Hearing Flashcards
1
Q
what is frequency?
A
- Frequency in hearing refers to the number of cycles of a sound wave per second.
- It is measured in Hertz, where 1Hz refers to one wave cycle per second.
- It relates to the psychological experience of pitch but the two are not the same.
- Low frequencies generally sound low.
- High frequencies generally sound high.
2
Q
what are sound waves?
A
- Sound waves can be defined by:
- Frequency
- Amplitude
- Phase - These three physical properties can affect the subject experience of a sound wave.
3
Q
what is amplitude?
A
- This corresponds to the amount of change in pressure created by sound.
- It relates to the psychological experience of loudness but the two are not the same.
4
Q
what is phase?
A
- One complete cycle of a wave spans 360o, where:
- 0o and 180o are the resting levels
- 90o is max pressure
- 270o is min pressure
5
Q
what are two waves?
A
- Phase can be used to comparing two waves.
- We use the distance between their starting points.
6
Q
what are pure tones vs complex sounds?
A
- Pure tones a sine wave sound with a single frequency.
- Complex sounds are a large collection of simple sine waves added together. - These component simple sine waves could have different frequencies, amplitudes and phases.
- FOURIER ANALYSIS
7
Q
what is the fourier analysis?
A
- The complex wave presented can be broken down into sine wave components.
- You can present all waves in terms of time or frequency domain.
8
Q
what are harmonics?
A
- The pure sine-wave components of complex sounds are called harmonics.
- The lowest sinusoidal frequency is called the sound’s fundamental frequency or first harmonic.
- The other sinusoidal components are called the overtones.
- The harmonics are described in relation to the fundamental’s frequency f. So, the 3rd harmonic has a frequency that is 3x higher than the fundamental 3f.
9
Q
what is sounds like across time and frequency?
A
- Not all sounds are continuous like a sinusoidal wave, but even clicks and tone burst can be broken down into different frequencies.
10
Q
what is noise cancellation?
A
- When you add two waves that are 180o out of phase, you can cancel them out.
- Noise cancellation devices, measure the sounds from the environment and creates counter pressures to nullify the noise.
11
Q
what is pitch?
A
- Simple and complex waves can create the perceptual experience of pitch.
- This is the attribute of a sound that can be ordered on a scale from low to high.
- While pitch can be related to frequency, they are not the same.
12
Q
what is the difference between pitch and frequencies?
A
- The pitch of a complex tone can be determined by the fundamental frequency.
- But there are instances, where the pitch of a complex sound is determined by a missing fundamental.
- On the right, you have a complex wave with a 100Hz fundamental, i.e. 1cycle/10ms.
13
Q
what is the missing fundamental?
A
- Here are the harmonic components (200 Hz, 300 Hz, 400 Hz, and 500 Hz) of the complex wave shown earlier, but there is no 100Hz wave here.
- The 100Hz wave is referred to as the missing fundamental.
- This example demonstrates how the perceived pitch of a complex wave does not directly correspond to a single physical frequency present in the sound.
- Pitch therefore relates not only to the frequency components of complex sound but also to the relationship between them.
14
Q
what is loudness?
A
- Loudness like pitch is another perceptual experience, and it corresponds to how intense something sounds.
- 20 micropascals (20μPa) is considered the minimum sound pressure require for human hearing.
- But human ear is able to hear and distinguish sounds over an enormous range of amplitudes.
15
Q
what is sound pressure level?
A
- Because the range of what we can hear is so broad, scientists developed a Sound Pressure Level (SPL) scale that is logarithmic, and can more easily be used to compare sound-level differences.
16
Q
how do you calculate SPL?
A
- Amplitude is measured in dB, and this is calculated using:
- dB = 20 log (p1/p0), where p0 is the reference hearing threshold (20μPa) and p1 is another sound.
- If we consider a rustling of leaves that is 10x the pressure of the reference (p1= 20μPa10), then using the formula we can estimate its dB:
- dB=20 log(20μPa10/ 20μPa)
- dB=20 log(20μPa*10/ 20μPa)
- dB=20 log(10)=20 - We say: The rustling of the leaves is 20 dB.
17
Q
what is the minimum audible field?
A
- You can measure loudness sensitivity by presenting different frequencies at different pressure levels and recording the minimum pressure required for you to just hear each frequency. This is called the Minimum Audible Field (MAF).
- When measuring MAF, you are measuring the sensitivity of both ears.
18
Q
what is minimum audible pressure?
A
- You can also measure the Minimum Audible Pressure by placing a tiny microphone inside the ear canal and checking what pressure level is required for you to hear it.
- Here you are only measuring the sensitivity of one ear.
19
Q
what is the difference between MAF vs MAP?
A
- When you compare the MAF and MAP curves (aka audibility curves), you see an increased sensitivity (lower thresholds) for MAF because two ears are more sensitive than one.
- But there is an area where the two curves diverge.
- This is because when you are measuring MAP inside the ear canal by the eardrum, you do not benefit from the resonance created by the pinna and ear canal as when you measure MAF.
20
Q
how do we perceive loudness?
A
- The graph presents absolute sensitivity to different frequencies.
- The threshold curve at the bottom is the MAF curve. It represents the minimum sound level required for you to just hear different frequencies
- Above the threshold line are equal-loudness curves (ELC). They show the sound levels required for different frequencies to sound #dB much louder.
- For example, the 50 curve represents the sound pressure required to hear 50dB at different frequencies.
- Range of frequencies to which humans most sensitive (2–4 kHz) corresponds to the speech band.
21
Q
how do we localise sound?
A
- By comparing the loudness of the two ears, you can estimate the location of sound sources: sound energy arriving at the two ears from a single source will be more intense at the ear located nearest the source.
- This binaural disparity is called the Interaural Intensity Difference.
- This difference primarily arises from the fact that our head acts as a dense barrier where the transmission of sound is compromised.
22
Q
what is the interaural intensity difference?
A
- This cue is most effective for high frequency sounds as the effect of the head shadow is less effective for lower frequencies (bottom).
23
Q
what is the interaural time difference?
A
- You can also figure out the location of a sound from the time difference it reaches each ear. This is referred to as the Interaural Time Difference.
- If the sound is straight ahead, the air pressure will reach both ears at the same time.
- If the sound is located more to the left or right, the air pressure will reach the nearer ear first.
- This cue is most effective for low frequency sounds.
- This is because shifts in phases between the ears are more easily detected at lower frequencies.
- As the frequency of a sound increases, it becomes more difficult to determine which phase in the left ear matches the right.
24
Q
what is the minimum audible angle?
A
- This refers to the smallest difference of position of a sound sources which the listener can detect.
- MAA is smallest for sources straight ahead at around 1o for frequencies around 1kHz.
25
Q
what are phonemes?
A
- Phonemes: basic, abstract unit of phonological analysis and is often defined as the smallest phonetic unit that can distinguish the meaning of words.
- e.g. cat has three phonemes /k/ + /æ/ + /t/
- The English language uses about 42-47 phonemes.
- Two types: vowels and consonants
26
Q
what is speech perception?
A
- “The process of imposing a meaningful perceptual experience on an otherwiser meaningless speech input” (Massaro, 2001)
27
Q
how can you represent sound?
A
- You can present complex sounds with amplitude or frequency on the y-axis and time on the x-axis. The graph with frequency on the y-axis is referred to as a spectrogram.
28
Q
what is the traditional view of speech perception?
A
- Speech is a linear sequence of discrete and invariant acoustic patterns.
- “. . . there is so much evidence that speech is basically a sequence of discrete elements that it seems reasonable to limit consideration to mechanisms that break the stream of speech down into elements and identify each element as a member, or as probably a member, of one or another of a finite number of sets.” (Licklider, 1952)
- “The basic problem of interest to the linguist might be formulated as follows: What are the rules that would make it possible to go from the continuous acoustic signal that impinges on the ear to the symbolization of the utterance in terms of discrete units, e.g., phonemes or the letters of our alphabet? There can be no doubt that speech is a sequence of discrete entities, since in writing we perform the kind of symbolization just mentioned, while in reading aloud we execute the inverse of this operation; that is, we go from a discrete symbolization to a continuous acoustic signal.” (Halle, 1956)
- linear sequence: serial transmission of information.
- discrete: units are distinct from one another
- invariant: the symbols are always the same
29
Q
what is the problem with the traditional view?
A
- Phonemes are not produced in isolation (not discrete)
- Sounds blend into one another
- Example: linguistics [lɪŋ.gwɪs.tɪks]: /n/ –[ŋ]
- Acoustic information is “smeared” between segments
- Acoustic-phonetic invariancerefers to the concept that speech sounds (phonemes) should have consistent and identifiable acoustic properties, regardless of speaker variability or context.
- Speech sound’s acoustic patterns varies in a complex manner according to the preceding and following sounds.
30
Q
what is coarticulation?
A
- When we speak, we do not pronounce each phoneme separately, we start the next before we are done with the first.
- Segmentation problem: Where do words start and end???
31
Q
what is top-down processing in hearing?
A
- Our knowledge about grammar and words enable us to figure out when words begin and end.
- But when you are listening to a foreign language, this becomes more difficult.
32
Q
what are contextual effects?
A
- On the right, you have a simplified spectrogram of /di/ and /du/. Only the frequencies with the highest amplitudes are presented.
- The different clusters are called formants.
- Here the lower formants are the same but the upper formants are different.
- So although both utterances start with the same /d/ sound, the upper formant for rises for /di/ but falls for /du/.
- This shows that the d sound is different when followed by an “i” or a “u”. The context is therefore important.
- This is related to coarticulation.
33
Q
what is categorical perception?
A
- This is when your brain takes in continuous and variable stimuli and sorts them into distinct categories.
- Lieberman et al.(1957) presented subjects with 14 different sound stimuli (each shown in the spectrograms above).
- Although the patterns changed in a gradual manner, subjects perceived three distinctive phonemes
- Although the patterns changed in a gradual manner, subjects perceived three distinctive phonemes
34
Q
how do we use prior knowledge in hearing?
A
- Identification of sounds is helped by phonological knowledge. This is a top-down mechanism.
- When something is not clear, you can make a slip of the ear.
- For example you may hear “All the lonely Starbucks lovers” instead of “Got a long list of ex-lovers”.
35
Q
how do we use phonological knowledge?
A
- Identification of sounds is helped by phonological knowledge. This is another top-down mechanism.
- Listeners know what to expect in terms of phonotactics, there are rules on how sounds are combined. They are different for different languages.
- For example, in English “ng” can be at the end of a word, but it cannot be at the start of a word.
36
Q
how do we use knowledge about the lexicon?
A
- The words stored in our mental lexicon also help us to identify speech sounds.
37
Q
how do we use knowledge about context?
A
- We are more likely to hear certain words in some situations.
- Examples: */ɛd/
- Roses are */ɛd/.
- The baby is crying, has she ben */ɛd/?
- Ouch, I hit my */ɛd/.
38
Q
what is phonemic restoration?
A
- This is an auditory illusion where listeners perceive a phoneme in a word even though the signal is absent or masked by another sound.
- Can you hear the “s” in legislature?
- Warren (1970) removed the first “s” in “legislature” and inserted a cough of the same duration in its place. When the clips was played, participants claimed to hear the missing “s” clearly.
39
Q
what is phonemic restoration with context?
A
- Warren and Warren (1970) presented participants with four audio clips with a sound missing:
- It was found that the *eel was on the axle.
- It was found that the *eel was on the shoe.
- It was found that the *eel was on the orange.
- It was found that the *eel was on the table.
- Subjects reported to be unaware that a sound missing. Even when prompted to guess which sound was missing, they were unable to do so.
- They heard *eel as (1) wheel, (2) heel, (3) peel and (4) meal, depending on the linguistic context!
- It was found that the *eel was on the axle.
- It was found that the *eel was on the shoe.
- It was found that the *eel was on the orange.
- It was found that the *eel was on the table.
