How to determine the type of voice and what types exist? Psychology is easy

In our world, there is an increasing interest in human voice identification technologies. On the one hand, this is due to the implementation of high-performance systems that can read complex signals, such as voices. Methods for recognizing a person by voice have been around since a person learned to speak. We already know the pros and cons. The advantage of this method is ease of use. Today, more than 20 companies can provide their products with the implementation of this authentication method in information system. For example, Keyware Technologies sells products with a probabilistic error of 2-5%. this technology is perfectly implemented using the switched telephone network. The main problem associated with this approach is the accuracy of identification. The voice is created from a combination of behavioral and physiological factors. To date, voice identification has been implemented to control access to premises where an objective assessment of the degree of security is not critical. But for example, a person with a sore throat or a cold will have difficulty with such a system.

Voice recognition technology- perhaps the most practical solution for many applications. such systems analyze the characteristics after digitization, look at the pitch, tone and rhythm. The scheme for entering voice messages is shown in Fig.1. Despite the decrease in reliability in terms of recognition with the presence of noise, it is still beneficial economic solution, since sound cards and microphones have long been registered on the network.

Picture 1

There are many different microphones, but they all work on the same principle. sound wave hits the membrane, where the vibrations of the membrane are transmitted to the elastic element, which converts the vibrations into an electrical signal. The signal is amplified and fed to the input of the sound card. Sound card it is an analog-to-digital converter. The main parameters are the bit depth of the encoding and the sampling frequency. These parameters directly affect the quality of the recording, and as a result, the size of the recording itself. The voice identification system works according to the following scheme:

A user registration is created and a template calculation is done.
The time ranges of the speech stream are selected for analysis.
Primary signal processing is implemented.
The primary parameters are considered.
Created imprint- voice template.
The template is compared with other templates already in the database.

When registering, the user enters his identifier, for example, full name and says a key phrase several times. After the first processing, the fragments are compared and the similarities for the print are calculated.

When choosing sections of fragments, apply different ways. You can use the entire speech stream, excluding pauses. You can also choose fragments where the most powerful sounds, since there is the smallest probability of noise. You can also choose vowel sounds, as they can determine the nature of pronunciation, and so on.

In Fig.2. showing the probability of presence certain features personality voices in 18 phonemes.

Drawing - 2

In the process of the first signal processing, the spectral parameters of speech are analyzed. The basic procedure is narrow-band signal filtering and envelope recovery. When pronouncing the control phrase, the signal is brought to a single amplitude scale by the amplifier. Primary signal parameters have the following properties:

reflections of the speaker's personality
do not depend on noise
easily separated from the signal
be independent to the physical and emotional state of the announcer
little to imitate

Primary parameters can use frequency response, background, overtone spacing, formants, duration of individual sounds, and so on. During pronunciation, the pause between sounds can vary within 10 - 50%. To compensate for this instability, you can use the following methods:

Compression or stretching of individual sections.
Selecting the center of the sound area, then the measurements around the center do not play a strong role.

The disadvantage of such systems is also that the secret phrase is difficult to keep secret. Since when pronouncing a phrase, it can be recorded by different radio recording devices. The voice template takes about 2-5 KB, and the phrase lasts no longer than 2-3 seconds.

Hearing a voice stranger(for example, in a telephone conversation), we immediately imagine the parameters of his appearance. In particular, we associate low voices with more large size bodies, and tall ones with miniature ones. This conclusion was reached by a group of British scientists led by Dr. Yi Xu from University College London.

During the study, volunteers - both men and women - were asked to rate the attractiveness of voices whose timbres were changed. At the same time, the male group was allowed to listen to female voices of a high timbre, and the female group was given to listen to male voices. low tone. Representatives of the stronger sex called voices more attractive, characterized by high tonality, hoarseness and a wide range of sound vibrations: in their minds they were associated with small size body...

The weaker sex preferred voices of a low timbre with short sound ranges, associating them with impressive dimensions. But the ladies found even more attractive male voices hoarsely: apparently, it seemed to them that this softened the aggressiveness associated with large sizes.

According to experts, it's all about evolutionary instincts. Thus, earlier studies on birds and animals made it possible to establish that representatives of the fauna often receive information about the parameters and intentions of other creatures (say, the threat emanating from them) by the frequency, timbre and vibrations of the sounds they make. For example, a growl at a low frequency usually means that you are dealing with a large animal, a predator that is in a dominant position or preparing to attack, while clear sounds at a high frequency indicate small size and a subordinate position.

People may be different. So, we usually define a high sonorous voice as belonging to a young woman. And a low, deep-voiced masculine one is an aged man, or at least a respectable "gentleman." And it amazes us when an old retired woman turns out to be a sonorous voice (this happens too), and a narrow-shouldered squishy man turns out to be the owner of the bass.

Such situations are not uncommon in blind dating. Let's say you met someone on the Internet and agreed to call. The man called, you liked his voice and you decided to meet. It’s not the expected beautiful slender woman who comes to the meeting, but a fat aunt with a double chin, and the alleged handsome man turns out to be short and nondescript…

Photographs, even if you have seen them, actually give even less information about a person than voices, because, firstly, much is not visible in the pictures (for example, it is not always possible to determine how tall a person is), and secondly, people have a habit of posting their best images on the web, which can be almost untrue.

Here's a voice - that gives a more or less objective picture. At the very least, most people confidently rate other people's voices as "pleasant" or "unpleasant." At the same time, a too low voice in a woman is associated with "masculinity", that is, an excess of male hormones (sexual hoarseness is somewhat different!), And a too high and even more rattling voice in a man - on the contrary, with "femininity", infantilism, weakness and other qualities that by no means make the representatives of the stronger sex attractive in the eyes of women.

Moreover, even if everything is in order with appearance, the voice can repel us from this person. So it is the voice that can become a decisive factor in a relationship.

TOO LOW

Surprisingly, the other extreme is usually beneficial - people treat a person with a deep voice (especially a man) with great respect. That's what leaders say, those who feel power over others, knows his own worth and has great potential. Be careful and don't overdo it! When the voice is too heavy, it starts to sound artificial, pretentious.

TOO HIGH

Men and women who communicate in "raised tones" (high, sharp, whistling) are not taken seriously. This voice is associated with insecure, weak, indecisive, immature people. It is believed that its owners are at odds with their sexuality - they deny or even block it. When we become aware of our sensual side, we automatically lower the tone of our voice, and if this never happens, the person is not in harmony with his desires.

DESIGNATELY SEXY

When a person "turns on" an insinuating, sexy voice, do not hesitate - in front of you is a real manipulator. This seducer has a huge ego and believes that can easily use others to his own advantage. Those who easily fall for his/her sexy bait will one day be very disappointed when they hear him/her speaking in a completely normal voice.

DEEP

omen enthusiasts and rich. The owner of a velvety, voluminous voice (as if accompanied by a light echo) declares to others about his sensuality and, at the same time, inspires emotional security. It seems that the interlocutor is in complete control own life which is why we love listening to it so much. To develop this “depth” in yourself, train the “u” sound more often.

TOO SWEET

Another trick is a voice so touching, with a soft tone and gentle words, that a sensitive interlocutor can earn "diabetic shock" upon hearing it for the first time. Although such people are perceived as very kind, it will be very difficult to trust someone who always sounds so sweet.

SMOOTH, "WITHOUT LIFE"

It would seem that it is very convenient to take a neutral position in relation to the voice and not stand out too much. However, a flat, monotonous sound gives the speaker such dissonant characteristics as apathy, indifference and even depression.

VERY QUIET

"Dark Horses" Do you think that a modest quiet sound is a sign of shy and insecure people? However, often behind this sheepskin"(provided that we are not talking about health problems), if not a "wolf", then a petty tyrant, who, with the help of a voice trick, on the contrary, is hiding wants to get everyone's attention. It is important for them that people ask to repeat what was said, to give words more significance and use them as a game of power.

VERY LOUD

Those who regularly yell in conversation do so in order to attract to oneself increased attention and create excitement. Usually such people are associated with arrogance, perceived as socially awkward, boastful and envious.

AGGRESSIVE

People who speak "with impertinence", irritably, as if throwing words at an opponent - persons with strong character . These are zealous and aggressive rivals who rarely compromise. Even against the backdrop of the most harmless, everyday dialogue, they, quite unexpectedly, can develop a theater of military operations.

QUIVERING

Anyone who has this type of voice either experiences nervous tension, or something very upset. Such a person constantly worrying about everything, he is especially concerned about the question of how outsiders will perceive him / her. Furthermore, similar people often inspire fear, giving the impression of dangerous neurotics.

NASAL

nasal sound - the sign of "nannies". They touch, but do not inspire, they are rarely taken seriously, and their opinions are never heeded. smart and educated people with such a tone of voice, it can be extremely difficult to give the impression of a person who is really knowledgeable in any matter.

Establishing the belonging of a recorded voice and sounding speech to a specific person (speaker identification) is one of the varieties of phonoscopic examination, which, in turn, is a set of studies regarding voice and sounding speech. With the help of this type of research, you can determine whose voice sounds on the audio recording. In addition to the main parameter (sounding voice/voices), this type of study allows you to establish a variety of additional characteristics of the recording, namely: the possibility of applying copying, editing, erasing to the analyzed material.

Examinations to establish the belonging of a recorded voice and sounding speech to a specific person (speaker identification) can be divided into two groups: identification and diagnostic. The first includes issues related to the identification of a person, sound sources and means of video and sound recording. The second includes issues related to the establishment of personality traits, materials and means of sound recording. The active dissemination of expertise to establish the belonging of a recorded voice and sounding speech to a specific person is associated with the development of the technological sphere, which has led not only to the improvement of audio recording equipment, but also to the creation of new tools and programs for speech and voice analysis. These tools made it possible to make the expert process objective, to increase its effectiveness many times over, and to increase the complexity of the study.

The subject of a phonoscopic examination when establishing the belonging of a recorded voice and speech to a specific person is an audio recording on a magnetic or digital medium of one or more conversations received in the course of operational-search activities, actions carried out in the course of litigation. An examination of an audio recording is carried out, as a rule, to help the court, the investigation establish the authenticity and reliability of the phonogram. However, the study is also carried out without the consent of the court on the initiative of the person who provided the phonogram.

Phonoscopic examination to establish the belonging of a recorded voice and sounding speech to a specific person (speaker identification) provides a wide range of possibilities:

identification of the speaker on the audio recording by voice, oral speech;
establishing the verbatim content of the audio recording;
determining the conditions for creating an audio recording;
detection of traces of falsification of the audio recording;
finding out information about the persons whose speech is presented on the audio recording (gender, anatomical features, nationality, professional activity, social status, psycho-emotional state);
establishing the type of equipment on which the audio recording was made;
definition of the type of speech: spontaneous, memorized or read out.

Who conducts an examination to establish the belonging of a recorded voice and sounding speech to a specific person (identification of the speaker)?

Conducting an examination of an audio recording is one of the labor-intensive types of research that requires special knowledge. The complexity of the work of an expert in the course of an examination to establish the belonging of a recorded voice and sounding speech to a specific person (identification of the speaker) is that the specialist must understand many nuances, because a person’s speech is unique: on this distinguishing feature the principles of examination of audio recordings are constructed. The voice of a person is individual and has a different sound at different times of the day. Also, the sound of the voice depends on the physical condition of a person (fatigue, illness), psychological and acoustic conditions. However, it is impossible to intentionally fake a voice. Experts in the field of examination of audio recordings must have the widest range knowledge in the field of various sciences: physics, technology, psychology, medicine, linguistics, psychology.

As a rule, an examination to establish the ownership of a recorded voice and sounding speech to a specific person (speaker identification) is usually carried out by two experts: an engineer and a linguist. The first works with physical voice characteristics (pitch, frequency, sound duration), revealing technical specifications equipment and audio recordings. An expert linguist evaluates the linguistic, social and psychophysiological components. It should be noted that expert research is somewhat different from scientific research. strict sense words, and is not intended to demonstrate the level of knowledge of a specialist or new information about the object of study. The purpose of the examination of an audio recording is different: to apply knowledge to describe the subject of research, to establish the truth in accordance with the question posed.

What needs to be submitted for examination to establish the ownership of the recorded voice and the sounding speech to a specific person (identification of the speaker)?

For examination, it is necessary to provide the following comparative samples:

Samples of speech and voice obtained as a result of investigative actions (Article 202 of the Code of Criminal Procedure).
Samples of speech and voice obtained not in the course of investigative actions carried out in accordance with Articles 176-178, 182-185 of the Code of Criminal Procedure (search, inspection, seizure).
Samples of audio recordings and video phonograms from personal or home archives, audio and video recordings, Internet broadcasts of speeches, conversations, interviews can also be provided.
Speech and voice samples classified as conditionally free (audio recordings of investigative actions, namely: interrogations, confrontations).
Samples for research must meet the following requirements: comparability, representativeness and certainty of origin.

Also, in order to conduct an examination to establish the belonging of a recorded voice and sounding speech to a specific person (identification of the speaker), the sample of the audio recording must meet the following parameters:

The duration of the actual speech segment should be at least 3 minutes.
Frequency range: 100-5000 Hz.
Signal-to-noise ratio of at least 8 dB or more.
When providing a digital recording, the resolution must be at least 8 bits.

Is the audio recording evidence that can be presented in court?

AT recent times There is an opinion that the audio recording is not evidence and is not accepted by the investigative and judicial authorities. This is a delusion, since it is legally allowed to present an audio recording as evidence, as evidenced by articles 59 and 60 of the Civil Procedure Code, 67 and 68 of the Arbitration Procedure Code. These articles talk about the possibility of providing an audio recording as evidence, but only after confirming its relevance, admissibility and reliability. It is also necessary to indicate the conditions for the recording, by whom and when the phonogram was made. This requirement is described in Article 77 of the Code of Civil Procedure. An examination of the audio recording can give an opinion on the conformity of the audio recording with the specified parameters.

In the case of an open court session, the parties, as well as the citizens present, may, without the consent of the presiding judge, make an audio recording of the course of the court session.

In order to speed up the process of attaching an audio recording to evidence, it is necessary to submit, together with the audio recording itself, a verbatim transcript of the content of the conversation and an expert opinion confirming the authenticity of the information.

What questions does the expert answer in the course of the examination to establish whether the recorded voice and sounding speech belongs to a specific person (speaker identification)?

Does person X own the voice recorded on the presented audio recording?
Which parts of the audio recording belong to specific individuals?
Was the audio recording under investigation created on the presented equipment?
Was the audio recording under investigation created on one or more devices?
What object does the non-speech sounds recorded on the audio recording belong to?
Does the audio recording submitted for research contain a continuous recording, or was it made with stops?
Is the speech presented on the audio recording free, read or memorized?
What type of equipment was used to record audio?
What defects speech apparatus possesses the person whose voice is presented on the audio recording?
In what locality does the person live or lived, taking into account his speech characteristics?
How many faces appear in the audio recording submitted for research?
Has the audio recording submitted for research been edited? If so, what parts were installed?
What are the features of the environment in the place where the recording was made for the phonoscopic examination?
What personal characteristics does the person (people) whose voice is identified on the phonogram possess(s)?
One or different people belongs to the speech presented on the audio recording?
What is the sound environment (room, street, car interior) in which the audio recording was made?
Does the speech (fragments of speech) presented on the audio recording belong to the speech samples presented on the audio recording N?
Are there signs of intentional distortion of the author's characteristics in the speaker's speech?

An examination to establish whether the recorded voice and the sounding speech belongs to a specific person (speaker identification) requires not only special knowledge, but also serious technical equipment. Our Center has all the necessary equipment for the examination of audio recordings. Specialists in the field of examination of audio recordings not only in high level will conduct a study, but also present the customer's rights to various stages judicial trial.

Any person can order an examination to establish the belonging of a recorded voice and sounding speech to a specific person (speaker identification) and to a non-state company. The advantages of a non-state company in the field of phonoscopic examination are obvious: significant opportunities for the implementation, use of the latest technical means and research methods, the lack of departmental instructions make phonoscopic examination more effective. At the same time, it is important to remember that the conclusions of the state and non-state companies have the same value.

The list of suggested questions is not exhaustive. If other questions arise, it is advisable to seek advice from an expert before appointing an examination.

By the voice of a person, one can determine his character, mood, and even spiritual properties. The timbre of the voice, like the unique pattern of fingerprints, is strictly individual, and in the USA, England and Italy, voice recording is considered an indisputable legal document that cannot be forged.

Someone is genetically given a pleasant voice, while someone has to live with the "nasty". But there is always a way out - you can work on your voice. With skillful use, speech and voice become powerful psychological weapons that act on the subconscious of the interlocutor.

Psychologists have identified the main emotional reactions of a person to various variations in the pitch and timbre of the interlocutor's voice.

A high and sonorous voice is associated with youth, energy and, alas, with immaturity and inexperience. According to statistics, women and (especially) men with such a voice are much less likely to be appointed to solid posts. An interesting observation was made, the higher the voice, the lower the position.

Moreover, a piercing, high-pitched sound is read by the subconscious mind as an alarm signal. Thus, a long-sounding high voice causes an involuntary feeling of discomfort in the interlocutor and reduces the degree of confidence in the words. The owner of a very high voice should, more than anyone else, pay attention to the content and intonation of his speech.

The owners of a low voice are much more fortunate: it is associated with self-sufficiency, confidence, and intelligence. A person with such a voice is perceived by others as knowledgeable, and, therefore, more authoritative. The lower the voice of a man, the stronger and more reliable he looks in the eyes of a woman. By the way, statistics also do not deny the influence of voice on success with the opposite sex. You have probably noted the emergence of romantic thoughts while talking on the phone with by a stranger with a pleasing voice.

The low, velvety voice of a person we have never seen alone can evoke the strongest feeling of love. This is due to the subconscious reading of the interlocutor's genetic code. The fact is that a low voice is due to an increased content of male sex hormones in the blood, therefore, the owner of such a voice is more temperamental. That is why a woman with a low, chesty voice, rich in intonations, seems to men more sexual than the owner of a monotonous-sounding thin voice.

But in order to convince or seduce an interlocutor, a woman does not have to have a beautiful low voice by nature, it is enough to arm herself with her own voice (any that is), mind (enough worldly) and patience. First, work on your voice: the sound of the voice is affected not only by the hormonal component of the blood, but also by the structure vocal cords, psychological mood and breathing patterns.

Watch your posture - with good posture, the respiratory organs are located correctly, which allows it to be light and free. The sound should come out of the chest. To check the correctness of the exercise, put your hand on your chest, if it vibrates, you are on the right track.

Strengthens breathing and improves the sound of the voice by singing. Left alone in the apartment with yourself and household chores, try to sing something loudly enough. inflate air balloons- it is very strengthening respiratory system, advises "Goroscope.ru".

Diversify your speech with intonations, especially if you tend to present information “on one note”. Read aloud a small segment of the text several times (preferably a poem), filling it with new intonations with each new reading. Learn to clearly formulate the idea that you want to convey to the listener. This will save you from treacherous "mooing" in the pauses between semantic segments of speech. Stock up on arguments in favor of your opinion, place accents correctly. It is impossible to convince someone of something by quietly chewing phrases. Remember that information given only half a decibel louder is much more successful in persuading the listener and is better absorbed by him. Set the optimal “volume level” for your voice. Count out loud, from one to ten, gradually raising your voice, when you hear the sound that suits you, remember it, and then try to follow it.