-/*\r
-\r
- * jfdctflt.c\r
-\r
- *\r
-\r
- * Copyright (C) 1994, Thomas G. Lane.\r
-\r
- * This file is part of the Independent JPEG Group's software.\r
-\r
- * For conditions of distribution and use, see the accompanying README file.\r
-\r
- *\r
-\r
- * This file contains a floating-point implementation of the\r
-\r
- * forward DCT (Discrete Cosine Transform).\r
-\r
- *\r
-\r
- * This implementation should be more accurate than either of the integer\r
-\r
- * DCT implementations. However, it may not give the same results on all\r
-\r
- * machines because of differences in roundoff behavior. Speed will depend\r
-\r
- * on the hardware's floating point capacity.\r
-\r
- *\r
-\r
- * A 2-D DCT can be done by 1-D DCT on each row followed by 1-D DCT\r
-\r
- * on each column. Direct algorithms are also available, but they are\r
-\r
- * much more complex and seem not to be any faster when reduced to code.\r
-\r
- *\r
-\r
- * This implementation is based on Arai, Agui, and Nakajima's algorithm for\r
-\r
- * scaled DCT. Their original paper (Trans. IEICE E-71(11):1095) is in\r
-\r
- * Japanese, but the algorithm is described in the Pennebaker & Mitchell\r
-\r
- * JPEG textbook (see REFERENCES section in file README). The following code\r
-\r
- * is based directly on figure 4-8 in P&M.\r
-\r
- * While an 8-point DCT cannot be done in less than 11 multiplies, it is\r
-\r
- * possible to arrange the computation so that many of the multiplies are\r
-\r
- * simple scalings of the final outputs. These multiplies can then be\r
-\r
- * folded into the multiplications or divisions by the JPEG quantization\r
-\r
- * table entries. The AA&N method leaves only 5 multiplies and 29 adds\r
-\r
- * to be done in the DCT itself.\r
-\r
- * The primary disadvantage of this method is that with a fixed-point\r
-\r
- * implementation, accuracy is lost due to imprecise representation of the\r
-\r
- * scaled quantization values. However, that problem does not arise if\r
-\r
- * we use floating point arithmetic.\r
-\r
- */\r
-\r
-\r
-\r
-#define JPEG_INTERNALS\r
-\r
-#include "jinclude.h"\r
-\r
-#include "radiant_jpeglib.h"\r
-\r
-#include "jdct.h" /* Private declarations for DCT subsystem */\r
-\r
-\r
-\r
-#ifdef DCT_FLOAT_SUPPORTED\r
-\r
-\r
-\r
-\r
-\r
-/*\r
-\r
- * This module is specialized to the case DCTSIZE = 8.\r
-\r
- */\r
-\r
-\r
-\r
-#if DCTSIZE != 8\r
-\r
- Sorry, this code only copes with 8x8 DCTs. /* deliberate syntax err */\r
-\r
-#endif\r
-\r
-\r
-\r
-\r
-\r
-/*\r
-\r
- * Perform the forward DCT on one block of samples.\r
-\r
- */\r
-\r
-\r
-\r
-GLOBAL void\r
-\r
-jpeg_fdct_float (FAST_FLOAT * data)\r
-\r
-{\r
-\r
- FAST_FLOAT tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;\r
-\r
- FAST_FLOAT tmp10, tmp11, tmp12, tmp13;\r
-\r
- FAST_FLOAT z1, z2, z3, z4, z5, z11, z13;\r
-\r
- FAST_FLOAT *dataptr;\r
-\r
- int ctr;\r
-\r
-\r
-\r
- /* Pass 1: process rows. */\r
-\r
-\r
-\r
- dataptr = data;\r
-\r
- for (ctr = DCTSIZE-1; ctr >= 0; ctr--) {\r
-\r
- tmp0 = dataptr[0] + dataptr[7];\r
-\r
- tmp7 = dataptr[0] - dataptr[7];\r
-\r
- tmp1 = dataptr[1] + dataptr[6];\r
-\r
- tmp6 = dataptr[1] - dataptr[6];\r
-\r
- tmp2 = dataptr[2] + dataptr[5];\r
-\r
- tmp5 = dataptr[2] - dataptr[5];\r
-\r
- tmp3 = dataptr[3] + dataptr[4];\r
-\r
- tmp4 = dataptr[3] - dataptr[4];\r
-\r
- \r
-\r
- /* Even part */\r
-\r
- \r
-\r
- tmp10 = tmp0 + tmp3; /* phase 2 */\r
-\r
- tmp13 = tmp0 - tmp3;\r
-\r
- tmp11 = tmp1 + tmp2;\r
-\r
- tmp12 = tmp1 - tmp2;\r
-\r
- \r
-\r
- dataptr[0] = tmp10 + tmp11; /* phase 3 */\r
-\r
- dataptr[4] = tmp10 - tmp11;\r
-\r
- \r
-\r
- z1 = (tmp12 + tmp13) * ((FAST_FLOAT) 0.707106781); /* c4 */\r
-\r
- dataptr[2] = tmp13 + z1; /* phase 5 */\r
-\r
- dataptr[6] = tmp13 - z1;\r
-\r
- \r
-\r
- /* Odd part */\r
-\r
-\r
-\r
- tmp10 = tmp4 + tmp5; /* phase 2 */\r
-\r
- tmp11 = tmp5 + tmp6;\r
-\r
- tmp12 = tmp6 + tmp7;\r
-\r
-\r
-\r
- /* The rotator is modified from fig 4-8 to avoid extra negations. */\r
-\r
- z5 = (tmp10 - tmp12) * ((FAST_FLOAT) 0.382683433); /* c6 */\r
-\r
- z2 = ((FAST_FLOAT) 0.541196100) * tmp10 + z5; /* c2-c6 */\r
-\r
- z4 = ((FAST_FLOAT) 1.306562965) * tmp12 + z5; /* c2+c6 */\r
-\r
- z3 = tmp11 * ((FAST_FLOAT) 0.707106781); /* c4 */\r
-\r
-\r
-\r
- z11 = tmp7 + z3; /* phase 5 */\r
-\r
- z13 = tmp7 - z3;\r
-\r
-\r
-\r
- dataptr[5] = z13 + z2; /* phase 6 */\r
-\r
- dataptr[3] = z13 - z2;\r
-\r
- dataptr[1] = z11 + z4;\r
-\r
- dataptr[7] = z11 - z4;\r
-\r
-\r
-\r
- dataptr += DCTSIZE; /* advance pointer to next row */\r
-\r
- }\r
-\r
-\r
-\r
- /* Pass 2: process columns. */\r
-\r
-\r
-\r
- dataptr = data;\r
-\r
- for (ctr = DCTSIZE-1; ctr >= 0; ctr--) {\r
-\r
- tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*7];\r
-\r
- tmp7 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*7];\r
-\r
- tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*6];\r
-\r
- tmp6 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*6];\r
-\r
- tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*5];\r
-\r
- tmp5 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*5];\r
-\r
- tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*4];\r
-\r
- tmp4 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*4];\r
-\r
- \r
-\r
- /* Even part */\r
-\r
- \r
-\r
- tmp10 = tmp0 + tmp3; /* phase 2 */\r
-\r
- tmp13 = tmp0 - tmp3;\r
-\r
- tmp11 = tmp1 + tmp2;\r
-\r
- tmp12 = tmp1 - tmp2;\r
-\r
- \r
-\r
- dataptr[DCTSIZE*0] = tmp10 + tmp11; /* phase 3 */\r
-\r
- dataptr[DCTSIZE*4] = tmp10 - tmp11;\r
-\r
- \r
-\r
- z1 = (tmp12 + tmp13) * ((FAST_FLOAT) 0.707106781); /* c4 */\r
-\r
- dataptr[DCTSIZE*2] = tmp13 + z1; /* phase 5 */\r
-\r
- dataptr[DCTSIZE*6] = tmp13 - z1;\r
-\r
- \r
-\r
- /* Odd part */\r
-\r
-\r
-\r
- tmp10 = tmp4 + tmp5; /* phase 2 */\r
-\r
- tmp11 = tmp5 + tmp6;\r
-\r
- tmp12 = tmp6 + tmp7;\r
-\r
-\r
-\r
- /* The rotator is modified from fig 4-8 to avoid extra negations. */\r
-\r
- z5 = (tmp10 - tmp12) * ((FAST_FLOAT) 0.382683433); /* c6 */\r
-\r
- z2 = ((FAST_FLOAT) 0.541196100) * tmp10 + z5; /* c2-c6 */\r
-\r
- z4 = ((FAST_FLOAT) 1.306562965) * tmp12 + z5; /* c2+c6 */\r
-\r
- z3 = tmp11 * ((FAST_FLOAT) 0.707106781); /* c4 */\r
-\r
-\r
-\r
- z11 = tmp7 + z3; /* phase 5 */\r
-\r
- z13 = tmp7 - z3;\r
-\r
-\r
-\r
- dataptr[DCTSIZE*5] = z13 + z2; /* phase 6 */\r
-\r
- dataptr[DCTSIZE*3] = z13 - z2;\r
-\r
- dataptr[DCTSIZE*1] = z11 + z4;\r
-\r
- dataptr[DCTSIZE*7] = z11 - z4;\r
-\r
-\r
-\r
- dataptr++; /* advance pointer to next column */\r
-\r
- }\r
-\r
-}\r
-\r
-\r
-\r
-#endif /* DCT_FLOAT_SUPPORTED */\r
-\r
+/*
+
+ * jfdctflt.c
+
+ *
+
+ * Copyright (C) 1994, Thomas G. Lane.
+
+ * This file is part of the Independent JPEG Group's software.
+
+ * For conditions of distribution and use, see the accompanying README file.
+
+ *
+
+ * This file contains a floating-point implementation of the
+
+ * forward DCT (Discrete Cosine Transform).
+
+ *
+
+ * This implementation should be more accurate than either of the integer
+
+ * DCT implementations. However, it may not give the same results on all
+
+ * machines because of differences in roundoff behavior. Speed will depend
+
+ * on the hardware's floating point capacity.
+
+ *
+
+ * A 2-D DCT can be done by 1-D DCT on each row followed by 1-D DCT
+
+ * on each column. Direct algorithms are also available, but they are
+
+ * much more complex and seem not to be any faster when reduced to code.
+
+ *
+
+ * This implementation is based on Arai, Agui, and Nakajima's algorithm for
+
+ * scaled DCT. Their original paper (Trans. IEICE E-71(11):1095) is in
+
+ * Japanese, but the algorithm is described in the Pennebaker & Mitchell
+
+ * JPEG textbook (see REFERENCES section in file README). The following code
+
+ * is based directly on figure 4-8 in P&M.
+
+ * While an 8-point DCT cannot be done in less than 11 multiplies, it is
+
+ * possible to arrange the computation so that many of the multiplies are
+
+ * simple scalings of the final outputs. These multiplies can then be
+
+ * folded into the multiplications or divisions by the JPEG quantization
+
+ * table entries. The AA&N method leaves only 5 multiplies and 29 adds
+
+ * to be done in the DCT itself.
+
+ * The primary disadvantage of this method is that with a fixed-point
+
+ * implementation, accuracy is lost due to imprecise representation of the
+
+ * scaled quantization values. However, that problem does not arise if
+
+ * we use floating point arithmetic.
+
+ */
+
+
+
+#define JPEG_INTERNALS
+
+#include "jinclude.h"
+
+#include "radiant_jpeglib.h"
+
+#include "jdct.h" /* Private declarations for DCT subsystem */
+
+
+
+#ifdef DCT_FLOAT_SUPPORTED
+
+
+
+
+
+/*
+
+ * This module is specialized to the case DCTSIZE = 8.
+
+ */
+
+
+
+#if DCTSIZE != 8
+
+ Sorry, this code only copes with 8x8 DCTs. /* deliberate syntax err */
+
+#endif
+
+
+
+
+
+/*
+
+ * Perform the forward DCT on one block of samples.
+
+ */
+
+
+
+GLOBAL void
+
+jpeg_fdct_float (FAST_FLOAT * data)
+
+{
+
+ FAST_FLOAT tmp0, tmp1, tmp2, tmp3, tmp4, tmp5, tmp6, tmp7;
+
+ FAST_FLOAT tmp10, tmp11, tmp12, tmp13;
+
+ FAST_FLOAT z1, z2, z3, z4, z5, z11, z13;
+
+ FAST_FLOAT *dataptr;
+
+ int ctr;
+
+
+
+ /* Pass 1: process rows. */
+
+
+
+ dataptr = data;
+
+ for (ctr = DCTSIZE-1; ctr >= 0; ctr--) {
+
+ tmp0 = dataptr[0] + dataptr[7];
+
+ tmp7 = dataptr[0] - dataptr[7];
+
+ tmp1 = dataptr[1] + dataptr[6];
+
+ tmp6 = dataptr[1] - dataptr[6];
+
+ tmp2 = dataptr[2] + dataptr[5];
+
+ tmp5 = dataptr[2] - dataptr[5];
+
+ tmp3 = dataptr[3] + dataptr[4];
+
+ tmp4 = dataptr[3] - dataptr[4];
+
+
+
+ /* Even part */
+
+
+
+ tmp10 = tmp0 + tmp3; /* phase 2 */
+
+ tmp13 = tmp0 - tmp3;
+
+ tmp11 = tmp1 + tmp2;
+
+ tmp12 = tmp1 - tmp2;
+
+
+
+ dataptr[0] = tmp10 + tmp11; /* phase 3 */
+
+ dataptr[4] = tmp10 - tmp11;
+
+
+
+ z1 = (tmp12 + tmp13) * ((FAST_FLOAT) 0.707106781); /* c4 */
+
+ dataptr[2] = tmp13 + z1; /* phase 5 */
+
+ dataptr[6] = tmp13 - z1;
+
+
+
+ /* Odd part */
+
+
+
+ tmp10 = tmp4 + tmp5; /* phase 2 */
+
+ tmp11 = tmp5 + tmp6;
+
+ tmp12 = tmp6 + tmp7;
+
+
+
+ /* The rotator is modified from fig 4-8 to avoid extra negations. */
+
+ z5 = (tmp10 - tmp12) * ((FAST_FLOAT) 0.382683433); /* c6 */
+
+ z2 = ((FAST_FLOAT) 0.541196100) * tmp10 + z5; /* c2-c6 */
+
+ z4 = ((FAST_FLOAT) 1.306562965) * tmp12 + z5; /* c2+c6 */
+
+ z3 = tmp11 * ((FAST_FLOAT) 0.707106781); /* c4 */
+
+
+
+ z11 = tmp7 + z3; /* phase 5 */
+
+ z13 = tmp7 - z3;
+
+
+
+ dataptr[5] = z13 + z2; /* phase 6 */
+
+ dataptr[3] = z13 - z2;
+
+ dataptr[1] = z11 + z4;
+
+ dataptr[7] = z11 - z4;
+
+
+
+ dataptr += DCTSIZE; /* advance pointer to next row */
+
+ }
+
+
+
+ /* Pass 2: process columns. */
+
+
+
+ dataptr = data;
+
+ for (ctr = DCTSIZE-1; ctr >= 0; ctr--) {
+
+ tmp0 = dataptr[DCTSIZE*0] + dataptr[DCTSIZE*7];
+
+ tmp7 = dataptr[DCTSIZE*0] - dataptr[DCTSIZE*7];
+
+ tmp1 = dataptr[DCTSIZE*1] + dataptr[DCTSIZE*6];
+
+ tmp6 = dataptr[DCTSIZE*1] - dataptr[DCTSIZE*6];
+
+ tmp2 = dataptr[DCTSIZE*2] + dataptr[DCTSIZE*5];
+
+ tmp5 = dataptr[DCTSIZE*2] - dataptr[DCTSIZE*5];
+
+ tmp3 = dataptr[DCTSIZE*3] + dataptr[DCTSIZE*4];
+
+ tmp4 = dataptr[DCTSIZE*3] - dataptr[DCTSIZE*4];
+
+
+
+ /* Even part */
+
+
+
+ tmp10 = tmp0 + tmp3; /* phase 2 */
+
+ tmp13 = tmp0 - tmp3;
+
+ tmp11 = tmp1 + tmp2;
+
+ tmp12 = tmp1 - tmp2;
+
+
+
+ dataptr[DCTSIZE*0] = tmp10 + tmp11; /* phase 3 */
+
+ dataptr[DCTSIZE*4] = tmp10 - tmp11;
+
+
+
+ z1 = (tmp12 + tmp13) * ((FAST_FLOAT) 0.707106781); /* c4 */
+
+ dataptr[DCTSIZE*2] = tmp13 + z1; /* phase 5 */
+
+ dataptr[DCTSIZE*6] = tmp13 - z1;
+
+
+
+ /* Odd part */
+
+
+
+ tmp10 = tmp4 + tmp5; /* phase 2 */
+
+ tmp11 = tmp5 + tmp6;
+
+ tmp12 = tmp6 + tmp7;
+
+
+
+ /* The rotator is modified from fig 4-8 to avoid extra negations. */
+
+ z5 = (tmp10 - tmp12) * ((FAST_FLOAT) 0.382683433); /* c6 */
+
+ z2 = ((FAST_FLOAT) 0.541196100) * tmp10 + z5; /* c2-c6 */
+
+ z4 = ((FAST_FLOAT) 1.306562965) * tmp12 + z5; /* c2+c6 */
+
+ z3 = tmp11 * ((FAST_FLOAT) 0.707106781); /* c4 */
+
+
+
+ z11 = tmp7 + z3; /* phase 5 */
+
+ z13 = tmp7 - z3;
+
+
+
+ dataptr[DCTSIZE*5] = z13 + z2; /* phase 6 */
+
+ dataptr[DCTSIZE*3] = z13 - z2;
+
+ dataptr[DCTSIZE*1] = z11 + z4;
+
+ dataptr[DCTSIZE*7] = z11 - z4;
+
+
+
+ dataptr++; /* advance pointer to next column */
+
+ }
+
+}
+
+
+
+#endif /* DCT_FLOAT_SUPPORTED */
+
projects / xonotic / netradiant.git / blobdiff