| 1 | /* |
| 2 | * (I)RDFT transforms |
| 3 | * Copyright (c) 2009 Alex Converse <alex dot converse at gmail dot com> |
| 4 | * |
| 5 | * This file is part of FFmpeg. |
| 6 | * |
| 7 | * FFmpeg is free software; you can redistribute it and/or |
| 8 | * modify it under the terms of the GNU Lesser General Public |
| 9 | * License as published by the Free Software Foundation; either |
| 10 | * version 2.1 of the License, or (at your option) any later version. |
| 11 | * |
| 12 | * FFmpeg is distributed in the hope that it will be useful, |
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 15 | * Lesser General Public License for more details. |
| 16 | * |
| 17 | * You should have received a copy of the GNU Lesser General Public |
| 18 | * License along with FFmpeg; if not, write to the Free Software |
| 19 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 20 | */ |
| 21 | #include <stdlib.h> |
| 22 | #include <math.h> |
| 23 | #include "libavutil/mathematics.h" |
| 24 | #include "rdft.h" |
| 25 | |
| 26 | /** |
| 27 | * @file |
| 28 | * (Inverse) Real Discrete Fourier Transforms. |
| 29 | */ |
| 30 | |
| 31 | /* sin(2*pi*x/n) for 0<=x<n/4, followed by n/2<=x<3n/4 */ |
| 32 | #if !CONFIG_HARDCODED_TABLES |
| 33 | SINTABLE(16); |
| 34 | SINTABLE(32); |
| 35 | SINTABLE(64); |
| 36 | SINTABLE(128); |
| 37 | SINTABLE(256); |
| 38 | SINTABLE(512); |
| 39 | SINTABLE(1024); |
| 40 | SINTABLE(2048); |
| 41 | SINTABLE(4096); |
| 42 | SINTABLE(8192); |
| 43 | SINTABLE(16384); |
| 44 | SINTABLE(32768); |
| 45 | SINTABLE(65536); |
| 46 | #endif |
| 47 | static SINTABLE_CONST FFTSample * const ff_sin_tabs[] = { |
| 48 | NULL, NULL, NULL, NULL, |
| 49 | ff_sin_16, ff_sin_32, ff_sin_64, ff_sin_128, ff_sin_256, ff_sin_512, ff_sin_1024, |
| 50 | ff_sin_2048, ff_sin_4096, ff_sin_8192, ff_sin_16384, ff_sin_32768, ff_sin_65536, |
| 51 | }; |
| 52 | |
| 53 | /** Map one real FFT into two parallel real even and odd FFTs. Then interleave |
| 54 | * the two real FFTs into one complex FFT. Unmangle the results. |
| 55 | * ref: http://www.engineeringproductivitytools.com/stuff/T0001/PT10.HTM |
| 56 | */ |
| 57 | static void rdft_calc_c(RDFTContext *s, FFTSample *data) |
| 58 | { |
| 59 | int i, i1, i2; |
| 60 | FFTComplex ev, od; |
| 61 | const int n = 1 << s->nbits; |
| 62 | const float k1 = 0.5; |
| 63 | const float k2 = 0.5 - s->inverse; |
| 64 | const FFTSample *tcos = s->tcos; |
| 65 | const FFTSample *tsin = s->tsin; |
| 66 | |
| 67 | if (!s->inverse) { |
| 68 | s->fft.fft_permute(&s->fft, (FFTComplex*)data); |
| 69 | s->fft.fft_calc(&s->fft, (FFTComplex*)data); |
| 70 | } |
| 71 | /* i=0 is a special case because of packing, the DC term is real, so we |
| 72 | are going to throw the N/2 term (also real) in with it. */ |
| 73 | ev.re = data[0]; |
| 74 | data[0] = ev.re+data[1]; |
| 75 | data[1] = ev.re-data[1]; |
| 76 | for (i = 1; i < (n>>2); i++) { |
| 77 | i1 = 2*i; |
| 78 | i2 = n-i1; |
| 79 | /* Separate even and odd FFTs */ |
| 80 | ev.re = k1*(data[i1 ]+data[i2 ]); |
| 81 | od.im = -k2*(data[i1 ]-data[i2 ]); |
| 82 | ev.im = k1*(data[i1+1]-data[i2+1]); |
| 83 | od.re = k2*(data[i1+1]+data[i2+1]); |
| 84 | /* Apply twiddle factors to the odd FFT and add to the even FFT */ |
| 85 | data[i1 ] = ev.re + od.re*tcos[i] - od.im*tsin[i]; |
| 86 | data[i1+1] = ev.im + od.im*tcos[i] + od.re*tsin[i]; |
| 87 | data[i2 ] = ev.re - od.re*tcos[i] + od.im*tsin[i]; |
| 88 | data[i2+1] = -ev.im + od.im*tcos[i] + od.re*tsin[i]; |
| 89 | } |
| 90 | data[2*i+1]=s->sign_convention*data[2*i+1]; |
| 91 | if (s->inverse) { |
| 92 | data[0] *= k1; |
| 93 | data[1] *= k1; |
| 94 | s->fft.fft_permute(&s->fft, (FFTComplex*)data); |
| 95 | s->fft.fft_calc(&s->fft, (FFTComplex*)data); |
| 96 | } |
| 97 | } |
| 98 | |
| 99 | av_cold int ff_rdft_init(RDFTContext *s, int nbits, enum RDFTransformType trans) |
| 100 | { |
| 101 | int n = 1 << nbits; |
| 102 | |
| 103 | s->nbits = nbits; |
| 104 | s->inverse = trans == IDFT_C2R || trans == DFT_C2R; |
| 105 | s->sign_convention = trans == IDFT_R2C || trans == DFT_C2R ? 1 : -1; |
| 106 | |
| 107 | if (nbits < 4 || nbits > 16) |
| 108 | return -1; |
| 109 | |
| 110 | if (ff_fft_init(&s->fft, nbits-1, trans == IDFT_C2R || trans == IDFT_R2C) < 0) |
| 111 | return -1; |
| 112 | |
| 113 | ff_init_ff_cos_tabs(nbits); |
| 114 | s->tcos = ff_cos_tabs[nbits]; |
| 115 | s->tsin = ff_sin_tabs[nbits]+(trans == DFT_R2C || trans == DFT_C2R)*(n>>2); |
| 116 | #if !CONFIG_HARDCODED_TABLES |
| 117 | { |
| 118 | int i; |
| 119 | const double theta = (trans == DFT_R2C || trans == DFT_C2R ? -1 : 1) * 2 * M_PI / n; |
| 120 | for (i = 0; i < (n >> 2); i++) |
| 121 | s->tsin[i] = sin(i * theta); |
| 122 | } |
| 123 | #endif |
| 124 | s->rdft_calc = rdft_calc_c; |
| 125 | |
| 126 | if (ARCH_ARM) ff_rdft_init_arm(s); |
| 127 | |
| 128 | return 0; |
| 129 | } |
| 130 | |
| 131 | av_cold void ff_rdft_end(RDFTContext *s) |
| 132 | { |
| 133 | ff_fft_end(&s->fft); |
| 134 | } |