faandct.c
Go to the documentation of this file.
1 /*
2  * Floating point AAN DCT
3  * this implementation is based upon the IJG integer AAN DCT (see jfdctfst.c)
4  *
5  * Copyright (c) 2003 Michael Niedermayer <michaelni@gmx.at>
6  * Copyright (c) 2003 Roman Shaposhnik
7  *
8  * Permission to use, copy, modify, and/or distribute this software for any
9  * purpose with or without fee is hereby granted, provided that the above
10  * copyright notice and this permission notice appear in all copies.
11  *
12  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
13  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
14  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
15  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
16  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
17  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
18  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
19  */
20 
28 #include "dsputil.h"
29 #include "faandct.h"
30 
31 #define FLOAT float
32 #ifdef FAAN_POSTSCALE
33 # define SCALE(x) postscale[x]
34 #else
35 # define SCALE(x) 1
36 #endif
37 
38 //numbers generated by simple c code (not as accurate as they could be)
39 /*
40 for(i=0; i<8; i++){
41  printf("#define B%d %1.20llf\n", i, (long double)1.0/(cosl(i*acosl(-1.0)/(long double)16.0)*sqrtl(2)));
42 }
43 */
44 #define B0 1.00000000000000000000
45 #define B1 0.72095982200694791383 // (cos(pi*1/16)sqrt(2))^-1
46 #define B2 0.76536686473017954350 // (cos(pi*2/16)sqrt(2))^-1
47 #define B3 0.85043009476725644878 // (cos(pi*3/16)sqrt(2))^-1
48 #define B4 1.00000000000000000000 // (cos(pi*4/16)sqrt(2))^-1
49 #define B5 1.27275858057283393842 // (cos(pi*5/16)sqrt(2))^-1
50 #define B6 1.84775906502257351242 // (cos(pi*6/16)sqrt(2))^-1
51 #define B7 3.62450978541155137218 // (cos(pi*7/16)sqrt(2))^-1
52 
53 
54 #define A1 0.70710678118654752438 // cos(pi*4/16)
55 #define A2 0.54119610014619698435 // cos(pi*6/16)sqrt(2)
56 #define A5 0.38268343236508977170 // cos(pi*6/16)
57 #define A4 1.30656296487637652774 // cos(pi*2/16)sqrt(2)
58 
59 static const FLOAT postscale[64]={
60 B0*B0, B0*B1, B0*B2, B0*B3, B0*B4, B0*B5, B0*B6, B0*B7,
61 B1*B0, B1*B1, B1*B2, B1*B3, B1*B4, B1*B5, B1*B6, B1*B7,
62 B2*B0, B2*B1, B2*B2, B2*B3, B2*B4, B2*B5, B2*B6, B2*B7,
63 B3*B0, B3*B1, B3*B2, B3*B3, B3*B4, B3*B5, B3*B6, B3*B7,
64 B4*B0, B4*B1, B4*B2, B4*B3, B4*B4, B4*B5, B4*B6, B4*B7,
65 B5*B0, B5*B1, B5*B2, B5*B3, B5*B4, B5*B5, B5*B6, B5*B7,
66 B6*B0, B6*B1, B6*B2, B6*B3, B6*B4, B6*B5, B6*B6, B6*B7,
67 B7*B0, B7*B1, B7*B2, B7*B3, B7*B4, B7*B5, B7*B6, B7*B7,
68 };
69 
70 static av_always_inline void row_fdct(FLOAT temp[64], DCTELEM * data)
71 {
72  FLOAT tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
73  FLOAT tmp10, tmp11, tmp12, tmp13;
74  FLOAT z2, z4, z11, z13;
75  FLOAT av_unused z5;
76  int i;
77 
78  for (i=0; i<8*8; i+=8) {
79  tmp0= data[0 + i] + data[7 + i];
80  tmp7= data[0 + i] - data[7 + i];
81  tmp1= data[1 + i] + data[6 + i];
82  tmp6= data[1 + i] - data[6 + i];
83  tmp2= data[2 + i] + data[5 + i];
84  tmp5= data[2 + i] - data[5 + i];
85  tmp3= data[3 + i] + data[4 + i];
86  tmp4= data[3 + i] - data[4 + i];
87 
88  tmp10= tmp0 + tmp3;
89  tmp13= tmp0 - tmp3;
90  tmp11= tmp1 + tmp2;
91  tmp12= tmp1 - tmp2;
92 
93  temp[0 + i]= tmp10 + tmp11;
94  temp[4 + i]= tmp10 - tmp11;
95 
96  tmp12 += tmp13;
97  tmp12 *= A1;
98  temp[2 + i]= tmp13 + tmp12;
99  temp[6 + i]= tmp13 - tmp12;
100 
101  tmp4 += tmp5;
102  tmp5 += tmp6;
103  tmp6 += tmp7;
104 
105 #if 0
106  z5= (tmp4 - tmp6) * A5;
107  z2= tmp4*A2 + z5;
108  z4= tmp6*A4 + z5;
109 #else
110  z2= tmp4*(A2+A5) - tmp6*A5;
111  z4= tmp6*(A4-A5) + tmp4*A5;
112 #endif
113  tmp5*=A1;
114 
115  z11= tmp7 + tmp5;
116  z13= tmp7 - tmp5;
117 
118  temp[5 + i]= z13 + z2;
119  temp[3 + i]= z13 - z2;
120  temp[1 + i]= z11 + z4;
121  temp[7 + i]= z11 - z4;
122  }
123 }
124 
126 {
127  FLOAT tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
128  FLOAT tmp10, tmp11, tmp12, tmp13;
129  FLOAT z2, z4, z11, z13;
130  FLOAT av_unused z5;
131  FLOAT temp[64];
132  int i;
133 
134  emms_c();
135 
136  row_fdct(temp, data);
137 
138  for (i=0; i<8; i++) {
139  tmp0= temp[8*0 + i] + temp[8*7 + i];
140  tmp7= temp[8*0 + i] - temp[8*7 + i];
141  tmp1= temp[8*1 + i] + temp[8*6 + i];
142  tmp6= temp[8*1 + i] - temp[8*6 + i];
143  tmp2= temp[8*2 + i] + temp[8*5 + i];
144  tmp5= temp[8*2 + i] - temp[8*5 + i];
145  tmp3= temp[8*3 + i] + temp[8*4 + i];
146  tmp4= temp[8*3 + i] - temp[8*4 + i];
147 
148  tmp10= tmp0 + tmp3;
149  tmp13= tmp0 - tmp3;
150  tmp11= tmp1 + tmp2;
151  tmp12= tmp1 - tmp2;
152 
153  data[8*0 + i]= lrintf(SCALE(8*0 + i) * (tmp10 + tmp11));
154  data[8*4 + i]= lrintf(SCALE(8*4 + i) * (tmp10 - tmp11));
155 
156  tmp12 += tmp13;
157  tmp12 *= A1;
158  data[8*2 + i]= lrintf(SCALE(8*2 + i) * (tmp13 + tmp12));
159  data[8*6 + i]= lrintf(SCALE(8*6 + i) * (tmp13 - tmp12));
160 
161  tmp4 += tmp5;
162  tmp5 += tmp6;
163  tmp6 += tmp7;
164 
165 #if 0
166  z5= (tmp4 - tmp6) * A5;
167  z2= tmp4*A2 + z5;
168  z4= tmp6*A4 + z5;
169 #else
170  z2= tmp4*(A2+A5) - tmp6*A5;
171  z4= tmp6*(A4-A5) + tmp4*A5;
172 #endif
173  tmp5*=A1;
174 
175  z11= tmp7 + tmp5;
176  z13= tmp7 - tmp5;
177 
178  data[8*5 + i]= lrintf(SCALE(8*5 + i) * (z13 + z2));
179  data[8*3 + i]= lrintf(SCALE(8*3 + i) * (z13 - z2));
180  data[8*1 + i]= lrintf(SCALE(8*1 + i) * (z11 + z4));
181  data[8*7 + i]= lrintf(SCALE(8*7 + i) * (z11 - z4));
182  }
183 }
184 
186 {
187  FLOAT tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
188  FLOAT tmp10, tmp11, tmp12, tmp13;
189  FLOAT temp[64];
190  int i;
191 
192  emms_c();
193 
194  row_fdct(temp, data);
195 
196  for (i=0; i<8; i++) {
197  tmp0 = temp[8*0 + i] + temp[8*1 + i];
198  tmp1 = temp[8*2 + i] + temp[8*3 + i];
199  tmp2 = temp[8*4 + i] + temp[8*5 + i];
200  tmp3 = temp[8*6 + i] + temp[8*7 + i];
201  tmp4 = temp[8*0 + i] - temp[8*1 + i];
202  tmp5 = temp[8*2 + i] - temp[8*3 + i];
203  tmp6 = temp[8*4 + i] - temp[8*5 + i];
204  tmp7 = temp[8*6 + i] - temp[8*7 + i];
205 
206  tmp10 = tmp0 + tmp3;
207  tmp11 = tmp1 + tmp2;
208  tmp12 = tmp1 - tmp2;
209  tmp13 = tmp0 - tmp3;
210 
211  data[8*0 + i] = lrintf(SCALE(8*0 + i) * (tmp10 + tmp11));
212  data[8*4 + i] = lrintf(SCALE(8*4 + i) * (tmp10 - tmp11));
213 
214  tmp12 += tmp13;
215  tmp12 *= A1;
216  data[8*2 + i] = lrintf(SCALE(8*2 + i) * (tmp13 + tmp12));
217  data[8*6 + i] = lrintf(SCALE(8*6 + i) * (tmp13 - tmp12));
218 
219  tmp10 = tmp4 + tmp7;
220  tmp11 = tmp5 + tmp6;
221  tmp12 = tmp5 - tmp6;
222  tmp13 = tmp4 - tmp7;
223 
224  data[8*1 + i] = lrintf(SCALE(8*0 + i) * (tmp10 + tmp11));
225  data[8*5 + i] = lrintf(SCALE(8*4 + i) * (tmp10 - tmp11));
226 
227  tmp12 += tmp13;
228  tmp12 *= A1;
229  data[8*3 + i] = lrintf(SCALE(8*2 + i) * (tmp13 + tmp12));
230  data[8*7 + i] = lrintf(SCALE(8*6 + i) * (tmp13 - tmp12));
231  }
232 }