285 lines
11 KiB
C++
285 lines
11 KiB
C++
/*
|
|
* jrevdct.c
|
|
*
|
|
* Copyright (C) 1991, 1992, 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 the basic inverse-DCT transformation subroutine.
|
|
*
|
|
* This implementation is based on an algorithm described in
|
|
* C. Loeffler, A. Ligtenberg and G. Moschytz, "Practical Fast 1-D DCT
|
|
* Algorithms with 11 Multiplications", Proc. Int'l. Conf. on Acoustics,
|
|
* Speech, and Signal Processing 1989 (ICASSP '89), pp. 988-991.
|
|
* The primary algorithm described there uses 11 multiplies and 29 adds.
|
|
* We use their alternate method with 12 multiplies and 32 adds.
|
|
* The advantage of this method is that no data path contains more than one
|
|
* multiplication; this allows a very simple and accurate implementation in
|
|
* scaled fixed-point arithmetic, with a minimal number of shifts.
|
|
*/
|
|
|
|
/* Modified 2007-2016 for usage in Mednafen */
|
|
|
|
#include "defs.h"
|
|
#include "jrevdct.h"
|
|
|
|
namespace MDFN_IEN_PCFX
|
|
{
|
|
/*
|
|
* This routine is specialized to the case DCTSIZE = 8.
|
|
*/
|
|
#define DCTSIZE 8
|
|
|
|
/*
|
|
* A 2-D IDCT can be done by 1-D IDCT on each row followed by 1-D IDCT
|
|
* 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.
|
|
*
|
|
* The poop on this scaling stuff is as follows:
|
|
*
|
|
* Each 1-D IDCT step produces outputs which are a factor of sqrt(N)
|
|
* larger than the true IDCT outputs. The final outputs are therefore
|
|
* a factor of N larger than desired; since N=8 this can be cured by
|
|
* a simple right shift at the end of the algorithm. The advantage of
|
|
* this arrangement is that we save two multiplications per 1-D IDCT,
|
|
* because the y0 and y4 inputs need not be divided by sqrt(N).
|
|
*
|
|
* We have to do addition and subtraction of the integer inputs, which
|
|
* is no problem, and multiplication by fractional constants, which is
|
|
* a problem to do in integer arithmetic. We multiply all the constants
|
|
* by CONST_SCALE and convert them to integer constants (thus retaining
|
|
* CONST_BITS bits of precision in the constants). After doing a
|
|
* multiplication we have to divide the product by CONST_SCALE, with proper
|
|
* rounding, to produce the correct output. This division can be done
|
|
* cheaply as a right shift of CONST_BITS bits. We postpone shifting
|
|
* as long as possible so that partial sums can be added together with
|
|
* full fractional precision.
|
|
*
|
|
* The outputs of the first pass are scaled up by PASS1_BITS bits so that
|
|
* they are represented to better-than-integral precision. These outputs
|
|
* require BITS_IN_JSAMPLE + PASS1_BITS + 3 bits; this fits in a 16-bit word
|
|
* with the recommended scaling. (To scale up 12-bit sample data further, an
|
|
* intermediate int32 array would be needed.)
|
|
*
|
|
* To avoid overflow of the 32-bit intermediate results in pass 2, we must
|
|
* have BITS_IN_JSAMPLE + CONST_BITS + PASS1_BITS <= 26. Error analysis
|
|
* shows that the values given below are the most effective.
|
|
*/
|
|
|
|
#define CONST_BITS 13
|
|
#define PASS1_BITS 2
|
|
|
|
#if ((8 + CONST_BITS + PASS1_BITS) > 26)
|
|
#error "Too many bits1!"
|
|
#endif
|
|
|
|
#define ONE ((int32) 1)
|
|
|
|
#define CONST_SCALE (ONE << CONST_BITS)
|
|
|
|
/* Convert a positive real constant to an integer scaled by CONST_SCALE. */
|
|
|
|
#define FIX(x) ((int32) ((x) * CONST_SCALE + 0.5))
|
|
|
|
/* Some C compilers fail to reduce "FIX(constant)" at compile time, thus
|
|
* causing a lot of useless floating-point operations at run time.
|
|
* To get around this we use the following pre-calculated constants.
|
|
* If you change CONST_BITS you may want to add appropriate values.
|
|
* (With a reasonable C compiler, you can just rely on the FIX() macro...)
|
|
*/
|
|
|
|
#if CONST_BITS == 13
|
|
#define FIX_0_298631336 ((int32) 2446) /* FIX(0.298631336) */
|
|
#define FIX_0_390180644 ((int32) 3196) /* FIX(0.390180644) */
|
|
#define FIX_0_541196100 ((int32) 4433) /* FIX(0.541196100) */
|
|
#define FIX_0_765366865 ((int32) 6270) /* FIX(0.765366865) */
|
|
#define FIX_0_899976223 ((int32) 7373) /* FIX(0.899976223) */
|
|
#define FIX_1_175875602 ((int32) 9633) /* FIX(1.175875602) */
|
|
#define FIX_1_501321110 ((int32) 12299) /* FIX(1.501321110) */
|
|
#define FIX_1_847759065 ((int32) 15137) /* FIX(1.847759065) */
|
|
#define FIX_1_961570560 ((int32) 16069) /* FIX(1.961570560) */
|
|
#define FIX_2_053119869 ((int32) 16819) /* FIX(2.053119869) */
|
|
#define FIX_2_562915447 ((int32) 20995) /* FIX(2.562915447) */
|
|
#define FIX_3_072711026 ((int32) 25172) /* FIX(3.072711026) */
|
|
#else
|
|
#define FIX_0_298631336 FIX(0.298631336)
|
|
#define FIX_0_390180644 FIX(0.390180644)
|
|
#define FIX_0_541196100 FIX(0.541196100)
|
|
#define FIX_0_765366865 FIX(0.765366865)
|
|
#define FIX_0_899976223 FIX(0.899976223)
|
|
#define FIX_1_175875602 FIX(1.175875602)
|
|
#define FIX_1_501321110 FIX(1.501321110)
|
|
#define FIX_1_847759065 FIX(1.847759065)
|
|
#define FIX_1_961570560 FIX(1.961570560)
|
|
#define FIX_2_053119869 FIX(2.053119869)
|
|
#define FIX_2_562915447 FIX(2.562915447)
|
|
#define FIX_3_072711026 FIX(3.072711026)
|
|
#endif
|
|
|
|
|
|
/* Descale and correctly round an int32 value that's scaled by N bits.
|
|
* We assume RIGHT_SHIFT rounds towards minus infinity, so adding
|
|
* the fudge factor is correct for either sign of X.
|
|
*/
|
|
|
|
#define RIGHT_SHIFT(x,shft) ((x) >> (shft))
|
|
#define DESCALE(x,n) RIGHT_SHIFT((x) + (ONE << ((n)-1)), n)
|
|
#define MULTIPLY(var,const) ((var) * (const))
|
|
|
|
|
|
/*
|
|
* Perform the inverse DCT on one block of coefficients.
|
|
*/
|
|
|
|
void j_rev_dct(DCTBLOCK data)
|
|
{
|
|
int32 tmp0, tmp1, tmp2, tmp3;
|
|
int32 tmp10, tmp11, tmp12, tmp13;
|
|
int32 z1, z2, z3, z4, z5;
|
|
register DCTELEM *dataptr;
|
|
int rowctr;
|
|
|
|
/* Pass 1: process rows. */
|
|
/* Note results are scaled up by sqrt(8) compared to a true IDCT; */
|
|
/* furthermore, we scale the results by 2**PASS1_BITS. */
|
|
|
|
dataptr = data;
|
|
for (rowctr = DCTSIZE-1; rowctr >= 0; rowctr--)
|
|
{
|
|
/* Even part: reverse the even part of the forward DCT. */
|
|
/* The rotator is sqrt(2)*c(-6). */
|
|
|
|
z2 = (int32) dataptr[2];
|
|
z3 = (int32) dataptr[6];
|
|
|
|
z1 = MULTIPLY(z2 + z3, FIX_0_541196100);
|
|
tmp2 = z1 + MULTIPLY(z3, - FIX_1_847759065);
|
|
tmp3 = z1 + MULTIPLY(z2, FIX_0_765366865);
|
|
|
|
tmp0 = ((uint32) dataptr[0] + (uint32) dataptr[4]) << CONST_BITS;
|
|
tmp1 = ((uint32) dataptr[0] - (uint32) dataptr[4]) << CONST_BITS;
|
|
|
|
tmp10 = tmp0 + tmp3;
|
|
tmp13 = tmp0 - tmp3;
|
|
tmp11 = tmp1 + tmp2;
|
|
tmp12 = tmp1 - tmp2;
|
|
|
|
/* Odd part per figure 8; the matrix is unitary and hence its
|
|
* transpose is its inverse. i0..i3 are y7,y5,y3,y1 respectively.
|
|
*/
|
|
|
|
tmp0 = (int32) dataptr[7];
|
|
tmp1 = (int32) dataptr[5];
|
|
tmp2 = (int32) dataptr[3];
|
|
tmp3 = (int32) dataptr[1];
|
|
|
|
z1 = tmp0 + tmp3;
|
|
z2 = tmp1 + tmp2;
|
|
z3 = tmp0 + tmp2;
|
|
z4 = tmp1 + tmp3;
|
|
z5 = MULTIPLY(z3 + z4, FIX_1_175875602); /* sqrt(2) * c3 */
|
|
|
|
tmp0 = MULTIPLY(tmp0, FIX_0_298631336); /* sqrt(2) * (-c1+c3+c5-c7) */
|
|
tmp1 = MULTIPLY(tmp1, FIX_2_053119869); /* sqrt(2) * ( c1+c3-c5+c7) */
|
|
tmp2 = MULTIPLY(tmp2, FIX_3_072711026); /* sqrt(2) * ( c1+c3+c5-c7) */
|
|
tmp3 = MULTIPLY(tmp3, FIX_1_501321110); /* sqrt(2) * ( c1+c3-c5-c7) */
|
|
z1 = MULTIPLY(z1, - FIX_0_899976223); /* sqrt(2) * (c7-c3) */
|
|
z2 = MULTIPLY(z2, - FIX_2_562915447); /* sqrt(2) * (-c1-c3) */
|
|
z3 = MULTIPLY(z3, - FIX_1_961570560); /* sqrt(2) * (-c3-c5) */
|
|
z4 = MULTIPLY(z4, - FIX_0_390180644); /* sqrt(2) * (c5-c3) */
|
|
|
|
z3 += z5;
|
|
z4 += z5;
|
|
|
|
tmp0 += z1 + z3;
|
|
tmp1 += z2 + z4;
|
|
tmp2 += z2 + z3;
|
|
tmp3 += z1 + z4;
|
|
|
|
/* Final output stage: inputs are tmp10..tmp13, tmp0..tmp3 */
|
|
|
|
dataptr[0] = (DCTELEM) DESCALE(tmp10 + tmp3, CONST_BITS-PASS1_BITS);
|
|
dataptr[7] = (DCTELEM) DESCALE(tmp10 - tmp3, CONST_BITS-PASS1_BITS);
|
|
dataptr[1] = (DCTELEM) DESCALE(tmp11 + tmp2, CONST_BITS-PASS1_BITS);
|
|
dataptr[6] = (DCTELEM) DESCALE(tmp11 - tmp2, CONST_BITS-PASS1_BITS);
|
|
dataptr[2] = (DCTELEM) DESCALE(tmp12 + tmp1, CONST_BITS-PASS1_BITS);
|
|
dataptr[5] = (DCTELEM) DESCALE(tmp12 - tmp1, CONST_BITS-PASS1_BITS);
|
|
dataptr[3] = (DCTELEM) DESCALE(tmp13 + tmp0, CONST_BITS-PASS1_BITS);
|
|
dataptr[4] = (DCTELEM) DESCALE(tmp13 - tmp0, CONST_BITS-PASS1_BITS);
|
|
|
|
dataptr += DCTSIZE; /* advance pointer to next row */
|
|
}
|
|
|
|
/* Pass 2: process columns. */
|
|
/* Note that we must descale the results by a factor of 8 == 2**3, */
|
|
/* and also undo the PASS1_BITS scaling. */
|
|
|
|
dataptr = data;
|
|
for (rowctr = DCTSIZE-1; rowctr >= 0; rowctr--) {
|
|
/* Even part: reverse the even part of the forward DCT. */
|
|
/* The rotator is sqrt(2)*c(-6). */
|
|
|
|
z2 = (int32) dataptr[DCTSIZE*2];
|
|
z3 = (int32) dataptr[DCTSIZE*6];
|
|
|
|
z1 = MULTIPLY(z2 + z3, FIX_0_541196100);
|
|
tmp2 = z1 + MULTIPLY(z3, - FIX_1_847759065);
|
|
tmp3 = z1 + MULTIPLY(z2, FIX_0_765366865);
|
|
|
|
tmp0 = ((uint32) dataptr[DCTSIZE*0] + (uint32) dataptr[DCTSIZE*4]) << CONST_BITS;
|
|
tmp1 = ((uint32) dataptr[DCTSIZE*0] - (uint32) dataptr[DCTSIZE*4]) << CONST_BITS;
|
|
|
|
tmp10 = tmp0 + tmp3;
|
|
tmp13 = tmp0 - tmp3;
|
|
tmp11 = tmp1 + tmp2;
|
|
tmp12 = tmp1 - tmp2;
|
|
|
|
/* Odd part per figure 8; the matrix is unitary and hence its
|
|
* transpose is its inverse. i0..i3 are y7,y5,y3,y1 respectively.
|
|
*/
|
|
|
|
tmp0 = (int32) dataptr[DCTSIZE*7];
|
|
tmp1 = (int32) dataptr[DCTSIZE*5];
|
|
tmp2 = (int32) dataptr[DCTSIZE*3];
|
|
tmp3 = (int32) dataptr[DCTSIZE*1];
|
|
|
|
z1 = tmp0 + tmp3;
|
|
z2 = tmp1 + tmp2;
|
|
z3 = tmp0 + tmp2;
|
|
z4 = tmp1 + tmp3;
|
|
z5 = MULTIPLY(z3 + z4, FIX_1_175875602); /* sqrt(2) * c3 */
|
|
|
|
tmp0 = MULTIPLY(tmp0, FIX_0_298631336); /* sqrt(2) * (-c1+c3+c5-c7) */
|
|
tmp1 = MULTIPLY(tmp1, FIX_2_053119869); /* sqrt(2) * ( c1+c3-c5+c7) */
|
|
tmp2 = MULTIPLY(tmp2, FIX_3_072711026); /* sqrt(2) * ( c1+c3+c5-c7) */
|
|
tmp3 = MULTIPLY(tmp3, FIX_1_501321110); /* sqrt(2) * ( c1+c3-c5-c7) */
|
|
z1 = MULTIPLY(z1, - FIX_0_899976223); /* sqrt(2) * (c7-c3) */
|
|
z2 = MULTIPLY(z2, - FIX_2_562915447); /* sqrt(2) * (-c1-c3) */
|
|
z3 = MULTIPLY(z3, - FIX_1_961570560); /* sqrt(2) * (-c3-c5) */
|
|
z4 = MULTIPLY(z4, - FIX_0_390180644); /* sqrt(2) * (c5-c3) */
|
|
|
|
z3 += z5;
|
|
z4 += z5;
|
|
|
|
tmp0 += z1 + z3;
|
|
tmp1 += z2 + z4;
|
|
tmp2 += z2 + z3;
|
|
tmp3 += z1 + z4;
|
|
|
|
/* Final output stage: inputs are tmp10..tmp13, tmp0..tmp3 */
|
|
|
|
dataptr[DCTSIZE*0] = (DCTELEM) DESCALE(tmp10 + tmp3, CONST_BITS+PASS1_BITS+1);
|
|
dataptr[DCTSIZE*7] = (DCTELEM) DESCALE(tmp10 - tmp3, CONST_BITS+PASS1_BITS+1);
|
|
dataptr[DCTSIZE*1] = (DCTELEM) DESCALE(tmp11 + tmp2, CONST_BITS+PASS1_BITS+1);
|
|
dataptr[DCTSIZE*6] = (DCTELEM) DESCALE(tmp11 - tmp2, CONST_BITS+PASS1_BITS+1);
|
|
dataptr[DCTSIZE*2] = (DCTELEM) DESCALE(tmp12 + tmp1, CONST_BITS+PASS1_BITS+1);
|
|
dataptr[DCTSIZE*5] = (DCTELEM) DESCALE(tmp12 - tmp1, CONST_BITS+PASS1_BITS+1);
|
|
dataptr[DCTSIZE*3] = (DCTELEM) DESCALE(tmp13 + tmp0, CONST_BITS+PASS1_BITS+1);
|
|
dataptr[DCTSIZE*4] = (DCTELEM) DESCALE(tmp13 - tmp0, CONST_BITS+PASS1_BITS+1);
|
|
|
|
dataptr++; /* advance pointer to next column */
|
|
}
|
|
}
|
|
|
|
}
|