Scribd
Upload
204 views5 pages

Understanding Spectrograms in Speech

A spectrogram visually represents the spectrum of frequencies in a sound over time. It displays amplitude as color or intensity. Vowels have clearly defined formant bars that may merge for some vowels. Fricatives appear as chaotic noise across frequencies. Plosives are silent and the release looks like thin fricatives. Aspiration resembles voiceless vowels. Nasals and laterals resemble faint vowels. Spectrograms aid in identifying sounds.

Uploaded by

Marko
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
204 views5 pages

Understanding Spectrograms in Speech

A spectrogram visually represents the spectrum of frequencies in a sound over time. It displays amplitude as color or intensity. Vowels have clearly defined formant bars that may merge for some vowels. Fricatives appear as chaotic noise across frequencies. Plosives are silent and the release looks like thin fricatives. Aspiration resembles voiceless vowels. Nasals and laterals resemble faint vowels. Spectrograms aid in identifying sounds.

Uploaded by

Marko
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd

A spectrogram is a visual representation of the spectrum of frequencies in a sound or other signal

as they vary with time or some other variable. The spectrogram is a basic tool in audio spectral
analysis and other fields. It has been applied extensively in speech analysis. The instrument that
generates a spectrogram is called a spectrograph/ spectrometer. A common format is a graph with
two geometric dimensions: the horizontal axis represents time, the vertical axis is frequency; and
the third dimension indicates the amplitude of a particular frequency at a particular time, which is
represented by the intensity or colour of each point in the image.

Identifying sounds in spectrograms


Let's look at how various kinds of sounds appear on a spectrogram.

Vowels
Vowels usually have very clearly defined formant bars, as in the following:

In dipthongs, you can see the formants change frequency as the tongue body moves
through the mouth:

You can't always tell reliably which formant you're looking at -- F1, F2, F3, etc. -- unless
you already have a good idea of where to expect them. But the existence of formants is
usually obvious enough that you can at least be sure you're looking at a vowel.
(There are some especially common difficulties in identifying formants. In [], and
sometimes other back vowels, F1 and F2 are often so close together that they appear as a
single wide formant band. In [i], F2 and F3 also often appear merged together in a single
wide band.)

Fricatives
Fricatives are easy. The turbulent airstream of fricatives creates a chaotic mix of random
frequencies, each lasting for a very brief time. The result sounds much like static noise, and
on a spectrogram it looks like the kind of static noise you might see on a TV screen.
While each momentary burst of energy occurs at a random frequency, there are tendencies
in which frequencies the random bursts cluster around. [s] has a higher average frequency
than [] does; and both are higher than [f] or [].

Voiced fricatives show aspects of both regular vocal fold vibrations and a randomly
turbulent airstream.

[h]
[h] is really a voiceless version of the preceding or following vowel. On a spectrogram, it
looks a little like a cross between a fricative and a vowel. It will have a lot of random noise
that looks like static, but through the static you can usually see the faint bands of the
voiceless vowel's formants.

Plosives
The medial phase of a voiceless plosive is complete silence. On a spectrogram, this will
appear as a white blank.
The quiet vocal fold vibrations in a voiced plosive will sometimes appear as a faint band
along the bottom of the spectrogram at the frequency of f0. (But very often you won't see
anything there, either because the voicing got lost in the background noise or because the
recording or computer equipment cut off frequencies that low.)

To tell the difference between plosives, listeners rely on the release burst and on formant
transitions. On a spectrogram, the release burst looks like a very, very thin fricative. The
formant transitions (if you can see them) look like the formants have been distorted away
from the frequencies they have during most of the vowel.

Aspiration will look like a period of [h] between the blank gap and the vowel -- specifically,
a voiceless version of the following vowel. (Recall that the tongue body is in position for the
following vowel and that aspiration is just a delay in the onset of voicing.)
NB: Aspiration is not the same as the release burst. The period of aspiration (which only
some voiceless plosives have) is much longer than the very short release burst (which all
released plosives have).
The above spectrogram is of the English word attack [tk].

The periods of time labelled are:

A: the initial schwa


B: the medial phase of the [t] (silence)
C: the release burst of the [t]
D: the aspiration (delay of the onset of voicing for [])
E: the [] -- voicing has finally started. Right at the end of the vowel, you can see F2 and F3
start to approach one another in a formant transition pattern (often called the "velar
pinch") that usually marks the onset phase of a velar consonant.
F: the medial phase of the [k] (again, silence)
G: the release burst of the [k] (which I pronounced as released for the purposes of this
spectrogram)

Nasals and [l]


Nasals and [l] usually look like quite faint vowels, without a lot of amplitude in the higher
frequencies.
You can still see some things that look like formants. But the acoustic properties of tubes
with branches and side-chambers are much more complicated, with anti-formants as well
as formants, so the formant bands will appear in different positions and usually be fainter.
Which nasal or lateral it is usually isn't something you can figure out looking at just a
spectrogram.

Common questions

Powered by AI

A spectrogram distinguishes voiced from voiceless fricatives by capturing the dual characteristics of vibrations and turbulence. Voiced fricatives show aspects of regular vocal fold vibrations in addition to the chaotic, random frequencies due to a turbulent airstream. This combination results in visible periodic vocal striations or bands alongside noise patterns, differentiating them from purely voiceless fricatives, which only display randomness resembling static noise without such striations .

Identifying formants in back vowels like [ɑ] can be challenging because F1 and F2 often occur so close together that they appear as a single broad formant band on a spectrogram. This close positioning causes difficulty in distinguishing individual formants and can lead to misidentification unless there is pre-existing knowledge about the expected formant location. These challenges highlight the need for an analytical understanding of spectrogram interpretation aided by contextual linguistic knowledge .

Amplitude, reflecting the intensity of sound frequencies, plays a critical role in the analysis of speech sounds using spectrograms. Visually, amplitude is indicated by the brightness or color intensity of points on the spectrogram, with higher amplitude sounds appearing brighter or more colorful. This allows for the identification of sound energy levels over time, aiding in distinguishing between voiced consonants, vowels, and noise-like sounds such as fricatives based on their relative intensity .

The release burst is a critical feature in identifying plosives on a spectrogram. It appears as a very thin fricative and is crucial for differentiating plosives because it marks the precise moment when the built-up air pressure behind the closure is released. This feature helps listeners understand the identity of the plosive even if the voicing characteristics are masked by noise or equipment constraints . The presence of formant transitions, which look like the distortion of formants away from their stable frequencies, further assists in recognizing plosives when visible .

Vowels on a spectrogram exhibit clearly defined formant bars, which represent the resonant frequencies of the vocal tract. These formants are indicative of the specific vowel sound being produced, with diphthongs showing a change in formant frequencies as the tongue moves . In contrast, fricatives appear as a chaotic mix of random frequencies resembling static noise due to their turbulent airstream. High-frequency fricatives like [s] have a higher average frequency compared to [ʃ], [f], or [θ]. Thus, the presence of structured formants helps in identifying vowels, while random noise patterns assist in identifying fricatives.

Nasals and lateral approximants present identification challenges on a spectrogram because they display acoustic properties that differ markedly from vowels. While vowels exhibit strong, clearly defined formant bands, nasals and laterals appear as faint vowels with low amplitude at higher frequencies . This reduced amplitude is due to the complex resonant properties involving tubes with branches and side-chambers, resulting in both formants and anti-formants. These anti-formants create less pronounced formant bands, making it difficult to distinguish specific nasals or laterals solely from a spectrogram .

Formant transitions on a spectrogram are indicative of the onset or release phases of consonants such as [t] and [k]. At the end of a vowel, formant transitions exhibit changes wherein formant frequencies are distorted away from their steady-state positions as they move towards or away from the frequencies associated with the consonant articulation. For example, towards the end of the vowel [æ], the F2 and F3 formants move towards each other, indicating the onset of a velar consonant, a pattern often referred to as the "velar pinch" . This transition aids in identifying the place of articulation for the following or preceding consonant.

Aspiration and release bursts for voiceless plosives differ significantly in duration and representation on a spectrogram. The release burst is a very brief occurrence that resembles a thin fricative and indicates the release of occluded air pressure . Aspiration, however, lasts longer than the release burst and appears as noise resembling an [h], positioned between the silent phase and vowel onset on the spectrogram. It represents the delay in voicing onset, characterized by an absence of vocal fold vibration despite the vocal tract being in position for the next vowel .

The medial phase of a voiceless plosive is represented on a spectrogram as a region of complete silence, which appears as a white blank space. This phase signifies the moment of occlusion where no sound energy is released before the plosive's release burst occurs . Following this silent gap, the release burst and potential aspiration mark the plosive's continuation, helping to visually segment the sound on the spectrogram.

The distribution of static noise on a spectrogram differs between fricatives like [s] and [f] based on their average frequency characteristics. [s] exhibits a higher average frequency compared to [f], resulting in a more concentrated area of bright noise at higher frequencies on the spectrogram. In contrast, [f] displays noise more prominently at lower frequencies, with a less concentrated high-frequency band . This variation in noise distribution aids in distinguishing specific fricative sounds through their unique frequency signatures.

You might also like