KEMBAR78
Musical Sound Processing | PDF
Scribd
Upload
17 views7 pages

Musical Sound Processing

Uploaded by

71762103227
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
17 views7 pages

Musical Sound Processing

Uploaded by

71762103227
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 7
~ Web Resources cients (d)) match the power spectral the sequence of the FIR filter e Program 16. 5.m first the nately, applications, mation [Kum93]. Unfor true model order in many Iusieal programs are produced in basicaly ‘in Seotion 1.4.1 that eet Fecorded in an acoustically inert studio on each individual instrun nals from each track are manipulated by the von) ‘ ‘Then, the pe in a mix-down system to finally genes, are combi (Ear86]. The audio effects are artificial ip le78) ss \d they are increasingly being perto, devices, |, (Orf96]. Signal Procesing Tobe HOU IaMBe 9 — ju, he ly generated med using Web Resources of Digital Signal Processing ee Chapter 14: Applications a(n] vet ; ee 3 i") ost ; S % i * 1Sn Qn OE Reet ean (b) © | Figure 14.15: Single echo filter: (a) filter structure, (b) typical impulse response, and (c) magnitude response for i =8anda=08. | Some of the special audio effects that can be implemented digitally are reviewed in this section. 14.5.1 Time-Domain Operations Commonly used time-domain operations carried on musical sound signals are echo generation, reverberation, flanging, chorus generation, and phasing. In each of these operations, the basic building block is a delay. Single Echo Filter 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 14.15(a), which is characterized by the difference equation yin] = xin] +axtn—R], lal< 1, (44) or, equivalently, by the transfer function Hiz)=1+@: (14.48) In Eqs. (14.47) and (14.48), the delay parameter R denotes the time the sound wave takes to travel from the sound source to the listener after bouncing back from the reflecting wall, whereas the parameter with Ia < 1, represents the signal loss caused by propagation and reflection. ‘The impulse response of the single echo filter is sketched in Figure 14.15(b). The magnitude response of a single echo FIR filter for @= 0.8 and R = 8 is shown in Figure 14.15(c). The magnitude response exhibits R peaks and R dips in the range 0 < @< 27, with the peaks occurring at @= 2sk/R and the dips occurring at @= (2k + 1)Z/R, k=0, 1, .. R~ 1, Because of the comblike shape of the magni response, such a filter is also known as a comb filter. The maximum and minimum values of the magnitude response are given by 1 + a= 1.8 and 1 ~ @=02, respectively, Program 14.6 can be used to investigate the effect of a si i p. Figure 1.16. nee ech on the a ‘Reproduced with permission of Prof. Dale Callahan, University of Alabama, Biri ai 145 Musical Sound Processing ‘. 823 v{n) cel Cealiegs 0 @5 30 Time index 0 14.16: Multiple echo filter generating N ‘: meme ing N— 1 echoes: (a) filter structure and (b) impulse response with c= 0.8 Multiple Echo Filter ‘Togenerate a fixed number of multiple echoes : spaced R sampling period i B amplitudes, one can use an FIR filter with a transfer function Scene al ian le Hq) = ltaz* +07 +.-+a% (14.49) An IR realization of this filter is sketched in Figure 14.16(a). The impulse response of multiple echo fer with of = 0.8 for N = 6 and R = 4 is shown in Figure 14.16(0). ‘Aninfinite 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 H@ = Vaz ® 402s hore + 1 lemme) lal <1, (14.50) Figure 14.17(a) shows one possible realization of samples for R = 4 are indicated in Figure 14.17(b). sketched in Figure 14.17(c). The magnitude response with the peaks occurring at @ k/R and the dips occurring at @= (2k + IAPR, k= 0,1, oo R= Le The Sratimitm ahd roinirmum values of the magnitude response 2° given by Il ~ @) = 5 and 1(1+ @ 0.5556, respectively. : fundame y of the IIR multiple echo filte frequency fa ing musical instrument, such as the drum beat. For a specified Fy the delay fake aati 2 F/Fy sesulting in atime delay of RT= RIF, seconds [Or96]. ma He be ai A ewes Heck of multiple echos on the speech signal shown in 4.75 can be u the above IIR filter whose first 61 impulse response rhe magnitude response of this IR filter for R=7 is exhibits R peaks and R dips in the range 0< @< 27, hing the listener in a closed space, such as a concert hall, 4 1d reverberation. ‘The early reflections are reacl i on sound, early reflections, an wham, AL. University of Alabams Birmin; 1 of Eq, 14.50 is given by Fx=Fr/ web Rese cS Web Resources Program 14.8. + 14: Applications of Digital Signal Process, Chapte 824 atl eae osm € 1s a Normalized frequency () nite number of echoes: (a) filter structure, (b) impulse response wih Time index 9 (b) Figure 14.17: IIR filter generating an infi ‘@= 08 for R=4, and (c) magnitude response with a= 0.8 for R = 7. that are basically delayed and attenuated copies of the direct sound, whereas the reverberation is composed of densely packed echoes. The sound recorded in an inert studio is different from that recorded inside a closed space, and, as a result, the former does not sound “‘patural” to a listener, However, digital filtering can be employed to convert the sound recorded in an inert studio into a natural-sounding one by artificially creating the echoes and adding them to the original signal. The TR comb filter of Figure 14.17(a) by itself does not provide natural-sounding reverberations for two reasons [Sch62]. First, as can be seen from Figure 14.17(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 [Sch62]. To develop a more realistic reverberation, a reverberator with an allpass structure, as indicated in Figure 14.17(a), has been proposed (Sch62]° Its transfer function is given by composed of several closely spaced echoes ee a H() = lear lol< 4st) In the steady state, the spectral balance of the sound signal remains unchanged due to the unity magnitude response of the allpass reverberator. Program 14.87 can be used to investigate the effect of an allpass reverberator on the speech signal shown in Figure 1.16. The IIR comb filter of Figure 14,17(a) and the allpass reverberator of Figure 14.14(a) are basie reverberator units that are suitably interconnected to develop a natural-sounding reverberation. Figure 14.19 shows one such interconnection composed of a parallel connection of four IIR echo generators it! cascade with two allpass reverberators [Sch62], By choosing different values for the delays in each ©The structures shown here are the canonic single multiplier realizat 3 See also Section 8.6.1 ation ofa firstorder allpass transfer function (Mit#8h 7Reproduced with permission of Prof. Dale Callahan, University of Alabama, Birmi oa , musical Sound Processing 825 Pie Time index n pique 14.18: Allpass reverberator: (a) bl eS Hock diagram representation and (b) impulse re sponse with c= 0.8 for R=4. ra 1g reverberator scheme. Figure 14.19: A proposed natural-soundiny tants @, it is possible to arrive at a pleasant ific closed space, such as a concert hall, he multiplier cons hove natural-sounding reverberator on the Program 14 9. ‘tion (obtai ion (obtained by adjusting R,) and ‘occurring in a spec! und Prose reverberation, duplicating that |p, QB 14.9% can be used to vnvestigate the effect of te a Signal shown in Figure 1.16. Alabama, Birmingham, AL. | Tillie lae oh dengaeine , "Reproduced with ra

You might also like