Structure of human ear

Download Report

Transcript Structure of human ear

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.
FREQUENCY ANALYSIS
The amplitude of sound pressure at which
neural sensation is starting is called
hearing threshold. A normal hearing is
considered 0 to 10 dB 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 dB. However unpleasant sensation
is starting to occur at 120dB.
The
difference between the weakest and the
strongest audible sound is called dynamic
range of auditory system and is said to be
120 dB SPL or 1,000,000 times.
The amplitude of sound pressure at
which neural sensation is starting is
called hearing threshold. A normal
hearing is considered 0 to 10 dB 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
dB. However unpleasant sensation is
starting to occur at 120dB. The
difference between the weakest and
the strongest audible sound is called
dynamic range of auditory system and
is said to be 120 dB SPL or 1,000,000
times.
Hearing threshold is frequency
dependant. For 1000 Hz it is
approximately 2.10-5 Ра; for 20
Hz and 20 kHz 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.
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 = C1 • log(S) + C2 ; C1 and C2 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) = C1 • log(S) + C2 ; C1 and C2 are constants
Or rewritten as ratio:
e
N = (S / S0) ,S0 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:
I
N  10  lg
I0
For sound pressure level, particle velocity, electric current and voltage,
charge etc., the logarithmic level is determined according to:
p
N  20  lg
p0
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 (dBSPL). Note that the terms loudness and
later introduced loudness level relate to different concepts.
Loudness
Unit:Son
Loudness level
Unit:Phon
English
Loudness
Loudness level
Deutsch
Lautheit
Lautstärkepegel
Française
Sonie
Niveau de Sonie
Equal loudness contours; loudness level
140
Sound pressure level, dB
120
100
80
Equal loudness
60
40 phon
40
20
0
10
20
50
100
200
500
1000
Frequency, Hz
2000
5000
20000
10000
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.
100
The Stevens power low for
sound sensation is written
with the following formula:
Loudness, son
10
1
The first part of the curve is
described by:
0.1
Stephens power
law for normal
hearing
0.01
-20
0
Hearing threshold
for normal
hearing
20
40
60
Loudness Level, phon
(= SPL in dB at 1 kHz)
80
100
100
Reference level:
1 Son = 40 Phon
Loudness, son
10
2
1
10 dB increase of
the loudness level
leads to doubling
the loudness
0.1
0.01
-20
0
20
40
60
Loudness Level, phon
(= SPL in dB at 1 kHz)
80
100
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.
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.
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 dB
80
Hearing
threshold
Masked
hearing
threshold
60
40
20
0
100
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-S3.5 – standard USA.
Envelope of speech signal
sound
time

time
”Fine structure”
”Envelope”
time