| /* | |
| * Copyright (c) 2011 The WebRTC project authors. All Rights Reserved. | |
| * | |
| * Use of this source code is governed by a BSD-style license | |
| * that can be found in the LICENSE file in the root of the source | |
| * tree. An additional intellectual property rights grant can be found | |
| * in the file PATENTS. All contributing project authors may | |
| * be found in the AUTHORS file in the root of the source tree. | |
| */ | |
| /****************************************************************** | |
| iLBC Speech Coder ANSI-C Source Code | |
| WebRtcIlbcfix_GetLspPoly.c | |
| ******************************************************************/ | |
| #include "modules/audio_coding/codecs/ilbc/get_lsp_poly.h" | |
| #include "modules/audio_coding/codecs/ilbc/defines.h" | |
| /*----------------------------------------------------------------* | |
| * Construct the polynomials F1(z) and F2(z) from the LSP | |
| * (Computations are done in Q24) | |
| * | |
| * The expansion is performed using the following recursion: | |
| * | |
| * f[0] = 1; | |
| * tmp = -2.0 * lsp[0]; | |
| * f[1] = tmp; | |
| * for (i=2; i<=5; i++) { | |
| * b = -2.0 * lsp[2*i-2]; | |
| * f[i] = tmp*f[i-1] + 2.0*f[i-2]; | |
| * for (j=i; j>=2; j--) { | |
| * f[j] = f[j] + tmp*f[j-1] + f[j-2]; | |
| * } | |
| * f[i] = f[i] + tmp; | |
| * } | |
| *---------------------------------------------------------------*/ | |
| void WebRtcIlbcfix_GetLspPoly( | |
| int16_t *lsp, /* (i) LSP in Q15 */ | |
| int32_t *f) /* (o) polonymial in Q24 */ | |
| { | |
| int32_t tmpW32; | |
| int i, j; | |
| int16_t high, low; | |
| int16_t *lspPtr; | |
| int32_t *fPtr; | |
| lspPtr = lsp; | |
| fPtr = f; | |
| /* f[0] = 1.0 (Q24) */ | |
| (*fPtr) = (int32_t)16777216; | |
| fPtr++; | |
| (*fPtr) = WEBRTC_SPL_MUL((*lspPtr), -1024); | |
| fPtr++; | |
| lspPtr+=2; | |
| for(i=2; i<=5; i++) | |
| { | |
| (*fPtr) = fPtr[-2]; | |
| for(j=i; j>1; j--) | |
| { | |
| /* Compute f[j] = f[j] + tmp*f[j-1] + f[j-2]; */ | |
| high = (int16_t)(fPtr[-1] >> 16); | |
| low = (int16_t)((fPtr[-1] & 0xffff) >> 1); | |
| tmpW32 = 4 * high * *lspPtr + 4 * ((low * *lspPtr) >> 15); | |
| (*fPtr) += fPtr[-2]; | |
| (*fPtr) -= tmpW32; | |
| fPtr--; | |
| } | |
| *fPtr -= *lspPtr * (1 << 10); | |
| fPtr+=i; | |
| lspPtr+=2; | |
| } | |
| return; | |
| } |