Vol. 48 No. 3 Mar. 2007 1 Evaluation of Music-noise Assimilation Playback for Portable Audio Players Akifumi Inoue, Shohei Bise, Satoshi Ichimura and Yutaka Matsushita Though the population of portable audio listeners keeps growing, it still carries issues such as distractive background noise. This paper proposes a background noise reduction method using knowledge obtained by musical instrument performances. The method is based on our experience that listening to music feels comfortable when the rhythm accents are matched up at the correct timing. In our method, noise with a periodic rhythm takes the role of the rhythm section such as drums and basses. Then the tempo of the current track is converted at an unnoticeable amount so that the accents of the track can synchronize with that of the noise. We examined the reduction effects by carrying out five experiments, in which the traveling sound of a train was set as the periodical noise. Results of the experiments showed that the noise was faded out to a pleasant level without carrying out any spectrum noise reduction. We also showed that unreasonable changes in the tempo were unnecessary to keep an constant effectiveness. 1. 300 1) School of Computer Science, Tokyo University of Technology ALICE FORUM 20 1 1 1251
1252 Mar. 2007 2 3 4 5 6 7 2. 2.1 1 1 3 5 7 2.2 beatmania 3. 3.1 1 1 1 (a) 2) 1(b) 1(c) 3.2 1997 DJ 2001
Vol. 48 No. 3 1253 1 Fig. 1 Noise effects on melody line. Fig. 2 2 Overview of the proposed method. 2 3.1 BPM: Beat Per Minute BPM (1) (2) 2 3 Fig. 3 System flow. 2 2(a) BPM BPM 2(b) BPM 3.3 3 BPM BPM BPM BPM
1254 Mar. 2007 BPM 3.4 250 4,000 khz 0.2% TSM Time Scale Modification BPM 2 DJ Disk Jockey DJ 3) ±10% 4. Windows Visual C++ WAVE 44.1 khz 16 bit1 1 3 4 4 5 4 Fig. 4 HPF for train noise. 5 Fig. 5 The waveform of filtered noise. 1 Table 1 SOLA Parameters for SOLA. α 0.9 <α<1.1 S a 2,000 / α S s 2,000 k min 400 k max 400 N 2,800 BPM 65 4),5) BPM TSM SOLA Synchronous OverLap and Add 6) SOLA 7) OLA SOLA α S a S s k min k max N 8) 1 α 3.4 6 Screen-
Vol. 48 No. 3 1255 6 Fig. 6 Amplitudes of music and noise after the synchronization. blast Movie Studio BPM 5. 5.1 20 / / 5 7 150 Hz 2,000 4,000 Hz 200 Hz 2,000 Hz 11,000 Hz 20 0 0.015 7 Fig. 7 Frequency characteristics of noise and music-beats. 1 A BPM B 2 5 15 2 A B BPM 65 8 2 1 A B 15 10 7 5 0dB 5 9 3 3 BPM BPM 1 C 1 D 3 C D BPM C D C: BPM=65 D: BPM=130 10 1 4 BPM 120 BPM 60 2
1256 Mar. 2007 8 1 11 4 Fig. 8 Test 1 (availability). Fig. 11 Test 4 (back-beat). 9 2 Fig. 9 Test 2 (noise resistance). 12 5 Fig. 12 Test 5 (temporal characteristic). I 12 I 9) 10 3 BPM Fig. 10 Test 3 (The case where BPM difference between music and noise is larger). E F 3 3 E F E F BPM: 65 11 5 1 4 0 I 25 50 75 100 125 / 10) 5.2 13 1 A 1 BPM B 4 A: 1.60 B: 3.55 B 40 300
Vol. 48 No. 3 1257 13 1 Fig. 13 The result of test 1. 15 5 Fig. 15 The result of test 5. 14 2 Fig. 14 The result of test 2. 14 2 A B A: 10 db B: 7dB A 3 C 6 5 D 9 D C D BPM 4 E 7 8 F 5 3 ID3 15 5 I 88.75 70.0 5.3 11),12) 7
1258 Mar. 2007 MLD 13) 1 3 2.2 3.1 5 1 2 5.4 1 I I 4 I T 1/2 6. 6.1 6.2 BPM BPM BPM BPM 3 4 10% 12 5 21 22.4 10.2 5 50 1 1 3 4 1 10 12% 5
Vol. 48 No. 3 1259 40 6.3 6.4 3 1 BPM 1 BGM 7. 4) 5) 14) 1 1 15) 16 1) Schafer, R.M. (1996). 2) (2000). 3) 2006-GN-058 pp.133 138 (2006). 4)
1260 Mar. 2007 2003- MUS-53 pp.55 60 (2003). 5) 2004-MUS-56 pp.49 56 (2004). 6) Roucos, S. and Wilgus, A.M.: High Quality Time-Scame Modification for Speech, IEEE International Conference on ICASSP 85, Vol.10, pp.493 496 (1985). 7) Wong, J.W., Au, O.C. and Wong, P.H.: Fast Time Scale Modification using Envelope- Matching Technique (EM-TSM), IEEE International Symposium on ISCAS 98, Vol.5, pp.550 553 (1998). 8) C MAGAZINE, Vol.3, pp.13 41 (2002). 9) 1 (1990). 10) Massaro, D.: Preperceptual images, processing time and perceptual units in auditory perception, Psychological Review, Vol.79, No.2, pp.124 145 (1972). 11) pp.h 93 78, MA93 31 (1993). 12) 7 pp.447 448 (1995). 13) Moore, B. (1994). 14) Vol.J77- D-II, No.5, pp.901 911 (1994). 15) Vol.3, No.1, pp.108 148 (2004). 1999 2001 2005 ACM 2006 1966 1989 1994 1997 1999 2002 4 IT TEXT Web IT TEXT Web DICOMO 2003 2005 2006 ACM ( 18 6 21 ) ( 18 12 7 )
Vol. 48 No. 3 1261 1963 1968 1989 2002 3 2002 4 2003 4 2006 3 2006 4 MIS ITS LAN 201x 1993 1995 2000 2000 10 20 40 90 2000 10 2 2000 10 VR 2001 5 2002 3 IEEE ACM