Musical Sound Processing

1. Introduction
The utilization of visual and audio technology has increased dramatically along with the advent
of digital signal processing technology. The usage of recording and processing musical signals
became more adaptable and practical with the development of digital signal processing
technologies. The two main areas of musical signal processing are how to employ digital signal
processing technology to add delay and reverberation to the recorded music, and how to
ensure that the music recorded in soundproof recording studios can better replicate the music
beautifully performed in music hall. Another is that the recording director should be able to
balance process the recorded signal in the frequency domain while mixing the sound and edit
the music created by the solitary musician.
First, the sound from each individual instrument is recorded on a single track of a multitrack
tape recorder in an acoustically inert studio. The sound engineer will then alter the signals from
each track to add unique audio effects before combining them in a mix-down system to create
the stereo recording on a two-track tape recorder. The audio effects are produced artificially
utilizing a variety of signal processing circuits and devices, with more and more of these
operations being carried out using digital signal processing methods.

2. Time-Domain Operations
2.1 Single Echo Filter

Figure 1: Single echo filter: (a) filter structure, (b) typical impulse response, and (c) magnitude response
for R = 8 and a = 0.8
Echoes are simply generated by delay units. For example, the direct sound and a single echo
appearing R sampling periods later can be simply generated by the FIR filter of Figure 1(a),
which is characterized by the difference equation.

Equation 1 Equation 2

Figure 2: The network structure, unit impulse response and frequency response amplitude compose of
single echo filter.

2.2 Multiple Echo Filter

Figure 3: Multiple echo filter generating N-1 echoes: (a)filter structure (b) impulse response with a = 0.8
for N = 6 and R = 4.

To generate a fixed number of multiple echoes spaced R sampling periods apart with
exponentially decaying amplitudes, one can use an FIR filter with a transfer function of the
form.
 Equation 3

An IIR realization of this filter is sketched in Figure 3(a). The impulse response of a multiple echo
filter with a = 0:8 for N = 6 and R = 4 is shown in Figure 3(b). An infinite number of echoes
spaced R sampling periods apart with exponentially decaying amplitudes can be created by an
IIR filter with a transfer function of the form.

Figure 4: The network structure, unit impulse response and frequency response amplitude compose of
multiple echo filter.

2.3 Reverberation
In a concert hall or other enclosed area, the sound that reaches the listener is made up of direct
sound, early reflections, and reverberation. While the reverberation is made up of tightly
packed echoes, the early reflections are made up of several closely spaced echoes that are
essentially delayed and attenuated duplicates of the original sound. A listener will not perceive
the sound recorded in an inert studio as seeming "natural" in the same way as sound captured
in a closed area. However, by artificially producing the echoes and adding them to the original
Figure 5: IIR filter generating an infinite number of echoes: (a) filter structure, (b) impulse response with
a = 0:8 for R = 4, and (c) magnitude response with a = 0:8 for R = 7

signal, digital filtering can be used to transform the sound recorded in an inert studio into a
natural-sounding one.
The IIR comb filter of Figure 5(a) by itself does not provide natural-sounding reverberations for
two reasons. First, as can be seen from Figure 5(c), its magnitude response is not constant for
all frequencies, resulting in a “coloration” of many musical sounds that are often unpleasant for
listening purposes. Second, the output echo density, given by the number of echoes per second
generated by a unit impulse at the input, is much lower than that observed in a real room, thus
causing a “fluttering” of the composite sound. It has been observed that approximately 1000
echoes per second are necessary to create a reverberation that sounds free of flutter. To
develop a more realistic reverberation, a reverberator with an allpass structure, as indicated in
Figure 5(a), has been proposed. Its transfer function is given by.

Equation 4

The IIR comb filter of Figure 5(a) and the allpass reverberator of Figure 6(a) are basic
reverberator units that are suitably interconnected to develop a natural-sounding
reverberation.
 Figure 6: Allpass reverberator: (a) block diagram representation and (b) impulse response with a = 0:8
for R = 4

Figure 7: A proposed natural-sounding reverberator scheme.

Figure 7 shows one such interconnection composed of a parallel connection of four IIR echo
generators in cascade with two allpass reverberators. By choosing different values for the
delays in each section (obtained by adjusting Ri) and the multiplier constants a, it is possible to
arrive at a pleasant-sounding reverberation, duplicating that occurring in a specific closed
space, such as a concert hall.

2.4 Flanging
There are several special sound effects that are often used in the mix-down process. One such
effect is called flanging. Originally, it was created by feeding the same musical piece to two tape
recorders and then combining their delayed outputs while varying the difference t between
their delay times. One way of varying t is to slow down one of the tape recorders by placing the
operator’s thumb on the flange of the feed reel, which led to the name flanging. The FIR comb
filter of Figure 1(a) can be modified to create the flanging effect. In this case, the unit
generating the delay of R samples, or equivalently, a delay of RT seconds, where T is the
sampling period, is made a time-varying delay B(n), as indicated in Figure 8. The corresponding
input–output relation is then given by.
Equation 5

Figure 8: Generation of a flanging effect

Periodically varying the delay B(n) between 0 and R with a low frequency Wo such as

Equation 6
generates a flanging effect on the sound. It should be noted that, as the value of B(n) at an
instant n; in general, has a non-integer value, in an actual implementation, the output sample
value y(n) should be computed using some type of interpolation method such as that outlined.

2.5 Chorus Generator

 Figure 9: Generation of a chorus effect.

The chorus effect is achieved when several musicians are playing the same musical piece at the
same time but with small changes in the amplitudes and small timing differences between their
sounds. Such an effect can also be created synthetically by a chorus generator from the music
of a single musician. A simple modification of the digital filter of Figure 8 leads to a structure
that can be employed to simulate this sound effect. For example, the structure of Figure 9 can
effectively create a chorus of four musicians from the music of a single musician. To achieve this
effect, the delays B(n) are randomly varied with very slow variations. The phasing effect is
produced by processing the signal through a narrowband notch filter with variable notch
characteristics and adding a scaled portion of the notch filter output to the original signal, as
indicated in Figure 10. The phase of the signal at the notch filter output can dramatically alter
the phase of the combined signal, particularly around the notch frequency when it is varied
slowly. The tunable notch filter can be implemented using the technique described. The notch
filter in Figure 10 can be replaced with a cascade of tunable notch filters to provide an effect
like flanging. However, in flanging, the swept notch frequencies are always equally spaced,
whereas in phasing, the locations of the notch frequencies and their corresponding 3-dB
bandwidths are varied independently.

Figure 10: Generation of the phasing effect.

4. Time-Domain Operations
Almost all musical programs are produced basically in two stages:

• Sound from each individual instrument is recorded in an acoustically inert studio on a

single track of a multitrack tape recorder.

• Signals from each track are manipulated by the sound engineer by adding special audio
effects and then combined in a mix-down system to generate the final stereo recording

on a two-track tape recorder.

The special audio effects are either generated using time-domain operations or frequency domain
operations.

• Commonly used time-domain operations are echo generation, reverberation, flanging,

chorus generation, and phasing.

• Commonly used frequency-domain operations are filters and equalizers.

Frequency Frequency-Domain Operations Domain Operations

The frequency responses of individually recorded instruments or musical sounds of performers are often
modified by the sound engineer during the mix-down process.These effect are achieved by passing the
original signals through an equalizer.The equalizer provides “presence” by peaking the mid-frequency
components in the range 1.5 to 3 kHz and modifies the bass-treble relationships by providing “boost” or
“cut” to components outside this range.The equalizer is usually formed by cascading first-order and
second-order filters with adjustable frequency responses.

First-Order Digital Filters and Order Digital Filters and Equalizers Equalizers
Low-frequency Filters and Equalizers –
The transfer function of a first-order low frequency shelving filter for boost is given by

Where

First-Order Digital Filters and Order Digital Filters and Equalizers Equalizers
The tuning parameter is given by

where is the cutoff frequency and T is the sampling period.

The transfer function of a first-order low frequency shelving filter for cut is given by

Where

The tuning parameter is given by

where is the cutoff frequency and T is the sampling period.

The gain responses of the first-order low pass shelving filter are shown in the next slide for various
values of the tuning parameters.Gain responses of the low-frequency shelving filter for boost and cut
are shown below.

• Note: (1) The parameter K controls the amount of boost or cut at low frequencies.

• (2) The parameter controls the boost bandwidth, while the parameter controls the cut bandwidth.

A realization of the low-frequency shelving filter is shown below where for boost and
for cut

Gain responses of the high-frequency shelving filter for boost and cut are shown below
• Note: (1) The parameter K controls the amount of boost or cut at high frequencies.
• (2) The parameter controls the boost bandwidth, while the parameter controls the cut bandwidth.

Second-Order Digital Filters and Order Digital Filters and Equalizers

The transfer function of the second-order peak filter for boost is given by

where controls the center angular frequency where the band pass response peaks and

• Here, the parameter is related to the 3-dB bandwidth of the band pass response through

• Likewise, the transfer function of the second order peak filter for cut is given by Bw

In the previous expression, the center angular frequency , where the band stop response

dips, is related to the parameter β through and

cos( ) β = ωo
Here, the parameter is related to the 3-dB bandwidth of the band pass response

Through Bw

Note: The peak or the dip of the gain response occurs at the frequency which is independently controlled
by the parameter β
Note: The 3-dB bandwidth of the gain response is solely determined by the parameter
for boost or by the parameter for cut ωo, Bw, αB

The height of the peak of the magnitude response for boost is given by

The height of the dip of the magnitude response for cut is given by

Figure below show the gain responses of the second-order peak filter obtained by varying the parameter
ωo

Figure below show the gain responses of the second-order peak filter obtained by varying the
parameters K and Bw
Higher-Order Equalizers Order Equalizers
A graphic equalizer with tunable gain responses can be built using a cascade of first order and second-
order equalizers with external control of the maximum gain values of each section in the cascade.
Figure below shows the block diagram of a typical graphic equalizer:

Figure below shows the gain response of the equalizer shown in the previous slide for some

typical values of the parameter K (maximum gain values) of the individual sections

5. Advantages and Disadvantages of Musical Sound Processing

Advantages

 Information can be conveyed, displayed, or manipulated.

 Perfect reproducibility-identical performance from unit to unit.
 Guaranteed accuracy-accuracy is only determined by the number of bits used.
 Stored almost indefinitely without loss of information.

Disadvantages

 Speed and cost be expensive when large bandwidth signals.

 DSP designs can be consuming plus need the necessary resources (software etc).
  Finite word-length problems.

6. Conclusion
We compare the differences between FIR and IIR filters. IIR is comparable to a multi-peak filter, whereas
FIR is comparable to a comb filter. As a result, IIR can have more natural decay characteristics whereas
FIR might retain more of the original frequency. A compromise, indeed. Everything relies on the kind of
need we desire to fulfil. In instance 5, a reverberation paradigm that sounds natural is attempted to be
replicated. By combining IIR and All pass filters with the right decay ratios and delay units, we may
generate a reverberant atmosphere in any setting conceivable, such as a tiny room or a concert hall.