.48). The quality of speech is evaluated by the control phrases given in GOST R 50840-95. Each control phrase is transmitted twice:

• 
  once the signal is transmitted through the evaluated path,
• 
  the other - through the reference path.

Table 9.3. Correspondence between the speech quality of the vocal tract and the score for the paired comparison method

The order of alternation of paths is random. The control phrases uttered by the announcer can be recorded on a magnetic tape and then listened to by the experts in the recording. Accepted pauses between phrases 2...3 s, between pairs of phrases 4...5 s. However, the specificity of digital encoding is such that noticeable distortions are sometimes created. Although they do not interfere with its perception, they are easily distinguishable compared to the usual telephone path. Therefore, they provide 100% preference to the telephone path. Here the expert's assessment of "better" or "worse" is incorrect. Therefore, the quality of speech according to the method of paired comparisons is evaluated according to a 5-point system with an evaluation step of 0.1 points. As a benchmark for comparison, the quality of a standard analog audio path is taken when working from a telephone set with a dynamic microphone and at a nominal level. Its quality is estimated at 4 points. The correspondence between the speech quality of the vocal tract and the score for the paired comparison method is given in Table. 9.3.

The visibility of the distortions introduced by the codec is also evaluated by measurements phrasal intelligibility at an accelerated pace. Tables with the corresponding short phrases (three to four words) are given in GOST R 50840-95.) consist of short phrases of three to four words. The announcer reads one table of phrases at a normal pronunciation rate (one phrase in 2.4 s) and the second table at an accelerated pace (one phrase in 1.5 ... 1.6 s). The pause between phrases is at least 5...6 s. The expert first listens to the table read by the speaker at a normal pace, then the table read by the same speaker. A phrase is considered incorrectly accepted if at least one word is perceived incorrectly by the expert, omitted or added. Phrase intelligibility is determined by calculating the percentage of correctly received phrases for normal and accelerated pronunciation rates.

It is quite obvious that the requirements for high statistical reliability make the articulation method quite laborious. High demands are also placed on the qualifications of experts. However, all this pays off with the high reliability of the results obtained.
9.4. Audiometry

Audiometry is a special type of acoustic measurement designed to study the properties of human hearing, for example, to detect its deviations from normal. Audiometry is the main content of medical acoustics. Testing a person's hearing is especially important for people who are required to perceive sound, for example, speech, signals - drivers of vehicles, workers in manufacturing and construction enterprises, military personnel, etc. If necessary, audiometric tests are carried out on creative workers of sound and television broadcasting: sound engineers, sound engineers, sound engineers, music editors, actors, musicians, singers.

Considering the delicate nervous structure and slight mental vulnerability of representatives of creative professions, these tests should be carried out very delicately.

Audiometry as a method of medical research is used not only to diagnose hearing deviations from the norm, but also to identify some other human diseases.

In audiometry, subjective research methods are most often used, less often objective ones. Subjective methods are based on the test subject's oral responses to the measuring signals given by the operator, objective methods are based on the control of physiological reactions caused by sound stimuli with the help of special devices. At the same time, methods of the theory of conditioned reflexes developed by I.P. Pavlov and his students are widely used.

For subjective studies, special devices called audiometers are most often used. During research, the subject is placed in a chamber well isolated from external noise. Tone bursts of different frequencies, "white" noise strips, and speech signals are used as measuring signals.
^ 9.5. Audiometers with sinusoidal test signals

Audiometric methods are mainly used to detect and study deviations in hearing acuity from normal. Therefore, the measurement results are most often expressed as hearing loss compared to hearing acuity for normal average hearing. Devices for testing hearing acuity - audiometers - contain the following main parts:

• 
  audio frequency generator,
• 
  signal level control device,
• 
  a device for delivering an audio signal to the ear of the subject.

Hearing acuity is tested alternately for the right and left ear. The other ear is closed with a plug during the test, the sound pressure level is gradually reduced. The subject must answer the question: when he stops hearing the tone of a particular frequency. The graphs obtained as a result of the tests are not equal loudness curves for the threshold of hearing. They only reflect hearing loss compared to normal hearing acuity.

An example of such graphs is shown in Fig. 9.5. Graph 1 characterizes hearing acuity by bone conduction. This indicator indicates the normal sensitivity of hearing by bone conduction. It can be concluded that the main organ of the ear - the cochlea - is not damaged. Graph 2 is based on air conduction, i.e. involving the outer and middle ear. Hearing loss is approximately 50 dB over the entire range of audible sounds.

Rice. 9.5. . Graph 1 characterizes hearing acuity by bone conduction. Graph 2 is based on air conduction
Along with the above, there is another technique and, accordingly, there is a different type of audiometer. In it, instead of a continuous measuring signal, a group of sound pulses is generated. The number of pulses at the request of the operator can be changed. Most often, 4-5 pulses are set in a group. The operator fixes the number of pulses by flashing the LED or a lot of signal device. The subject does not see these flashes, but sometimes before the start of a group of impulses he is given a light signal.

This test procedure is believed to give more repeatable results. The subject, instead of answers of the form "I hear - I do not hear," fixes on the form the number of impulses heard.

This method of testing takes more time, but it allows you to test a group of subjects equipped with headphones at once. Everyone notes the result on their form.

In some modern models of audiometers, phonograms of sinusoidal voltages of different frequencies or narrow (half-octave or one-third octave) noise strips with a changing average frequency are used as a source of measuring signals. This improvement has simplified the procedures for testing hearing acuity, but some doctors and hygienists believe that the innovation hinders the ability to experiment on patients' hearing.
^ 9.6. Audiometer Calibration

Calibration of audiometers consists of two operations:

Establishing a correspondence between the numerical values ​​of the frequency scale and the true value of the frequency.

in establishing correspondence between the value of the scale of the level controller 0 background, the threshold of hearing at each measuring frequency.

The operation of correcting the frequency scale to the true frequency values ​​is relatively simple. The allowable difference between these two values ​​should not exceed + 2.5% of the true frequency value. A more complicated operation is the calibration (verification) of the instrument scale by sensitivity (by level).

For this operation, an auxiliary device is used - the so-called. "artificial ear", which is a small cylindrical acoustic chamber. Its volume is approximately equal to the volume of the cavity of the outer ear and auditory canal (~6 cm3), and the acoustic impedance is equal to the acoustic impedance of this volume of the ear. A microphone is inserted into the lower opening of the cylindrical chamber - a pressure receiver with a known sensitivity E = U / p, mV / Pa.

The voltage at the microphone clamps is measured with a voltmeter. A measuring earphone is attached to the upper opening of the camera.

Theoretically, the calibration process should be as follows. Knowing the sensitivity of the “artificial ear” microphone, we would set the voltage on the camera microphone to correspond to zero on the level scale at each frequency. This would correspond to zero on the level scale at these frequencies, i.e. would simulate the hearing threshold curve of the ear. But the measurement of such low voltages would be accompanied by a large error due to the influence of acoustic and electrical noise.

Therefore, the calibration is carried out at levels exceeding the threshold of hearing by 20-40 backgrounds, and then by recalculation, the values ​​\u200b\u200bof 0 background are set on the scale of the audiometer level control.

The procedure for calibrating an audiometer by level is complex and tedious, and takes a lot of time. However, without it, one cannot vouch for the accuracy of measuring hearing acuity. Of course, in the serial production of audiometers, the calibration operations are repeated to a limited extent, at 2–3 frequencies.
^ 9.7. Speech audiometers

The structure of speech audiometers is similar to the structure of audiometers, in which an audio frequency generator is used as the source of the measuring signal. The difference lies in the fact that the "live" speech of the speaker or the soundtrack of this speech is used as a signal source. The content of phonograms can be meaningful speech or a set of meaningless phrases.

The announcer reads the text, trying to maintain a constant volume level. To do this, he observes the readings of the level meter. Sometimes an automatic level control is introduced into the structure of a speech audiometer. It maintains the average signal level unchanged and thereby levels the speech volume level. Therefore, the measurement error is reduced.

A speech audiometer is necessary for the diagnosis of certain brain diseases, when a person hears speech sounds, but does not understand their meaning.

It has been established that for a person with a healthy psyche, the results of hearing tests on sinusoidal signals (tones) correlate well with the results of tests on speech signals. In some brain diseases or deviations from the normal psyche, this correlation is disturbed. This serves as the basis for a deeper examination of brain activity.

Speech audiometry procedures are sometimes used to evaluate the quality of communication and broadcast systems and devices. In this case, they are close to the articulation test procedures. Speech audiometry can detect hearing loss that is not possible with audiometers that generate pure tones. With the help of the first, direct hearing loss in speech is determined, which is very important for the patient.

A common drawback with all audiometric tests is that they are subjective and cannot be verified objectively. This goal is served by other methods developed on the basis of the methods of the school of I.P. Pavlov. This purpose is served by some physiological reactions to sound stimulation.
^ 9.8. Acoustic chambers for audiometry

The main requirement for acoustic chambers for audiometry is good protection from external acoustic noise. When performing audiometric tests with a loudspeaker, the noise level entering the chamber shall not exceed 20 dB above the hearing threshold. Such sound insulation can only be obtained with a box-in-box design of the chamber, i.e., with double massive barriers. It is especially difficult to provide good sound insulation at lower measurement frequencies. Due to the physical properties of barriers, sound insulation decreases as the frequency decreases. Recall that our own soundproofing:

where ω is the circular frequency, ρ is the barrier density, d is its thickness.

Fortunately, the sensitivity of hearing also decreases markedly with decreasing frequency. This facilitates the requirements for protection against penetrating noise.

Soundproofing requirements are reduced when using a headset for listening to test signals and using a plug on the other ear. The level of penetrating noise in this case can reach 40...45 dB, which is ensured by single barriers. The main source of penetrating noise is the door. A tight porch must be provided around the entire perimeter. The presence of ventilation holes reduces sound insulation to almost zero. Therefore, the volume of the chamber is chosen based on the presence of a volume of air sufficient for breathing. The volume of the chamber should be 20...25 m 3 . The camera should not have an extravagant interior. It should resemble a regular doctor's office. This is necessary to maintain a calm mental state of the subject. The chamber should be finished with effective sound-absorbing materials to avoid rare resonances with such a small volume, especially unpleasant at low frequencies. The reverberation time should be approximately 0.3 - 0.4 s.
^ 9.9. The phenomenon of juvenile deafness

Audiometry discovered and explained one new phenomenon in medicine.

Approximately 15 years ago, doctors began to notice a persistent hearing loss in a significant part of young people, in the age group from 14 to 20 years. V.A. Merzlovskaya (MIPT Student Polyclinic) and other doctors involved in the medical control of people entering higher educational institutions tried to discover the causes of this phenomenon and came to the conclusion that in 70 - 75% of cases it is due to the fact that young people abuse listening to music programs at high volume with the help of portable devices - players.

In the auditory canal of the ear, closed on one side by the ear telephone and on the other by the tympanic membrane, excessive sound pressure is created, causing great mechanical stress on the auditory nerves of the cochlea of ​​the inner ear.

If hearing is exposed to sounds with a sound pressure level of 100 dB (at the peaks of sound) for 2 hours, then by the end of this exposure, hearing sensitivity decreases by about 40 dB, and even 2 hours after the end of sound exposure, hearing sensitivity is not fully restored. .

Long-term listening to music programs at high levels leads to permanent hearing loss. Hearing sensitivity starts to drop off at about 800 Hz, and by 4 kHz this drop reaches about 40 dB.

Hygienists recommend limiting listening to music programs using players to two hours a day. However, manufacturers of such devices do not yet seek to include these recommendations in the instructions for using the players.
^ 9.10. Subjective assessments of the acoustic properties of a room

9.10.1. Disadvantages of objective methods

Many parameters have been invented and introduced into scientific use, designed to objectively reflect the acoustic properties of rooms:

• 
  reverberation time, its frequency course,
• 
  equivalent (effective) reverberation time,
• 
  acoustic attitude and others.

Therefore, along with objective ones, subjective indicators and methods for assessing the acoustic properties of rooms are widely used.

Strange as it may seem, the objective parameters and method of studying room acoustics do not give an unambiguous answer to the question: Is the room acoustically good or bad?

The numerical values ​​of the reverberation time, which are considered optimal, sometimes differ by 30 ... 40%, These differences can be explained by artistic inclinations, the habits of musicians and experts involved in optimizing the parameter, which is considered the main one.

Views regarding the frequency response of the reverberation time differ significantly. In American practice, it is considered useful to have a reverberation time rise at a frequency of 125 Hz ~ 40 ... 50% in relation to the reverberation time at a frequency of 500 Hz, and a slightly smaller rise (by 30 ... It is believed that these rises to some extent compensate for the decline in hearing sensitivity at the edge sections of the frequency range of audible sounds.

In Europe, the horizontal frequency response of the reverberation time is considered more acceptable. Only a slight rise at lower frequencies is considered acceptable. Some performers and conductors appreciate halls that even have rollbacks in the frequency response of the reverberation time at low and high frequencies.

The aesthetic assessment of the sound of music in two halls, in which the reverberation time is brought to the optimum by design and design measures, can differ significantly. It turns out that the time of arrival of the initial reflections to the listeners has a significant impact on the assessment of the hall. If the geometry of the hall is such that the delay times are close to the recommended ones, the music and speech sound good, despite the fact that the reverberation time is far from optimal.

An important role is played by the directions of arrival of waves reflected from obstacles. If most of the energy of the initial reflections comes to the listeners from the side of the stage or stage, the sound turns out to be "flat", spatial sensations are suppressed. Even worse, if the energy of the initial reflections comes from behind, when there are strong reflections from the back wall of the hall.

For an aesthetic assessment of the acoustic properties of the hall, it is more important to pay attention to obvious acoustic shortcomings: “mumbling” due to strongly pronounced hall resonances at low frequencies, sound focusing, the presence or absence of a “fluttering echo”, strong absorption of sound energy at medium and high frequencies by those sitting in the hall listeners (spectators). This shortcoming is called "the effect of listening (viewing) places". It is caused by the fact that the main part of the energy propagates parallel to the stalls plane, and not by reflections from the ceiling and walls of the room, and therefore is strongly absorbed by the listeners (spectators).

There are other acoustic disadvantages that are not taken into account by the accepted objective parameters.

For a long time it was believed that in order to obtain good acoustic properties, it is necessary to strive for high diffuseness, i.e. the most even distribution of sound energy throughout the volume of the room. The works of Thiele, Dreizen, Kacherovich, Furduev dispelled this misconception. You can read about this in the section "Measuring the degree of diffuseness of the sound field" chap. 2.

Consequently, this introduced parameter "Sound field diffuseness" does not give an unambiguous assessment of the acoustic properties of the premises. Other, subjective parameters are needed. They complement the objective parameters and characteristics.
^ 9.10.2. Concepts used in the subjective assessment of the acoustic properties of rooms

Subjective parameters are mostly of a qualitative, aesthetic nature. It should be emphasized that they are introduced specifically to assess the acoustic properties of the premises, and not to assess the quality of signals that have already passed through the broadcast channel or path. Experts making a subjective assessment of the acoustic properties of rooms must be specially trained for their tasks. It is important that they unambiguously understand the verbal definitions used in subjective examinations. The number of subjective concepts reaches many tens. Research is underway to reduce the number of expert assessments. The conditions for observation are unified and stipulated, musical and speech works are specially selected so that the results of examinations carried out in different halls can be compared.

To assess the duration of the echo, the following definitions are used: excessive, normal, underestimated. Sometimes a more subtle nuance of definitions is used.

The spatial impression is defined by the words: voluminous, airy, distributed in depth, "collected in a heap." Spatial sensations are enhanced with minimal coherence of the signals entering the right and left ears, with a significant proportion of the energy of the reflected waves.

The clarity of the sound is characterized by a good separability of the sound of orchestral and choral groups: the separability of the sound of individual instruments and the voices of singers. Sometimes they use the definition: detail. Paired concepts (antonyms) are widely used: legible - illegible, separately - together, in detail - blurry.

There is a balance of sonorities and timbre (tonal) balance. The balance of sonorities is understood as the proportionality of the sound of orchestral groups or groups of voices of the choir, the absence of excessive emphasis on the sonorities of individual instruments. Timbre (tonal) balance is characterized by the following associative concepts: neutral - colored, light - dull, loud - deaf, soft - hard, sharp - soft, full - liquid. As for shortcomings, they say: loudly, difficultly, sharply.

During articulation tests, attention is paid to clarity, clarity of perception of speech sounds and intonation: rich - poor, warm - cold, expressive - inexpressive, lively - dead, refined - rude, cheerful - sad. Sometimes they evaluate the loudness of the sound of the ensemble or soloists in the hall.

The loudness of the sound is determined not only by the intensity of the sounds extracted from the instruments, but also by the sound absorption of the hall, the intensity of the initial reflections, the uniformity or unevenness of the sound field in the listening positions. Loudness itself is not a quality parameter. But the distinguishability of useful sound in the presence of acoustic noise depends on it, and in this understanding, loudness characterizes the sound quality.
^ 9.10.3. Relationship between subjective and objective parameters

There are many works devoted to the subjective assessment of the acoustic quality of halls and its relationship with objective parameters. Of these, the studies carried out by groups of acousticians and musicians under the direction of Beranek, Kremer, Reichardt, and Schroeder deserve the most attention.

The result of the work of the Schroeder group was a technique that allows you to evaluate the acoustic qualities of a room according to two criteria:

• 
  clarity, defined as the ratio of the energy arriving at the receiving point during the initial period of resound (50 ms) to the total energy that arrived during the resound;
• 
  interural coherence, i.e. the degree of similarity of sounds coming to the right and left ears during the duration of the echo.

Kremer's group obtained different results. Expert musicians were offered 150 pairs of possible criteria. After critical discussion, the number of pairs was reduced to 19, and then to 4 criteria. However, it was not possible to find out with what weighting factors these criteria should be included in the overall quality assessment.

The largest amount of research on the subjective assessment of the acoustic quality of the halls was carried out by Beranek's group. Beranek ranked 47 most acoustically successful halls in the world in terms of quality. It was checked whether there is a correlation between the quality of the halls and 18 criteria taken into consideration in subjective evaluations.

When processing the results, Beranek came to the conclusion that the number of subjective criteria can be reduced to eight.

Reichardt (Technical University Dresden) suggested that of the 18 criteria proposed by Beranek, four main ones can be distinguished. At the same time, he proceeded from the fact that the criteria characterizing noticeable and removable shortcomings should be excluded from consideration. To assess the remaining four criteria, the corresponding objective criteria were found:

• 
  transparency of the sound corresponds to the criterion of clarity C
• 
  spatial impression - meets the spatial impression criterion R
• 
  timbre coloration of the sound - the frequency response of the reverberation time T(f)
• 
  loudness - sound energy density in the room ε = E/V, where V is the volume of the room, E is the energy contained in this volume.

1. What is the difference between objective and subjective sound quality ratings?

2. What is meant by sound quality measurement accuracy?

3. What are the differences between tonal and articulation methods for assessing intelligibility.

4. Name the main criterion that determines the sound quality of a speech signal.

5. How many points were chosen in the hall of the State Academic Bolshoi Theater of Russia when assessing speech intelligibility in 2000?

6. What is the scoring scale for assessing the quality of speech using the method of paired comparisons?

7. What is the duration of reading each tabular phrase at a normal and accelerated pace when assessing the phrasal intelligibility of speech?

8. Name the criterion for a phrase correctly accepted by an expert when assessing the phrasal intelligibility of speech.

9. What is the difference between direct expert auditions and comparative auditions?

10. What is the name of the system for selecting sound quality experts according to their competence, proposed by V.V. Furduev?

11. List the reasons that prevent obtaining reliable results of sound quality assessment in a subjective assessment.

12. Purpose of blocks in the structure of the audiometer.

13. What are the advantages and disadvantages of audiometers, in which the test signals are presented in the form of a ready-made phonogram?

14. What operations are performed when calibrating audiometers?

15. Why are speech audiometers used in addition to audiometers in which tone measuring signals are generated?

16. What are the technical requirements for acoustic chambers in which audiometry is performed?

17. Why are subjective assessments of the quality of room acoustics introduced in addition to objective ones?

18. What subjective concepts characterize the acoustic qualities of rooms?

19. What are the relationships between subjective and objective parameters?

20. What subjective parameters and why can be excluded from further consideration?

21. What is the essence of the study of the acoustic properties of rooms, conducted under the guidance of various acousticians?

Methods for expert evaluation of the sound quality of recordings
Boris Meerzon (magazine "Sound engineer": 1999: #8)

The number of recording studios, including those specializing in recording music of various forms and genres, is currently growing rapidly. The acquisition of the equipment necessary for equipping studios is no longer a problem. And therefore, a large number of people who have recently come into the profession are now engaged in sound recording, sometimes they have a poor understanding of the intricacies of sound engineering, ineptly using microphones and audio signal processing devices, which can especially affect the quality of the recording, and even not able to objectively evaluate the results of their work.
This, obviously, partly explains the variegation of the recordings in terms of their quality, which is not difficult to verify by listening at your leisure to the CDs of some of our, and sometimes foreign, firms that appear on sale. A very different and not always favorable impression on sophisticated listeners is also left by the sound quality of many musical television and radio programs.
In this regard, it is very important to ensure that sound engineers have a unified approach to assessing the sound quality of phonograms, use uniform criteria for its assessment and a single terminology understandable to all. This will certainly help to raise the level of sound engineering in general. Moreover, there is such experience.
As part of the work of the international radio and television organizations CCIR (now ITU) and OIRT (the latter has ceased to exist), recommendations have been developed for the subjective assessment of the quality of musical phonograms. This was done then, first of all, for the possibility of a successful international exchange of radio and television programs. But these recommendations can be fully applied even now, for internal practice, because they promote mutual understanding between sound engineers of different studios and help them, in the mutual exchange of recordings, to speak the same language. According to these recommendations, all broadcasters and recording studios should have special, permanent listening groups. They should be composed of qualified and specially trained experts from sound engineers, musicians, recording engineers, acousticians, technical supervisors and other professionals.
Experience has shown that recorders with good hearing and extensive experience, after several joint listening and discussions of the sound quality of recorded musical works, acquire the ability to evaluate the recordings so that their subjective opinions practically coincide. Thus, the average estimates of a group of trained experts (if several people take part in the audition) can be considered conditionally objective to a certain extent. Therefore, in educational institutions that train professional sound engineers, the curriculum includes, as a compulsory subject, "Analysis of sound recordings and their evaluation."
To facilitate the task set before the experts, a method of subjective assessment of sound quality has been developed, based on a strict specification of individual parameters that determine the overall quality of the soundtrack. Both technical and artistic indicators, considered in aggregate and interconnected with each other, are subject to evaluation. Listening should be carried out in a room that meets the established acoustic standards through standard loudspeaker installations. Phonograms are evaluated according to the following parameters:

1. spatial impression;
2. transparency;
3. musical balance;
4. timbre;
5. interference;
6. execution;
7. stereophonic.
In special cases, additional parameters are evaluated:
8. Arrangement (for dance and popular music);
9. Technique of sound pickup and recording.

The results of the examination are recorded in special protocols with columns corresponding to all the listed parameters.
(See table.)

Sound quality is assessed on a five-point scale:
5 - excellent;
4 - good;
3 - satisfactory;
2 - bad;
1 is completely unusable.
To evaluate the "interference" parameter, the scale takes the form:
5 - invisible;
4 - noticeable, but do not interfere;
3 - interfere a little;
2 - interfere;
1 - strongly interfere.

Let us consider in more detail each of the main parameters mentioned above, listing, for the convenience of assessment, their composite, particular parameters. Spatial impression - estimated by the expert's impression of the acoustic environment that existed during the recording. In particular, they judge the conformity of the size of the studio with the number of performers and the nature of the musical work, the time and nature of the reverberation, as well as the acoustic balance (the ratio of direct and reflected sounds). An important advantage of musical recordings is the feeling of a sound perspective in depth, i.e. creating the illusion of different distances from the listener to certain groups of orchestra instruments. Such a "diversity" of the sound picture to a certain extent recreates the volume of sound, which, as you know, is inevitably lost in electro-acoustic sound transmission, especially monophonic. However, when controlling the reverberation process during recording, creating various sound plans, the sound engineer must beware of the appearance of the so-called "multi-space". This lack of sound engineering is manifested in the fact that various instruments of the orchestra sound as if from different rooms, differing in their acoustic properties.
Multi-space, unless it is specifically provided for by the director's plans to create the necessary mise-en-scenes or special effects, is perceived in recordings of orchestral, choral or chamber music as a significant violation of the naturalness of sound. The reason for the multi-spatial sound can be the unfortunate placement of microphones in the studio (with a poly-microphone recording method), as well as, as mentioned earlier, the immoderate and inept use of artificial reverb.
Transparency is understood as the good distinguishability of the sound of individual instruments in the orchestra, the clarity of the musical texture, and the intelligibility of the text.
Transparency is directly dependent on the acoustic environment during recording, musical and acoustic balances, to a large extent, on the instrumentation of the piece being performed, and, of course, on the quality of the performance. Musical balance is determined by the ratio of the volume levels of various orchestral groups and individual instruments. This ratio mainly depends on the levels of direct sounds coming directly from the performers to the microphone.
Finding the optimal musical balance during recording is one of the main, and moreover, difficult, tasks of a sound engineer. When listening to an orchestra directly in the studio, musical balance may be perceived differently than when listening through the microphone path, even if the microphone is installed in the studio at the same point where the listener is. This is due to the different perception of sound during direct "binaural" listening in the studio and when listening through the loudspeaker in the control room. Normal musical balance can be achieved with proper microphone placement and mixing mode selection, and this is achieved the easier, the better the sound of the orchestra itself in the studio is balanced.
The timbre of the sound of musical instruments and voices should be transmitted naturally, without distortion. Such an assessment, of course, can only apply to the recording of traditional musical instruments, since. electronic music cannot fit into the framework of familiar sounds. With the help of electronic devices, the musician in this case can create new, synthetic timbres, which can only be assessed in this way: the timbre is pleasant or unpleasant, or at best, whether the timbre is similar to the timbre of one or another of the usual instruments.
But back to traditional music. The quality of the timbre transmission depends on the location of the performers and microphones in the studio, the nature of the studio acoustics, the frequency response of the sound transmission and sound recording channel, the nature and dose of the artificial reverberation signal. The timbre can be significantly distorted with increased nonlinear distortions in the path, detonation during recording, as well as with distortions of non-stationary processes that occur in the equipment, which determine the attacks of sounds, their attenuation and transitions from one sound to another. The sound engineer in the process of recording must learn to determine the cause of the distortion of timbres, and, having correctly diagnosed, take measures to eliminate them. According to the noise parameter, the recording is evaluated in terms of the visibility of extraneous sounds that interfere with the perception of music.
Interferences include:
- Noises penetrating into the studio as a result of imperfect soundproofing, as well as those created by the performers themselves (the rustle of turned musical pages, the clicks of wind instrument valves, the creak of furniture, parquet, or stands for the choir, the noise of the auditorium during recordings from open concerts, etc. ). Such acoustic noises when listening through the speaker are perceived more clearly and have a much greater disturbing effect than when listening directly in the hall. That is why it is so important to maintain complete silence in the studio when recording.
- Electrical pickup, hum, amplifier noise, tape noise in pauses, modulation noise, copy effect, quantization noise in digital recordings, etc.
- Impulse interference - electrical crackles, clicks from random instantaneous magnetization of the tape (for example, from magnetized scissors when editing analog phonograms), etc.
- Strong non-linear distortion, audible detonation, interference with the operation of automatic level controls (limiters, compressors), cod that occurs when the level is exceeded during digital recording.
The performance parameter is not technical, it determines the aesthetic properties of the phonogram. But the overall assessment of the recording depends on the quality of performance, and sometimes this parameter turns out to be decisive. Indeed, if a phonogram is impeccable from the point of view of recording, but contains unacceptable performance errors, then it should be recognized as unsuitable, despite other advantages.
The performance is evaluated both by the performer's general interpretation of this work, and by particular parameters: tempo, nuance, purity of intonation, clarity of articulation among singers and other indicators.
The stereophonic quality of the recording is evaluated according to the following particular parameters:
- clarity of localization of apparent sound sources (a sense of the distribution of directions on individual instruments of the orchestra);
- the width of the sound image;
- stereophonic balance between the sides, first of all, the clarity of the feeling of the middle of the stage, and in performances the smoothness of the movement of the performers on the stage (without jumps);
- the absence of a sound "hole" in the middle of the ensemble of performers.
In addition, the compatibility of stereo recording with mono playback should be determined - in terms of level, timbre, musical balance, transparency and spatial impression.
In some cases, in addition to the main assessments, it is necessary to evaluate the suitability of a given work for recording, for example, the arrangement of pop material. Indeed, an overly saturated, overloaded arrangement can sometimes make a work so inconvenient for recording that the most advanced technology and any sound engineering techniques will not help to achieve a satisfactory musical balance and good transparency in the recording.
The technique of sound recording and recording is also evaluated only in necessary cases. Here, attention is paid to the correct choice and use of microphones, level maintenance, subjective perception of loudness, the use of artificial reverberation, automatic dynamic range controls and other special effects, the quality of mixing and editing, and other aspects of the process of creating a phonogram that are not reflected in the previous paragraphs.
The overall assessment of the recording is given after the expert has made a complete analysis of the sound quality in all respects. Next, the final score of the phonogram is calculated as the average of the scores given by all the experts of the listening group. Many years of experience in the work of the listening group at the State House of Radio Broadcasting and Sound Recording showed that the most reliable examination becomes, if its results are entered into the protocols by each expert independently, without consultations during the listening with colleagues. Otherwise, the manifestation of taste and pressure of authorities is inevitable.