Structure of human ear Understanding the processes of

Structure of human ear Understanding the processes of human auditory system are key for posting requirements for architectural acoustics. This gives us knowledge about speech and communication mechanisms, hearing range and loudness perception. Human auditory system consists of three parts – outer ear, middle ear and inner ear; however this division is relative.

Structure of human ear Outer ear includes pinna and the ear canal, which ends with the eardrum. Its main purpose is to collect sound and direct it towards eardrum. Middle ear is a hollow space. There are situated the smallest three bones in human body – staples, incus and malleus. They are transmitting the vibrations of the ear drum to the inner ear, thus achieving impedance matching from the sparse air medium to the dense liquid media inside the cochlea.

Inner ear starts with the oval window at the beginning of the cochlea. The cochlea is the most important part of the ear – here is the frequency analysis and loudness sensation taking part. The organ of Corti lies also here to which the hair cells and neural ends are connected.

FREQUENCY ANALYSIS

Loudnes perception The amplitude of sound pressure at which neural sensation is starting is called hearing threshold. A normal hearing is considered 0 to 10 d. B for young healthy human. The amplitude of sound pressure at which a painful sensation is starting is called pain threshold and this is accepted to 140 d. B. However unpleasant sensation is starting to occur at 120 d. B. The difference between the weakest and the strongest audible sound is called dynamic range of auditory system and is said to be 120 d. B SPL or 1, 000 times.

Loudnes perception The amplitude of sound pressure at which neural sensation is starting is called hearing threshold. A normal hearing is considered 0 to 10 d. B for young healthy human. The amplitude of sound pressure at which a painful sensation is starting is called pain threshold and this is accepted to 140 d. B. However unpleasant sensation is starting to occur at 120 d. B. The difference between the weakest and the strongest audible sound is called dynamic range of auditory system and is said to be 120 d. B SPL or 1, 000 times.

Hearing threshold is frequency dependant. For 1000 Hz it is approximately 2. 10 -5 Ра; for 20 Hz and 20 k. Hz is about 3000 times greater. Increasing the sound pressure leads to the pain threshold, which is about 3. 106 times greater than the hearing threshold. The picture on the left is known as Fletcher and Munson graph. Threshold of pain is almost equal for all frequencies, including infrasound and ultrasound. That means that even though we cannot hear some sounds, high intensity of these frequencies can cause pain.

Stephens power law In 1860 Weber and Fechner conclude that the increase of a physical stimulus S with the smallest step ΔS leads to the increase in the subjective sensation for sound strength N with ΔN. N = C 1 • log(S) + C 2 ; C 1 and C 2 are constants Around 1940 American scientist Stephens discovered that the relationship between physical stimulus S and the subjective reaction N is better described by the following formula: log(N) = C 1 • log(S) + C 2 ; C 1 and C 2 are constants Or rewritten as ratio: e N = (S / S 0) , S 0 is the reference value of the physical quantity As can be seen this is a double logarithmic expression. Stehpens power law is valid for all human sensations ( for example sense of weight, salt, temperature etc. ). This expression gives the relation between the power of sensation, which is a subjective quantity, and power of stimulus, which is physical quantity. In the above formula е is a constant, which differs for every sensation.

Logarithmic levels. The decibel scale. In acoustics and radio electronics, the decibel scale is widely used due to its close relation to Steven’s power law, describing the relation between human hearing and sound levels. For energetic quantities, such as power, energy, sound intensity etc. , logarithmic level is determined according to: For sound pressure level, particle velocity, electric current and voltage, charge etc. , the logarithmic level is determined according to:

Loudness The term loudness is related to the subjective sensation of sound strength. The measurement unit for loudness is Son (or Sone). The unit “Son” has been introduced in order to differentiate between subjective sensation and physical quantity sound pressure level in decibels (d. BSPL). Note that the terms loudness and later introduced loudness level relate to different concepts. Loudness Unit: Son Loudness level Unit: Phon English Loudness level Deutsch Lautheit Lautstärkepegel Française Sonie Niveau de Sonie

Equal loudness contours; loudness level Equal loudness 40 phon Sounds which cause equal sensations have different intensity for different frequencies. The term loudness level represents the level of equal loudness sound for 1000 Hz. The measurement unit of loudness level is Phon.

Loudness curve The Stevens power low for sound sensation is written with the following formula: The first part of the curve is described by: Stephens power law for normal hearing Hearing threshold for normal hearing

Крива на гръмкостта Reference level: 1 Son = 40 Phon 2 10 d. B increase of the loudness level leads to doubling the loudness

Critical bands The hearing threshold of closely spaced in the frequency domain groups of pure tones with equal intensity, depends on the number of these tones if they are within the region of a critical band. The neural sensation for these frequencies is as if there is only one tone, in the center of this region, but with intensity of the sum of all tones. All those tones act on the same place of the Basilar membrane. Frequency band within which the property of ear so sum frequency components , is called critical band.

Pitch sensation Pure tones and periodical sounds with complex shape are perceived as musical sounds with different pitch. The higher the frequency of sound, the thinner the sound appears. The ear is very sensitive to small variations of the pure tones frequency. Pitch sensation is also logarithmic, as the loudness sensation.

Masking Hearing a sound signal in the presence of another sound is much more difficult than in silence. This effect is called masking effect. The masking effect is more prominent when the masker sound is closer in frequency to second sound. Low frequency sounds mask better than the high frequency.

Masking from pure tone A sound can mask another sound when they have close frequencies.

Masking effect 120 Masking spread 100 SPL in d. B 80 Hearing threshold Masked hearing threshold 60 40 20 0 1000 Frequency in Hz 10000 The masking effect depends on the level of masker and masked sounds.

Spatial hearing; binaural effect Human auditory system can determine the direction of incident sound by the use of two mechanisms: • Time difference ( bottom left ) • Level difference ( bottom right ) Due to the dimension of human head, both mechanisms work well for mid and high frequencies.

Speech production

Vocal chords

Speech signal generation

Speech signal generation Source Vocal chords spectrum

Speech signal generation Спектър на гласни струни Tube resonances

Speech signal generation Спектър на гласни струни Тръбни резонанси Emitted sound spectrum

Human voice spectrum Male Female These graphs show long time averaged spectrum of human voice. Female voice has higher pitch, but both are having frequency components down to 100 Hz.

Human voice spectrum Depending on the speech strength, the spectrum of human voice changes. The graph is adopted from ANSI-S 3. 5 – standard USA.

Envelope of speech signal sound time ”Fine structure” ”Envelope” time

THANKS FOR THE ATTENTION!