Simple and complex tones, noise


We are used to calling a sound a tone if a definite pitch can be ascribed to it.For instance, we perceive sounds which are produced with a clarinet or a piano as a sequence of tones. A tone which is associated with a harmonic oscillation of the air pressure is called a simple or pure tone, and its pitch is closely related to the frequency of the oscillation. Several simple tones with different frequencies sounding simultaneously form what is called a complex tone.Virtually all tones produced by musical instruments are complex tones consisting of several or many simple tones the relative strengths and pitches of which determine their timbre. Many musical instruments as for instance string instruments or wood winds can produce steady tones associated with periodic pressure signals; the pitch of such a tone is determined by the reciprocal
of the period. According to Subsection 2.9.1 such a periodic signal can be split into a series of harmonic partials each of them corresponding to a pure tone (fundamental tone and overtones).
If, on the other hand, the spectrum of a sound signal is continuous, that is, if it does not contain discrete partials, the signal is aperiodic, and usually no clearly defined pitch can be attributed to it. This is a typical feature of noise. Examples of noise are bangs, or the rolling of thunder, the sounds produced by drums, by wind or other air flows, by flowing water in a river or in a water pipe. The latter sorts of noise are more or less stationary, and their aperiodic character is due to random processes involved in their generation. We can observe and analyse such sorts of noise but we cannot predict it in detail but only characterise them by average values as, for instance, their power spectrum. (The term noise has a two-fold meaning:
it denotes an aperiodic sound signal, or it means just any undesired sound. More will be said about the latter aspect of noise . Finally, periodic and aperiodic sound signals may occur in mixed form. The aperiodic component may be, for instance, the bowing noise of string instruments, or the air noise accompanied with the tone production in wind instruments, in particular, of flutes. Basically, this part of the sound signal is an unwanted side effect of tone production; nevertheless, it contributes often to the characteristics of the instrument’s sound. Figure 11.1 shows the time dependence of the sound pressure for a few sound signals. Speech consists of a succession of phonemes which often merge with each other. All vowels (/a/, /e/, /i/, /o/, /u/, etc.) are complex tones in our present terminology, and their spectra consist of discrete ‘spectral lines’. However, due to the speech intonation, their fundamental frequency and hence the frequency distance of overtones is continuously changing. Voiced consonants such as the voiced /s/ or /ch/ represent mixtures of both spectral types. Purely aperiodic phonemes are the unvoiced fricatives as well as the plosives (/p/, /k/ or /t/). According to what has been said earlier the unvoiced consonant /s/ could well be regarded as noise. The reader may have noted that it is difficult to arrive at a clear delimitation between tones and noise. This is even more so when it comes to distinguishing between desired and undesired sounds, that is, to define noise in the second meaning of the word. It is evident that here the subjective attitude of the listener plays an important role. Thus a motorbike freak will not assess the sound of his own vehicle as noise. On the other hand the most wonderful


EmoticonEmoticon