Knowledge (XXG)

Bit Rate Reduction

Source 📝

80:
of the block (see below) which controls the size of steps between the 16 possible values such that minute changes can be recorded if the 16 values are closer together but minute changes are lost if the 16 values are far apart, two bits indicating the
274: 85:(see below), and two bits of control information for the SPC700. The remaining eight bytes consist of 16 signed 4-bit nibbles which correspond to the 16 samples, packed in a 560:
The PlayStation APU and the Philips CD-i CDIC add another set of coefficients to the above and reorders them, for five unique of 8 filters total (these come from the
343: 310: 166: 539: 516: 493: 387: 363: 139: 119: 749: 675: 43: 557:. The denominators are powers of 2 to facilitate implementation with bit shifts as opposed to a more expensive hardware multiplier. 541:
bit downquantized version of the samples to the linear extrapolation from the last two samples (2nd order differential coding).
178: 549:
that is more resilient to errors in the encoded bitstream. Otherwise, any errors could propagate infinitely, as an
366: 561: 16:
This article is about the compression scheme used on the SPC-700. For other methods of bit-rate reduction, see
89:
manner. As 32 bytes of input become 9 bytes of output, the BRR algorithm yields a 3.56:1 compression ratio.
554: 744: 655: 465: 35: 518:
bit downquantized version of the samples to a lowered previous input (delta pack or differential coding).
69: 51: 545:
The coefficients of the above filters are specified as slightly less than 1 or 2 in order to realize a
76:. From most to least significant, the first byte of each block consists of four bits indicating the 687: 169: 315: 282: 550: 546: 58: 17: 144: 524: 501: 478: 372: 348: 345:
are the last-output and next-to-last-output PCM samples, respectively. The filter type
124: 104: 738: 47: 86: 55: 565: 39: 62: 73: 68:
BRR compresses each consecutive sequence of sixteen 16-bit
269:{\displaystyle s_{t}=2^{r}n+k_{1}s_{t-1}-k_{2}s_{t-2}} 527: 504: 481: 375: 351: 318: 285: 181: 147: 127: 107: 533: 510: 487: 381: 357: 337: 304: 268: 160: 133: 113: 654:These calculations are all done in signed 16.16 464:These calculations are all done in signed 16.16 8: 526: 503: 495:bit downquantized version of the samples. 480: 374: 350: 323: 317: 290: 284: 254: 244: 225: 215: 199: 186: 180: 152: 146: 126: 106: 46:, as well as the audio processors of the 570: 391: 667: 7: 141:should be decoded into a PCM sample 750:Super Nintendo Entertainment System 14: 168:using the following second-order 61:series. The method is a form of 475:Filter 0 linearly decodes the 42:sound coprocessor used in the 1: 553:of an ideal integrator is a 389:using the following table: 766: 72:samples into a block of 9 15: 369:prediction coefficients 34:, is a name given to an 521:Filters 2 and 3 add an 338:{\displaystyle s_{t-2}} 305:{\displaystyle s_{t-1}} 101:in a block with filter 93:Decompression algorithm 656:fixed-point arithmetic 535: 512: 489: 466:fixed-point arithmetic 383: 359: 339: 306: 270: 162: 135: 115: 688:SPC 700 Documentation 536: 513: 490: 384: 360: 340: 307: 271: 163: 161:{\displaystyle s_{t}} 136: 116: 731:US Patent 5,978,492 728:US Patent 5,519,166 725:US Patent 5,430,241 722:US Patent 5,303,374 719:US Patent 5,166,981 716:US Patent 5,128,963 713:US Patent 5,111,530 710:US Patent 5,086,475 707:US Patent 5,070,515 704:US Patent 5,041,830 701:US Patent 4,829,522 698:US Patent 4,797,902 695:US Patent 4,783,792 692:US Patent 4,685,115 525: 502: 479: 373: 349: 316: 283: 179: 145: 125: 105: 365:is translated into 38:method used on the 531: 508: 485: 379: 355: 335: 302: 266: 158: 131: 111: 24:Bit Rate Reduction 652: 651: 568:specifications): 534:{\displaystyle r} 511:{\displaystyle r} 498:Filter 1 adds an 488:{\displaystyle r} 462: 461: 382:{\displaystyle k} 358:{\displaystyle f} 170:linear prediction 134:{\displaystyle r} 114:{\displaystyle f} 36:audio compression 757: 680: 679: 672: 571: 551:impulse response 547:leaky integrator 540: 538: 537: 532: 517: 515: 514: 509: 494: 492: 491: 486: 392: 388: 386: 385: 380: 364: 362: 361: 356: 344: 342: 341: 336: 334: 333: 311: 309: 308: 303: 301: 300: 275: 273: 272: 267: 265: 264: 249: 248: 236: 235: 220: 219: 204: 203: 191: 190: 167: 165: 164: 159: 157: 156: 140: 138: 137: 132: 120: 118: 117: 112: 59:Macintosh Quadra 32:Bit Rate Reduced 18:data compression 765: 764: 760: 759: 758: 756: 755: 754: 735: 734: 684: 683: 674: 673: 669: 664: 593: 585: 523: 522: 500: 499: 477: 476: 414: 406: 371: 370: 347: 346: 319: 314: 313: 286: 281: 280: 250: 240: 221: 211: 195: 182: 177: 176: 148: 143: 142: 123: 122: 103: 102: 95: 21: 12: 11: 5: 763: 761: 753: 752: 747: 737: 736: 733: 732: 729: 726: 723: 720: 717: 714: 711: 708: 705: 702: 699: 696: 693: 690: 682: 681: 666: 665: 663: 660: 650: 649: 646: 643: 639: 638: 635: 632: 628: 627: 624: 621: 617: 616: 613: 610: 606: 605: 602: 599: 595: 594: 591: 586: 583: 578: 543: 542: 530: 519: 507: 496: 484: 460: 459: 456: 453: 449: 448: 445: 442: 438: 437: 434: 431: 427: 426: 423: 420: 416: 415: 412: 407: 404: 399: 378: 354: 332: 329: 326: 322: 299: 296: 293: 289: 277: 276: 263: 260: 257: 253: 247: 243: 239: 234: 231: 228: 224: 218: 214: 210: 207: 202: 198: 194: 189: 185: 155: 151: 130: 110: 94: 91: 30:, also called 13: 10: 9: 6: 4: 3: 2: 762: 751: 748: 746: 745:Digital audio 743: 742: 740: 730: 727: 724: 721: 718: 715: 712: 709: 706: 703: 700: 697: 694: 691: 689: 686: 685: 677: 671: 668: 661: 659: 657: 647: 644: 641: 640: 636: 633: 630: 629: 625: 622: 619: 618: 614: 611: 608: 607: 603: 600: 597: 596: 590: 587: 582: 579: 577: 573: 572: 569: 567: 563: 558: 556: 555:step function 552: 548: 528: 520: 505: 497: 482: 474: 473: 472: 471:Or in words: 469: 467: 457: 454: 451: 450: 446: 443: 440: 439: 435: 432: 429: 428: 424: 421: 418: 417: 411: 408: 403: 400: 398: 394: 393: 390: 376: 368: 352: 330: 327: 324: 320: 297: 294: 291: 287: 261: 258: 255: 251: 245: 241: 237: 232: 229: 226: 222: 216: 212: 208: 205: 200: 196: 192: 187: 183: 175: 174: 173: 171: 153: 149: 128: 108: 100: 92: 90: 88: 84: 79: 75: 71: 66: 64: 60: 57: 53: 49: 45: 41: 37: 33: 29: 25: 19: 670: 653: 588: 580: 575: 559: 544: 470: 463: 409: 401: 396: 278: 98: 96: 82: 77: 67: 48:Philips CD-i 31: 27: 23: 22: 566:Yellow Book 52:PlayStation 739:Categories 662:References 562:Green Book 172:equation: 121:and range 87:big-endian 54:, and the 328:− 295:− 259:− 238:− 230:− 97:A nibble 676:"68kMLA" 574:Filter 395:Filter 648:15/16 637:55/64 626:13/16 623:115/64 458:13/16 455:115/64 447:15/16 279:Here, 83:filter 50:, the 40:SPC700 645:61/32 634:49/32 612:15/16 444:61/32 433:15/16 78:range 74:bytes 63:ADPCM 56:Apple 26:, or 564:and 312:and 44:SNES 367:IIR 70:PCM 28:BRR 741:: 658:. 615:0 604:0 468:. 436:0 425:0 65:. 678:. 642:4 631:3 620:2 609:1 601:0 598:0 592:2 589:k 584:1 581:k 576:f 529:r 506:r 483:r 452:3 441:2 430:1 422:0 419:0 413:2 410:k 405:1 402:k 397:f 377:k 353:f 331:2 325:t 321:s 298:1 292:t 288:s 262:2 256:t 252:s 246:2 242:k 233:1 227:t 223:s 217:1 213:k 209:+ 206:n 201:r 197:2 193:= 188:t 184:s 154:t 150:s 129:r 109:f 99:n 20:.

Index

data compression
audio compression
SPC700
SNES
Philips CD-i
PlayStation
Apple
Macintosh Quadra
ADPCM
PCM
bytes
big-endian
linear prediction
IIR
fixed-point arithmetic
leaky integrator
impulse response
step function
Green Book
Yellow Book
fixed-point arithmetic
"68kMLA"
SPC 700 Documentation
Categories
Digital audio
Super Nintendo Entertainment System

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