/src/libjpeg-turbo.main/jidctflt.c
Line | Count | Source (jump to first uncovered line) |
1 | | /* |
2 | | * jidctflt.c |
3 | | * |
4 | | * This file was part of the Independent JPEG Group's software: |
5 | | * Copyright (C) 1994-1998, Thomas G. Lane. |
6 | | * Modified 2010 by Guido Vollbeding. |
7 | | * libjpeg-turbo Modifications: |
8 | | * Copyright (C) 2014, D. R. Commander. |
9 | | * For conditions of distribution and use, see the accompanying README.ijg |
10 | | * file. |
11 | | * |
12 | | * This file contains a floating-point implementation of the |
13 | | * inverse DCT (Discrete Cosine Transform). In the IJG code, this routine |
14 | | * must also perform dequantization of the input coefficients. |
15 | | * |
16 | | * This implementation should be more accurate than either of the integer |
17 | | * IDCT implementations. However, it may not give the same results on all |
18 | | * machines because of differences in roundoff behavior. Speed will depend |
19 | | * on the hardware's floating point capacity. |
20 | | * |
21 | | * A 2-D IDCT can be done by 1-D IDCT on each column followed by 1-D IDCT |
22 | | * on each row (or vice versa, but it's more convenient to emit a row at |
23 | | * a time). Direct algorithms are also available, but they are much more |
24 | | * complex and seem not to be any faster when reduced to code. |
25 | | * |
26 | | * This implementation is based on Arai, Agui, and Nakajima's algorithm for |
27 | | * scaled DCT. Their original paper (Trans. IEICE E-71(11):1095) is in |
28 | | * Japanese, but the algorithm is described in the Pennebaker & Mitchell |
29 | | * JPEG textbook (see REFERENCES section in file README.ijg). The following |
30 | | * code is based directly on figure 4-8 in P&M. |
31 | | * While an 8-point DCT cannot be done in less than 11 multiplies, it is |
32 | | * possible to arrange the computation so that many of the multiplies are |
33 | | * simple scalings of the final outputs. These multiplies can then be |
34 | | * folded into the multiplications or divisions by the JPEG quantization |
35 | | * table entries. The AA&N method leaves only 5 multiplies and 29 adds |
36 | | * to be done in the DCT itself. |
37 | | * The primary disadvantage of this method is that with a fixed-point |
38 | | * implementation, accuracy is lost due to imprecise representation of the |
39 | | * scaled quantization values. However, that problem does not arise if |
40 | | * we use floating point arithmetic. |
41 | | */ |
42 | | |
43 | | #define JPEG_INTERNALS |
44 | | #include "jinclude.h" |
45 | | #include "jpeglib.h" |
46 | | #include "jdct.h" /* Private declarations for DCT subsystem */ |
47 | | |
48 | | #ifdef DCT_FLOAT_SUPPORTED |
49 | | |
50 | | |
51 | | /* |
52 | | * This module is specialized to the case DCTSIZE = 8. |
53 | | */ |
54 | | |
55 | | #if DCTSIZE != 8 |
56 | | Sorry, this code only copes with 8x8 DCTs. /* deliberate syntax err */ |
57 | | #endif |
58 | | |
59 | | |
60 | | /* Dequantize a coefficient by multiplying it by the multiplier-table |
61 | | * entry; produce a float result. |
62 | | */ |
63 | | |
64 | 0 | #define DEQUANTIZE(coef, quantval) (((FAST_FLOAT)(coef)) * (quantval)) |
65 | | |
66 | | |
67 | | /* |
68 | | * Perform dequantization and inverse DCT on one block of coefficients. |
69 | | */ |
70 | | |
71 | | GLOBAL(void) |
72 | | jpeg_idct_float(j_decompress_ptr cinfo, jpeg_component_info *compptr, |
73 | | JCOEFPTR coef_block, JSAMPARRAY output_buf, |
74 | | JDIMENSION output_col) |
75 | 0 | { |
76 | 0 | FAST_FLOAT tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7; |
77 | 0 | FAST_FLOAT tmp10, tmp11, tmp12, tmp13; |
78 | 0 | FAST_FLOAT z5, z10, z11, z12, z13; |
79 | 0 | JCOEFPTR inptr; |
80 | 0 | FLOAT_MULT_TYPE *quantptr; |
81 | 0 | FAST_FLOAT *wsptr; |
82 | 0 | JSAMPROW outptr; |
83 | 0 | JSAMPLE *range_limit = cinfo->sample_range_limit; |
84 | 0 | int ctr; |
85 | 0 | FAST_FLOAT workspace[DCTSIZE2]; /* buffers data between passes */ |
86 | 0 | #define _0_125 ((FLOAT_MULT_TYPE)0.125) |
87 | | |
88 | | /* Pass 1: process columns from input, store into work array. */ |
89 | |
|
90 | 0 | inptr = coef_block; |
91 | 0 | quantptr = (FLOAT_MULT_TYPE *)compptr->dct_table; |
92 | 0 | wsptr = workspace; |
93 | 0 | for (ctr = DCTSIZE; ctr > 0; ctr--) { |
94 | | /* Due to quantization, we will usually find that many of the input |
95 | | * coefficients are zero, especially the AC terms. We can exploit this |
96 | | * by short-circuiting the IDCT calculation for any column in which all |
97 | | * the AC terms are zero. In that case each output is equal to the |
98 | | * DC coefficient (with scale factor as needed). |
99 | | * With typical images and quantization tables, half or more of the |
100 | | * column DCT calculations can be simplified this way. |
101 | | */ |
102 | |
|
103 | 0 | if (inptr[DCTSIZE * 1] == 0 && inptr[DCTSIZE * 2] == 0 && |
104 | 0 | inptr[DCTSIZE * 3] == 0 && inptr[DCTSIZE * 4] == 0 && |
105 | 0 | inptr[DCTSIZE * 5] == 0 && inptr[DCTSIZE * 6] == 0 && |
106 | 0 | inptr[DCTSIZE * 7] == 0) { |
107 | | /* AC terms all zero */ |
108 | 0 | FAST_FLOAT dcval = DEQUANTIZE(inptr[DCTSIZE * 0], |
109 | 0 | quantptr[DCTSIZE * 0] * _0_125); |
110 | |
|
111 | 0 | wsptr[DCTSIZE * 0] = dcval; |
112 | 0 | wsptr[DCTSIZE * 1] = dcval; |
113 | 0 | wsptr[DCTSIZE * 2] = dcval; |
114 | 0 | wsptr[DCTSIZE * 3] = dcval; |
115 | 0 | wsptr[DCTSIZE * 4] = dcval; |
116 | 0 | wsptr[DCTSIZE * 5] = dcval; |
117 | 0 | wsptr[DCTSIZE * 6] = dcval; |
118 | 0 | wsptr[DCTSIZE * 7] = dcval; |
119 | |
|
120 | 0 | inptr++; /* advance pointers to next column */ |
121 | 0 | quantptr++; |
122 | 0 | wsptr++; |
123 | 0 | continue; |
124 | 0 | } |
125 | | |
126 | | /* Even part */ |
127 | | |
128 | 0 | tmp0 = DEQUANTIZE(inptr[DCTSIZE * 0], quantptr[DCTSIZE * 0] * _0_125); |
129 | 0 | tmp1 = DEQUANTIZE(inptr[DCTSIZE * 2], quantptr[DCTSIZE * 2] * _0_125); |
130 | 0 | tmp2 = DEQUANTIZE(inptr[DCTSIZE * 4], quantptr[DCTSIZE * 4] * _0_125); |
131 | 0 | tmp3 = DEQUANTIZE(inptr[DCTSIZE * 6], quantptr[DCTSIZE * 6] * _0_125); |
132 | |
|
133 | 0 | tmp10 = tmp0 + tmp2; /* phase 3 */ |
134 | 0 | tmp11 = tmp0 - tmp2; |
135 | |
|
136 | 0 | tmp13 = tmp1 + tmp3; /* phases 5-3 */ |
137 | 0 | tmp12 = (tmp1 - tmp3) * ((FAST_FLOAT)1.414213562) - tmp13; /* 2*c4 */ |
138 | |
|
139 | 0 | tmp0 = tmp10 + tmp13; /* phase 2 */ |
140 | 0 | tmp3 = tmp10 - tmp13; |
141 | 0 | tmp1 = tmp11 + tmp12; |
142 | 0 | tmp2 = tmp11 - tmp12; |
143 | | |
144 | | /* Odd part */ |
145 | |
|
146 | 0 | tmp4 = DEQUANTIZE(inptr[DCTSIZE * 1], quantptr[DCTSIZE * 1] * _0_125); |
147 | 0 | tmp5 = DEQUANTIZE(inptr[DCTSIZE * 3], quantptr[DCTSIZE * 3] * _0_125); |
148 | 0 | tmp6 = DEQUANTIZE(inptr[DCTSIZE * 5], quantptr[DCTSIZE * 5] * _0_125); |
149 | 0 | tmp7 = DEQUANTIZE(inptr[DCTSIZE * 7], quantptr[DCTSIZE * 7] * _0_125); |
150 | |
|
151 | 0 | z13 = tmp6 + tmp5; /* phase 6 */ |
152 | 0 | z10 = tmp6 - tmp5; |
153 | 0 | z11 = tmp4 + tmp7; |
154 | 0 | z12 = tmp4 - tmp7; |
155 | |
|
156 | 0 | tmp7 = z11 + z13; /* phase 5 */ |
157 | 0 | tmp11 = (z11 - z13) * ((FAST_FLOAT)1.414213562); /* 2*c4 */ |
158 | |
|
159 | 0 | z5 = (z10 + z12) * ((FAST_FLOAT)1.847759065); /* 2*c2 */ |
160 | 0 | tmp10 = z5 - z12 * ((FAST_FLOAT)1.082392200); /* 2*(c2-c6) */ |
161 | 0 | tmp12 = z5 - z10 * ((FAST_FLOAT)2.613125930); /* 2*(c2+c6) */ |
162 | |
|
163 | 0 | tmp6 = tmp12 - tmp7; /* phase 2 */ |
164 | 0 | tmp5 = tmp11 - tmp6; |
165 | 0 | tmp4 = tmp10 - tmp5; |
166 | |
|
167 | 0 | wsptr[DCTSIZE * 0] = tmp0 + tmp7; |
168 | 0 | wsptr[DCTSIZE * 7] = tmp0 - tmp7; |
169 | 0 | wsptr[DCTSIZE * 1] = tmp1 + tmp6; |
170 | 0 | wsptr[DCTSIZE * 6] = tmp1 - tmp6; |
171 | 0 | wsptr[DCTSIZE * 2] = tmp2 + tmp5; |
172 | 0 | wsptr[DCTSIZE * 5] = tmp2 - tmp5; |
173 | 0 | wsptr[DCTSIZE * 3] = tmp3 + tmp4; |
174 | 0 | wsptr[DCTSIZE * 4] = tmp3 - tmp4; |
175 | |
|
176 | 0 | inptr++; /* advance pointers to next column */ |
177 | 0 | quantptr++; |
178 | 0 | wsptr++; |
179 | 0 | } |
180 | | |
181 | | /* Pass 2: process rows from work array, store into output array. */ |
182 | |
|
183 | 0 | wsptr = workspace; |
184 | 0 | for (ctr = 0; ctr < DCTSIZE; ctr++) { |
185 | 0 | outptr = output_buf[ctr] + output_col; |
186 | | /* Rows of zeroes can be exploited in the same way as we did with columns. |
187 | | * However, the column calculation has created many nonzero AC terms, so |
188 | | * the simplification applies less often (typically 5% to 10% of the time). |
189 | | * And testing floats for zero is relatively expensive, so we don't bother. |
190 | | */ |
191 | | |
192 | | /* Even part */ |
193 | | |
194 | | /* Apply signed->unsigned and prepare float->int conversion */ |
195 | 0 | z5 = wsptr[0] + ((FAST_FLOAT)CENTERJSAMPLE + (FAST_FLOAT)0.5); |
196 | 0 | tmp10 = z5 + wsptr[4]; |
197 | 0 | tmp11 = z5 - wsptr[4]; |
198 | |
|
199 | 0 | tmp13 = wsptr[2] + wsptr[6]; |
200 | 0 | tmp12 = (wsptr[2] - wsptr[6]) * ((FAST_FLOAT)1.414213562) - tmp13; |
201 | |
|
202 | 0 | tmp0 = tmp10 + tmp13; |
203 | 0 | tmp3 = tmp10 - tmp13; |
204 | 0 | tmp1 = tmp11 + tmp12; |
205 | 0 | tmp2 = tmp11 - tmp12; |
206 | | |
207 | | /* Odd part */ |
208 | |
|
209 | 0 | z13 = wsptr[5] + wsptr[3]; |
210 | 0 | z10 = wsptr[5] - wsptr[3]; |
211 | 0 | z11 = wsptr[1] + wsptr[7]; |
212 | 0 | z12 = wsptr[1] - wsptr[7]; |
213 | |
|
214 | 0 | tmp7 = z11 + z13; |
215 | 0 | tmp11 = (z11 - z13) * ((FAST_FLOAT)1.414213562); |
216 | |
|
217 | 0 | z5 = (z10 + z12) * ((FAST_FLOAT)1.847759065); /* 2*c2 */ |
218 | 0 | tmp10 = z5 - z12 * ((FAST_FLOAT)1.082392200); /* 2*(c2-c6) */ |
219 | 0 | tmp12 = z5 - z10 * ((FAST_FLOAT)2.613125930); /* 2*(c2+c6) */ |
220 | |
|
221 | 0 | tmp6 = tmp12 - tmp7; |
222 | 0 | tmp5 = tmp11 - tmp6; |
223 | 0 | tmp4 = tmp10 - tmp5; |
224 | | |
225 | | /* Final output stage: float->int conversion and range-limit */ |
226 | |
|
227 | 0 | outptr[0] = range_limit[((int)(tmp0 + tmp7)) & RANGE_MASK]; |
228 | 0 | outptr[7] = range_limit[((int)(tmp0 - tmp7)) & RANGE_MASK]; |
229 | 0 | outptr[1] = range_limit[((int)(tmp1 + tmp6)) & RANGE_MASK]; |
230 | 0 | outptr[6] = range_limit[((int)(tmp1 - tmp6)) & RANGE_MASK]; |
231 | 0 | outptr[2] = range_limit[((int)(tmp2 + tmp5)) & RANGE_MASK]; |
232 | 0 | outptr[5] = range_limit[((int)(tmp2 - tmp5)) & RANGE_MASK]; |
233 | 0 | outptr[3] = range_limit[((int)(tmp3 + tmp4)) & RANGE_MASK]; |
234 | 0 | outptr[4] = range_limit[((int)(tmp3 - tmp4)) & RANGE_MASK]; |
235 | |
|
236 | 0 | wsptr += DCTSIZE; /* advance pointer to next row */ |
237 | 0 | } |
238 | 0 | } |
239 | | |
240 | | #endif /* DCT_FLOAT_SUPPORTED */ |