jcdctmgr.cpp File Reference

#include "jinclude.h"
#include "jpeglib.h"
#include "jdct.h"

Include dependency graph for jcdctmgr.cpp:

Classes
struct	my_fdct_controller

Macros
#define	JPEG_INTERNALS
#define	CONST_BITS   14
#define	DIVIDE_BY(a, b)   if (a >= b) a /= b; else a = 0

Typedefs
typedef my_fdct_controller *	my_fdct_ptr

Functions
static void	start_pass_fdctmgr (j_compress_ptr cinfo)
static void	forward_DCT (j_compress_ptr cinfo, jpeg_component_info *compptr, JSAMPARRAY sample_data, JBLOCKROW coef_blocks, JDIMENSION start_row, JDIMENSION start_col, JDIMENSION num_blocks)
static void	forward_DCT_float (j_compress_ptr cinfo, jpeg_component_info *compptr, JSAMPARRAY sample_data, JBLOCKROW coef_blocks, JDIMENSION start_row, JDIMENSION start_col, JDIMENSION num_blocks)
void	jinit_forward_dct (j_compress_ptr cinfo)

Macro Definition Documentation

#define JPEG_INTERNALS

#define CONST_BITS   14

#define DIVIDE_BY	(		a,
			b
	)		if (a >= b) a /= b; else a = 0

Typedef Documentation

typedef my_fdct_controller* my_fdct_ptr

Function Documentation

static void start_pass_fdctmgr ( j_compress_ptr  cinfo )

static

{
  my_fdct_ptr fdct = (my_fdct_ptr) cinfo->fdct;
  int ci, qtblno, i;
  jpeg_component_info *compptr;
  JQUANT_TBL * qtbl;
  DCTELEM * dtbl;

  for (ci = 0, compptr = cinfo->comp_info; ci < cinfo->num_components;
       ci++, compptr++) {
    qtblno = compptr->quant_tbl_no;
    /* Make sure specified quantization table is present */
    if (qtblno < 0 || qtblno >= NUM_QUANT_TBLS ||
    cinfo->quant_tbl_ptrs[qtblno] == NULL)
      ERREXIT1(cinfo, JERR_NO_QUANT_TABLE, qtblno);
    qtbl = cinfo->quant_tbl_ptrs[qtblno];
    /* Compute divisors for this quant table */
    /* We may do this more than once for same table, but it's not a big deal */
    switch (cinfo->dct_method) {
#ifdef DCT_ISLOW_SUPPORTED
    case JDCT_ISLOW:
      /* For LL&M IDCT method, divisors are equal to raw quantization
       * coefficients multiplied by 8 (to counteract scaling).
       */
      if (fdct->divisors[qtblno] == NULL) {
    fdct->divisors[qtblno] = (DCTELEM *)
      (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
                      DCTSIZE2 * SIZEOF(DCTELEM));
      }
      dtbl = fdct->divisors[qtblno];
      for (i = 0; i < DCTSIZE2; i++) {
    dtbl[i] = ((DCTELEM) qtbl->quantval[i]) << 3;
      }
      break;
#endif
#ifdef DCT_IFAST_SUPPORTED
    case JDCT_IFAST:
      {
    /* For AA&N IDCT method, divisors are equal to quantization
     * coefficients scaled by scalefactor[row]*scalefactor[col], where
     *   scalefactor[0] = 1
     *   scalefactor[k] = cos(k*PI/16) * sqrt(2)    for k=1..7
     * We apply a further scale factor of 8.
     */
#define CONST_BITS 14
    static const int16_t aanscales[DCTSIZE2] = {
      /* precomputed values scaled up by 14 bits */
      16384, 22725, 21407, 19266, 16384, 12873,  8867,  4520,
      22725, 31521, 29692, 26722, 22725, 17855, 12299,  6270,
      21407, 29692, 27969, 25172, 21407, 16819, 11585,  5906,
      19266, 26722, 25172, 22654, 19266, 15137, 10426,  5315,
      16384, 22725, 21407, 19266, 16384, 12873,  8867,  4520,
      12873, 17855, 16819, 15137, 12873, 10114,  6967,  3552,
       8867, 12299, 11585, 10426,  8867,  6967,  4799,  2446,
       4520,  6270,  5906,  5315,  4520,  3552,  2446,  1247
    };
    SHIFT_TEMPS

    if (fdct->divisors[qtblno] == NULL) {
      fdct->divisors[qtblno] = (DCTELEM *)
        (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
                    DCTSIZE2 * SIZEOF(DCTELEM));
    }
    dtbl = fdct->divisors[qtblno];
    for (i = 0; i < DCTSIZE2; i++) {
      dtbl[i] = (DCTELEM)
        DESCALE(MULTIPLY16V16((int32_t) qtbl->quantval[i],
                  (int32_t) aanscales[i]),
            CONST_BITS-3);
    }
      }
      break;
#endif
#ifdef DCT_FLOAT_SUPPORTED
    case JDCT_FLOAT:
      {
    /* For float AA&N IDCT method, divisors are equal to quantization
     * coefficients scaled by scalefactor[row]*scalefactor[col], where
     *   scalefactor[0] = 1
     *   scalefactor[k] = cos(k*PI/16) * sqrt(2)    for k=1..7
     * We apply a further scale factor of 8.
     * What's actually stored is 1/divisor so that the inner loop can
     * use a multiplication rather than a division.
     */
    FAST_FLOAT * fdtbl;
    int row, col;
    static const double aanscalefactor[DCTSIZE] = {
      1.0, 1.387039845, 1.306562965, 1.175875602,
      1.0, 0.785694958, 0.541196100, 0.275899379
    };

    if (fdct->float_divisors[qtblno] == NULL) {
      fdct->float_divisors[qtblno] = (FAST_FLOAT *)
        (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
                    DCTSIZE2 * SIZEOF(FAST_FLOAT));
    }
    fdtbl = fdct->float_divisors[qtblno];
    i = 0;
    for (row = 0; row < DCTSIZE; row++) {
      for (col = 0; col < DCTSIZE; col++) {
        fdtbl[i] = (FAST_FLOAT)
          (1.0 / (((double) qtbl->quantval[i] *
               aanscalefactor[row] * aanscalefactor[col] * 8.0)));
        i++;
      }
    }
      }
      break;
#endif
    default:
      ERREXIT(cinfo, JERR_NOT_COMPILED);
      break;
    }
  }
}

static void forward_DCT ( j_compress_ptr  cinfo, jpeg_component_info *  compptr, JSAMPARRAY  sample_data, JBLOCKROW  coef_blocks, JDIMENSION  start_row, JDIMENSION  start_col, JDIMENSION  num_blocks  )

static

{
  /* This routine is heavily used, so it's worth coding it tightly. */
  my_fdct_ptr fdct = (my_fdct_ptr) cinfo->fdct;
  forward_DCT_method_ptr do_dct = fdct->do_dct;
  DCTELEM * divisors = fdct->divisors[compptr->quant_tbl_no];
  DCTELEM workspace[DCTSIZE2];  /* work area for FDCT subroutine */
  JDIMENSION bi;

  sample_data += start_row; /* fold in the vertical offset once */

  for (bi = 0; bi < num_blocks; bi++, start_col += DCTSIZE) {
    /* Load data into workspace, applying unsigned->signed conversion */
    { register DCTELEM *workspaceptr;
      register JSAMPROW elemptr;
      register int elemr;

      workspaceptr = workspace;
      for (elemr = 0; elemr < DCTSIZE; elemr++) {
    elemptr = sample_data[elemr] + start_col;
#if DCTSIZE == 8        /* unroll the inner loop */
    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
    *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
#else
    { register int elemc;
      for (elemc = DCTSIZE; elemc > 0; elemc--) {
        *workspaceptr++ = GETJSAMPLE(*elemptr++) - CENTERJSAMPLE;
      }
    }
#endif
      }
    }

    /* Perform the DCT */
    (*do_dct) (workspace);

    /* Quantize/descale the coefficients, and store into coef_blocks[] */
    { register DCTELEM temp, qval;
      register int i;
      register JCOEFPTR output_ptr = coef_blocks[bi];

      for (i = 0; i < DCTSIZE2; i++) {
    qval = divisors[i];
    temp = workspace[i];
    /* Divide the coefficient value by qval, ensuring proper rounding.
     * Since C does not specify the direction of rounding for negative
     * quotients, we have to force the dividend positive for portability.
     *
     * In most files, at least half of the output values will be zero
     * (at default quantization settings, more like three-quarters...)
     * so we should ensure that this case is fast.  On many machines,
     * a comparison is enough cheaper than a divide to make a special test
     * a win.  Since both inputs will be nonnegative, we need only test
     * for a < b to discover whether a/b is 0.
     * If your machine's division is fast enough, define FAST_DIVIDE.
     */
#ifdef FAST_DIVIDE
#define DIVIDE_BY(a,b)  a /= b
#else
#define DIVIDE_BY(a,b)  if (a >= b) a /= b; else a = 0
#endif
    if (temp < 0) {
      temp = -temp;
      temp += qval>>1;  /* for rounding */
      DIVIDE_BY(temp, qval);
      temp = -temp;
    } else {
      temp += qval>>1;  /* for rounding */
      DIVIDE_BY(temp, qval);
    }
    output_ptr[i] = (JCOEF) temp;
      }
    }
  }
}

static void forward_DCT_float ( j_compress_ptr  cinfo, jpeg_component_info *  compptr, JSAMPARRAY  sample_data, JBLOCKROW  coef_blocks, JDIMENSION  start_row, JDIMENSION  start_col, JDIMENSION  num_blocks  )

static

{
  /* This routine is heavily used, so it's worth coding it tightly. */
  my_fdct_ptr fdct = (my_fdct_ptr) cinfo->fdct;
  float_DCT_method_ptr do_dct = fdct->do_float_dct;
  FAST_FLOAT * divisors = fdct->float_divisors[compptr->quant_tbl_no];
  FAST_FLOAT workspace[DCTSIZE2]; /* work area for FDCT subroutine */
  JDIMENSION bi;

  sample_data += start_row; /* fold in the vertical offset once */

  for (bi = 0; bi < num_blocks; bi++, start_col += DCTSIZE) {
    /* Load data into workspace, applying unsigned->signed conversion */
    { register FAST_FLOAT *workspaceptr;
      register JSAMPROW elemptr;
      register int elemr;

      workspaceptr = workspace;
      for (elemr = 0; elemr < DCTSIZE; elemr++) {
    elemptr = sample_data[elemr] + start_col;
#if DCTSIZE == 8        /* unroll the inner loop */
    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
    *workspaceptr++ = (FAST_FLOAT)(GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
#else
    { register int elemc;
      for (elemc = DCTSIZE; elemc > 0; elemc--) {
        *workspaceptr++ = (FAST_FLOAT)
          (GETJSAMPLE(*elemptr++) - CENTERJSAMPLE);
      }
    }
#endif
      }
    }

    /* Perform the DCT */
    (*do_dct) (workspace);

    /* Quantize/descale the coefficients, and store into coef_blocks[] */
    { register FAST_FLOAT temp;
      register int i;
      register JCOEFPTR output_ptr = coef_blocks[bi];

      for (i = 0; i < DCTSIZE2; i++) {
    /* Apply the quantization and scaling factor */
    temp = workspace[i] * divisors[i];
    /* Round to nearest integer.
     * Since C does not specify the direction of rounding for negative
     * quotients, we have to force the dividend positive for portability.
     * The maximum coefficient size is +-16K (for 12-bit data), so this
     * code should work for either 16-bit or 32-bit ints.
     */
    output_ptr[i] = (JCOEF) ((int) (temp + (FAST_FLOAT) 16384.5) - 16384);
      }
    }
  }
}

void jinit_forward_dct

(

j_compress_ptr

cinfo

)

{
  my_fdct_ptr fdct;
  int i;

  fdct = (my_fdct_ptr)
    (*cinfo->mem->alloc_small) ((j_common_ptr) cinfo, JPOOL_IMAGE,
                SIZEOF(my_fdct_controller));
  cinfo->fdct = (struct jpeg_forward_dct *) fdct;
  fdct->pub.start_pass = start_pass_fdctmgr;

  switch (cinfo->dct_method) {
#ifdef DCT_ISLOW_SUPPORTED
  case JDCT_ISLOW:
    fdct->pub.forward_DCT = forward_DCT;
    fdct->do_dct = jpeg_fdct_islow;
    break;
#endif
#ifdef DCT_IFAST_SUPPORTED
  case JDCT_IFAST:
    fdct->pub.forward_DCT = forward_DCT;
    fdct->do_dct = jpeg_fdct_ifast;
    break;
#endif
#ifdef DCT_FLOAT_SUPPORTED
  case JDCT_FLOAT:
    fdct->pub.forward_DCT = forward_DCT_float;
    fdct->do_float_dct = jpeg_fdct_float;
    break;
#endif
  default:
    ERREXIT(cinfo, JERR_NOT_COMPILED);
    break;
  }

  /* Mark divisor tables unallocated */
  for (i = 0; i < NUM_QUANT_TBLS; i++) {
    fdct->divisors[i] = NULL;
#ifdef DCT_FLOAT_SUPPORTED
    fdct->float_divisors[i] = NULL;
#endif
  }
}