Knowledge (XXG)

Sub-band coding

Source đź“ť

40: 1488: 1478: 105: 220:
for sound is a popular example of mu-law encoding. Using 8-bit mu-law encoding would cut the per-channel bitrate of CD audio down to about 350 kbit/s, half the standard rate. Because this simple method only minimally exploits masking effects, it produces results that are often audibly inferior
244:
First, a digital filter bank divides the input signal spectrum into some number (e.g., 32) of subbands. The psychoacoustic model looks at the energy in each of these subbands, as well as in the original signal, and computes masking thresholds using psychoacoustic information. Each of the subband
240:
The basic idea of SBC is to enable a data reduction by discarding information about frequencies which are masked. The result differs from the original signal, but if the discarded information is chosen carefully, the difference will not be noticeable, or more importantly, objectionable.
245:
samples is quantized and encoded so as to keep the quantization noise below the dynamically computed masking threshold. The final step is to format all these quantized samples into groups of data called frames, to facilitate eventual playback by a decoder.
237:. Human ears are normally sensitive to a wide range of frequencies, but when a sufficiently loud signal is present at one frequency, the ear will not hear weaker signals at nearby frequencies. We say that the louder signal masks the softer ones. 196:
may be thought of as a type of noise, because they are effectively the difference between the original source and its binary representation. With PCM, the audible effects of these errors can be mitigated with
248:
Decoding is much easier than encoding, since no psychoacoustic model is involved. The frames are unpacked, subband samples are decoded, and a frequency-time mapping reconstructs an output audio signal.
201:
and by using enough bits to ensure that the noise is low enough to be masked either by the signal itself or by other sources of noise. A high quality signal is possible, but at the cost of a high
276:) within a bit rate of 64 kbit/s. In the SB-ADPCM technique, the frequency band is split into two sub-bands (higher and lower) and the signals in each sub-band are encoded using ADPCM. 209:
for one channel of CD audio). In effect, many bits are wasted in encoding masked portions of the signal because PCM makes no assumptions about how the human ear hears.
69: 312: 980: 791: 680: 216:. Small signals are digitized with finer granularity than are large ones; the effect is to add noise that is proportional to the signal strength. Sun's 192:
The more bits used to represent each sample, the finer the granularity in the digital representation, and thus the smaller the quantization error. Such
494: 261:(MPEG), developed standards for coding of both audio and video. Subband coding resides at the heart of the popular MP3 format (more properly known as 1186: 1009: 803: 1191: 768: 167:
recordings, and so on. Digitization transforms continuous signals into discrete ones by sampling a signal's amplitude at uniform intervals and
921: 212:
Coding techniques reduce bitrate by exploiting known characteristics of the auditory system. A classic method is nonlinear PCM, such as the
1298: 1036: 975: 786: 736: 559: 404: 419: 305: 91: 136:, and encodes each one independently. This decomposition is often the first step in data compression for audio and video signals. 1411: 1421: 1259: 1110: 1029: 823: 1394: 1014: 808: 596: 527: 1156: 484: 1491: 285: 1481: 1384: 926: 298: 258: 52: 474: 469: 229:
The utility of SBC is perhaps best illustrated with a specific example. When used for audio compression, SBC exploits
1416: 62: 56: 48: 1527: 1522: 1343: 1181: 1161: 1105: 763: 554: 357: 1517: 1426: 1367: 1293: 1141: 731: 726: 581: 499: 424: 73: 1431: 1004: 798: 630: 399: 1372: 743: 586: 382: 372: 140: 997: 748: 532: 377: 262: 133: 1401: 156: 1085: 547: 509: 330: 177: 1316: 1166: 1151: 1120: 1115: 1024: 931: 833: 818: 601: 1389: 1359: 1338: 1244: 1176: 1070: 758: 574: 564: 459: 439: 434: 184: 164: 970: 213: 1333: 1321: 1303: 1171: 1055: 992: 838: 753: 709: 670: 352: 113: 1308: 1264: 1237: 1232: 1207: 1090: 1075: 985: 894: 889: 864: 718: 451: 429: 321: 230: 125: 1227: 1041: 965: 916: 884: 850: 409: 347: 234: 172: 1019: 813: 542: 537: 394: 367: 339: 217: 129: 27:
This article is about the signal coding technique. For the Bluetooth audio codec, see
1511: 1326: 1274: 941: 936: 911: 843: 464: 362: 1447: 414: 389: 1406: 1284: 1080: 956: 906: 28: 1463: 1254: 1249: 1136: 1095: 901: 17: 160: 290: 104: 272:
codec which uses sub-band adaptive differential pulse code modulation (SB-
1377: 1222: 879: 168: 1146: 620: 569: 202: 660: 206: 198: 1495: 1269: 1100: 693: 640: 273: 269: 175:. This process is fundamentally inexact, and involves two errors: 650: 504: 489: 479: 294: 625: 591: 144: 33: 257:
Beginning in the late 1980s, a standardization body, the
171:
to the nearest value representable with the available
155:
The simplest way to digitally encode audio signals is
1456: 1440: 1358: 1283: 1215: 1206: 1129: 1063: 1054: 955: 872: 863: 779: 717: 708: 610: 520: 450: 338: 329: 128:that breaks a signal into a number of different 108:Sub-band coding and decoding signal flow diagram 61:but its sources remain unclear because it lacks 139:SBC is the core technique used in many popular 306: 8: 1212: 1060: 869: 714: 335: 313: 299: 291: 92:Learn how and when to remove this message 103: 7: 25: 1487: 1486: 1477: 1476: 182:from sampling at intervals, and 38: 268:Sub-band coding is used in the 1: 259:Moving Picture Experts Group 1544: 1368:Compressed data structures 690:RLE + BWT + MTF + Huffman 358:Asymmetric numeral systems 221:compared to the original. 26: 1472: 727:Discrete cosine transform 657:LZ77 + Huffman + context 1432:Smallest grammar problem 286:Sub-Band Coding Tutorial 159:(PCM), which is used on 47:This article includes a 1373:Compressed suffix array 922:Nyquist–Shannon theorem 141:lossy audio compression 132:, typically by using a 76:more precise citations. 263:MPEG-1 Audio Layer III 151:Encoding audio signals 134:fast Fourier transform 109: 1402:Kolmogorov complexity 1270:Video characteristics 647:LZ77 + Huffman + ANS 157:pulse-code modulation 143:algorithms including 107: 1492:Compression software 1086:Compression artifact 1042:Psychoacoustic model 178:discretization error 1482:Compression formats 1121:Texture compression 1116:Standard test image 932:Silence compression 194:quantization errors 1390:Information theory 1245:Display resolution 1071:Chroma subsampling 460:Byte pair encoding 405:Shannon–Fano–Elias 185:quantization error 110: 49:list of references 1528:Signal processing 1523:Audio engineering 1505: 1504: 1354: 1353: 1304:Deblocking filter 1202: 1201: 1050: 1049: 859: 858: 704: 703: 124:) is any form of 114:signal processing 102: 101: 94: 16:(Redirected from 1535: 1518:Data compression 1490: 1489: 1480: 1479: 1309:Lapped transform 1213: 1091:Image resolution 1076:Coding tree unit 1061: 870: 715: 336: 322:Data compression 315: 308: 301: 292: 265:), for example. 231:auditory masking 225:Basic principles 205:(e.g., over 700 126:transform coding 97: 90: 86: 83: 77: 72:this article by 63:inline citations 42: 41: 34: 21: 1543: 1542: 1538: 1537: 1536: 1534: 1533: 1532: 1508: 1507: 1506: 1501: 1468: 1452: 1436: 1417:Rate–distortion 1350: 1279: 1198: 1125: 1046: 951: 947:Sub-band coding 855: 780:Predictive type 775: 700: 667:LZSS + Huffman 617:LZ77 + Huffman 606: 516: 452:Dictionary type 446: 348:Adaptive coding 325: 319: 282: 255: 235:auditory system 227: 214:μ-law algorithm 189:from rounding. 153: 130:frequency bands 118:sub-band coding 98: 87: 81: 78: 67: 53:related reading 43: 39: 32: 23: 22: 15: 12: 11: 5: 1541: 1539: 1531: 1530: 1525: 1520: 1510: 1509: 1503: 1502: 1500: 1499: 1484: 1473: 1470: 1469: 1467: 1466: 1460: 1458: 1454: 1453: 1451: 1450: 1444: 1442: 1438: 1437: 1435: 1434: 1429: 1424: 1419: 1414: 1409: 1404: 1399: 1398: 1397: 1387: 1382: 1381: 1380: 1375: 1364: 1362: 1356: 1355: 1352: 1351: 1349: 1348: 1347: 1346: 1341: 1331: 1330: 1329: 1324: 1319: 1311: 1306: 1301: 1296: 1290: 1288: 1281: 1280: 1278: 1277: 1272: 1267: 1262: 1257: 1252: 1247: 1242: 1241: 1240: 1235: 1230: 1219: 1217: 1210: 1204: 1203: 1200: 1199: 1197: 1196: 1195: 1194: 1189: 1184: 1179: 1169: 1164: 1159: 1154: 1149: 1144: 1139: 1133: 1131: 1127: 1126: 1124: 1123: 1118: 1113: 1108: 1103: 1098: 1093: 1088: 1083: 1078: 1073: 1067: 1065: 1058: 1052: 1051: 1048: 1047: 1045: 1044: 1039: 1034: 1033: 1032: 1027: 1022: 1017: 1012: 1002: 1001: 1000: 990: 989: 988: 983: 973: 968: 962: 960: 953: 952: 950: 949: 944: 939: 934: 929: 924: 919: 914: 909: 904: 899: 898: 897: 892: 887: 876: 874: 867: 861: 860: 857: 856: 854: 853: 851:Psychoacoustic 848: 847: 846: 841: 836: 828: 827: 826: 821: 816: 811: 806: 796: 795: 794: 783: 781: 777: 776: 774: 773: 772: 771: 766: 761: 751: 746: 741: 740: 739: 734: 723: 721: 719:Transform type 712: 706: 705: 702: 701: 699: 698: 697: 696: 688: 687: 686: 683: 675: 674: 673: 665: 664: 663: 655: 654: 653: 645: 644: 643: 635: 634: 633: 628: 623: 614: 612: 608: 607: 605: 604: 599: 594: 589: 584: 579: 578: 577: 572: 562: 557: 552: 551: 550: 540: 535: 530: 524: 522: 518: 517: 515: 514: 513: 512: 507: 502: 497: 492: 487: 482: 477: 472: 462: 456: 454: 448: 447: 445: 444: 443: 442: 437: 432: 427: 417: 412: 407: 402: 397: 392: 387: 386: 385: 380: 375: 365: 360: 355: 350: 344: 342: 333: 327: 326: 320: 318: 317: 310: 303: 295: 289: 288: 281: 280:External links 278: 254: 251: 226: 223: 218:Au file format 173:number of bits 152: 149: 100: 99: 57:external links 46: 44: 37: 24: 18:Subband coding 14: 13: 10: 9: 6: 4: 3: 2: 1540: 1529: 1526: 1524: 1521: 1519: 1516: 1515: 1513: 1497: 1493: 1485: 1483: 1475: 1474: 1471: 1465: 1462: 1461: 1459: 1455: 1449: 1446: 1445: 1443: 1439: 1433: 1430: 1428: 1425: 1423: 1420: 1418: 1415: 1413: 1410: 1408: 1405: 1403: 1400: 1396: 1393: 1392: 1391: 1388: 1386: 1383: 1379: 1376: 1374: 1371: 1370: 1369: 1366: 1365: 1363: 1361: 1357: 1345: 1342: 1340: 1337: 1336: 1335: 1332: 1328: 1325: 1323: 1320: 1318: 1315: 1314: 1312: 1310: 1307: 1305: 1302: 1300: 1297: 1295: 1292: 1291: 1289: 1286: 1282: 1276: 1275:Video quality 1273: 1271: 1268: 1266: 1263: 1261: 1258: 1256: 1253: 1251: 1248: 1246: 1243: 1239: 1236: 1234: 1231: 1229: 1226: 1225: 1224: 1221: 1220: 1218: 1214: 1211: 1209: 1205: 1193: 1190: 1188: 1185: 1183: 1180: 1178: 1175: 1174: 1173: 1170: 1168: 1165: 1163: 1160: 1158: 1155: 1153: 1150: 1148: 1145: 1143: 1140: 1138: 1135: 1134: 1132: 1128: 1122: 1119: 1117: 1114: 1112: 1109: 1107: 1104: 1102: 1099: 1097: 1094: 1092: 1089: 1087: 1084: 1082: 1079: 1077: 1074: 1072: 1069: 1068: 1066: 1062: 1059: 1057: 1053: 1043: 1040: 1038: 1035: 1031: 1028: 1026: 1023: 1021: 1018: 1016: 1013: 1011: 1008: 1007: 1006: 1003: 999: 996: 995: 994: 991: 987: 984: 982: 979: 978: 977: 974: 972: 969: 967: 964: 963: 961: 958: 954: 948: 945: 943: 942:Speech coding 940: 938: 937:Sound quality 935: 933: 930: 928: 925: 923: 920: 918: 915: 913: 912:Dynamic range 910: 908: 905: 903: 900: 896: 893: 891: 888: 886: 883: 882: 881: 878: 877: 875: 871: 868: 866: 862: 852: 849: 845: 842: 840: 837: 835: 832: 831: 829: 825: 822: 820: 817: 815: 812: 810: 807: 805: 802: 801: 800: 797: 793: 790: 789: 788: 785: 784: 782: 778: 770: 767: 765: 762: 760: 757: 756: 755: 752: 750: 747: 745: 742: 738: 735: 733: 730: 729: 728: 725: 724: 722: 720: 716: 713: 711: 707: 695: 692: 691: 689: 684: 682: 679: 678: 677:LZ77 + Range 676: 672: 669: 668: 666: 662: 659: 658: 656: 652: 649: 648: 646: 642: 639: 638: 636: 632: 629: 627: 624: 622: 619: 618: 616: 615: 613: 609: 603: 600: 598: 595: 593: 590: 588: 585: 583: 580: 576: 573: 571: 568: 567: 566: 563: 561: 558: 556: 553: 549: 546: 545: 544: 541: 539: 536: 534: 531: 529: 526: 525: 523: 519: 511: 508: 506: 503: 501: 498: 496: 493: 491: 488: 486: 483: 481: 478: 476: 473: 471: 468: 467: 466: 463: 461: 458: 457: 455: 453: 449: 441: 438: 436: 433: 431: 428: 426: 423: 422: 421: 418: 416: 413: 411: 408: 406: 403: 401: 398: 396: 393: 391: 388: 384: 381: 379: 376: 374: 371: 370: 369: 366: 364: 361: 359: 356: 354: 351: 349: 346: 345: 343: 341: 337: 334: 332: 328: 323: 316: 311: 309: 304: 302: 297: 296: 293: 287: 284: 283: 279: 277: 275: 271: 266: 264: 260: 252: 250: 246: 242: 238: 236: 232: 224: 222: 219: 215: 210: 208: 204: 200: 195: 190: 188: 186: 181: 179: 174: 170: 166: 162: 158: 150: 148: 146: 142: 137: 135: 131: 127: 123: 119: 115: 106: 96: 93: 85: 75: 71: 65: 64: 58: 54: 50: 45: 36: 35: 30: 19: 1448:Hutter Prize 1412:Quantization 1317:Compensation 1111:Quantization 946: 834:Compensation 400:Shannon–Fano 340:Entropy type 267: 256: 253:Applications 247: 243: 239: 228: 211: 193: 191: 183: 176: 154: 138: 121: 117: 111: 88: 82:October 2011 79: 68:Please help 60: 1407:Prefix code 1260:Frame types 1081:Color space 907:Convolution 637:LZ77 + ANS 548:Incremental 521:Other types 440:Levenshtein 74:introducing 29:SBC (codec) 1512:Categories 1464:Mark Adler 1422:Redundancy 1339:Daubechies 1322:Estimation 1255:Frame rate 1177:Daubechies 1137:Chain code 1096:Macroblock 902:Companding 839:Estimation 759:Daubechies 465:Lempel–Ziv 425:Exp-Golomb 353:Arithmetic 1441:Community 1265:Interlace 651:Zstandard 430:Fibonacci 420:Universal 378:Canonical 161:audio CDs 1427:Symmetry 1395:Timeline 1378:FM-index 1223:Bit rate 1216:Concepts 1064:Concepts 927:Sampling 880:Bit rate 873:Concepts 575:Sequitur 410:Tunstall 383:Modified 373:Adaptive 331:Lossless 169:rounding 1385:Entropy 1334:Wavelet 1313:Motion 1172:Wavelet 1152:Fractal 1147:Deflate 1130:Methods 917:Latency 830:Motion 754:Wavelet 671:LHA/LZH 621:Deflate 570:Re-Pair 565:Grammar 395:Shannon 368:Huffman 324:methods 233:in the 203:bitrate 70:improve 1496:codecs 1457:People 1360:Theory 1327:Vector 844:Vector 661:Brotli 611:Hybrid 510:Snappy 363:Golomb 207:kbit/s 199:dither 1287:parts 1285:Codec 1250:Frame 1208:Video 1192:SPIHT 1101:Pixel 1056:Image 1010:ACELP 981:ADPCM 971:μ-law 966:A-law 959:parts 957:Codec 865:Audio 804:ACELP 792:ADPCM 769:SPIHT 710:Lossy 694:bzip2 685:LZHAM 641:LZFSE 543:Delta 435:Gamma 415:Unary 390:Range 274:ADPCM 270:G.722 55:, or 1299:DPCM 1106:PSNR 1037:MDCT 1030:WLPC 1015:CELP 976:DPCM 824:WLPC 809:CELP 787:DPCM 737:MDCT 681:LZMA 582:LDCT 560:DPCM 505:LZWL 495:LZSS 490:LZRW 480:LZJB 1344:DWT 1294:DCT 1238:VBR 1233:CBR 1228:ABR 1187:EZW 1182:DWT 1167:RLE 1157:KLT 1142:DCT 1025:LSP 1020:LAR 1005:LPC 998:FFT 895:VBR 890:CBR 885:ABR 819:LSP 814:LAR 799:LPC 764:DWT 749:FFT 744:DST 732:DCT 631:LZS 626:LZX 602:RLE 597:PPM 592:PAQ 587:MTF 555:DMC 533:CTW 528:BWT 500:LZW 485:LZO 475:LZ4 470:842 165:DAT 145:MP3 122:SBC 112:In 1514:: 1162:LP 993:FT 986:DM 538:CM 163:, 147:. 116:, 59:, 51:, 1498:) 1494:( 314:e 307:t 300:v 187:, 180:, 120:( 95:) 89:( 84:) 80:( 66:. 31:. 20:)

Index

Subband coding
SBC (codec)
list of references
related reading
external links
inline citations
improve
introducing
Learn how and when to remove this message

signal processing
transform coding
frequency bands
fast Fourier transform
lossy audio compression
MP3
pulse-code modulation
audio CDs
DAT
rounding
number of bits
discretization error
quantization error
dither
bitrate
kbit/s
ÎĽ-law algorithm
Au file format
auditory masking
auditory system

Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.

↑