/*
 *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
 *
 *  Use of this source code is governed by a BSD-style license
 *  that can be found in the LICENSE file in the root of the source
 *  tree. An additional intellectual property rights grant can be found
 *  in the file PATENTS.  All contributing project authors may
 *  be found in the AUTHORS file in the root of the source tree.
 */


#include "vpx_config.h"
#include "vp9/common/onyxc_int.h"
#include "onyx_int.h"
#include "vp9/common/systemdependent.h"
#include "quantize.h"
#include "vp9/common/alloccommon.h"
#include "mcomp.h"
#include "firstpass.h"
#include "psnr.h"
#include "vpx_scale/vpxscale.h"
#include "vp9/common/extend.h"
#include "ratectrl.h"
#include "vp9/common/quant_common.h"
#include "segmentation.h"
#if CONFIG_POSTPROC
#include "vp9/common/postproc.h"
#endif
#include "vpx_mem/vpx_mem.h"
#include "vp9/common/swapyv12buffer.h"
#include "vpx_ports/vpx_timer.h"
#include "temporal_filter.h"

#include "vp9/common/seg_common.h"
#include "mbgraph.h"
#include "vp9/common/pred_common.h"
#include "vp9/encoder/rdopt.h"
#include "bitstream.h"
#include "ratectrl.h"

#if CONFIG_NEWBESTREFMV
#include "vp9/common/mvref_common.h"
#endif

#if ARCH_ARM
#include "vpx_ports/arm.h"
#endif

#include <math.h>
#include <stdio.h>
#include <limits.h>

#if CONFIG_RUNTIME_CPU_DETECT
#define IF_RTCD(x) (x)
#define RTCD(x) &cpi->common.rtcd.x
#else
#define IF_RTCD(x) NULL
#define RTCD(x) NULL
#endif

extern void vp9_pick_filter_level_fast(YV12_BUFFER_CONFIG *sd, VP9_COMP *cpi);

extern void vp9_set_alt_lf_level(VP9_COMP *cpi, int filt_val);

extern void vp9_pick_filter_level(YV12_BUFFER_CONFIG *sd, VP9_COMP *cpi);

extern void vp9_cmachine_specific_config(VP9_COMP *cpi);

extern void vp9_deblock_frame(YV12_BUFFER_CONFIG *source,
                              YV12_BUFFER_CONFIG *post,
                              int filt_lvl, int low_var_thresh, int flag);

extern void print_tree_update_probs();

#if HAVE_ARMV7
extern void vp8_yv12_copy_frame_func_neon(YV12_BUFFER_CONFIG *src_ybc,
                                          YV12_BUFFER_CONFIG *dst_ybc);

extern void vp8_yv12_copy_src_frame_func_neon(YV12_BUFFER_CONFIG *src_ybc,
                                              YV12_BUFFER_CONFIG *dst_ybc);
#endif

int vp9_calc_ss_err(YV12_BUFFER_CONFIG *source, YV12_BUFFER_CONFIG *dest);

extern void vp9_temporal_filter_prepare_c(VP9_COMP *cpi, int distance);

static void set_default_lf_deltas(VP9_COMP *cpi);

#define DEFAULT_INTERP_FILTER EIGHTTAP  /* SWITCHABLE for better performance */
#define SEARCH_BEST_FILTER 0            /* to search exhaustively for
                                           best filter */
#define RESET_FOREACH_FILTER 0          /* whether to reset the encoder state
                                           before trying each new filter */
#define SHARP_FILTER_QTHRESH 0          /* Q threshold for 8-tap sharp filter */

#define ALTREF_HIGH_PRECISION_MV 1      /* whether to use high precision mv
                                           for altref computation */
#define HIGH_PRECISION_MV_QTHRESH 200   /* Q threshold for use of high precision
                                           mv. Choose a very high value for
                                           now so that HIGH_PRECISION is always
                                           chosen */

#if CONFIG_INTERNAL_STATS
#include "math.h"

extern double vp9_calc_ssim(YV12_BUFFER_CONFIG *source,
                            YV12_BUFFER_CONFIG *dest, int lumamask,
                            double *weight);


extern double vp9_calc_ssimg(YV12_BUFFER_CONFIG *source,
                             YV12_BUFFER_CONFIG *dest, double *ssim_y,
                             double *ssim_u, double *ssim_v);


#endif

// #define OUTPUT_YUV_REC

#ifdef OUTPUT_YUV_SRC
FILE *yuv_file;
#endif
#ifdef OUTPUT_YUV_REC
FILE *yuv_rec_file;
#endif

#if 0
FILE *framepsnr;
FILE *kf_list;
FILE *keyfile;
#endif

#if 0
extern int skip_true_count;
extern int skip_false_count;
#endif


#ifdef ENTROPY_STATS
extern int intra_mode_stats[VP9_BINTRAMODES][VP9_BINTRAMODES][VP9_BINTRAMODES];
#endif

#ifdef NMV_STATS
extern void init_nmvstats();
extern void print_nmvstats();
#endif

#ifdef SPEEDSTATS
unsigned int frames_at_speed[16] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
#endif

#if defined(SECTIONBITS_OUTPUT)
extern unsigned __int64 Sectionbits[500];
#endif
#ifdef MODE_STATS
extern INT64 Sectionbits[500];
extern unsigned int y_modes[VP9_YMODES];
extern unsigned int i8x8_modes[VP9_I8X8_MODES];
extern unsigned int uv_modes[VP9_UV_MODES];
extern unsigned int uv_modes_y[VP9_YMODES][VP9_UV_MODES];
extern unsigned int b_modes[B_MODE_COUNT];
extern unsigned int inter_y_modes[MB_MODE_COUNT];
extern unsigned int inter_uv_modes[VP9_UV_MODES];
extern unsigned int inter_b_modes[B_MODE_COUNT];
#endif

extern void vp9_init_quantizer(VP9_COMP *cpi);

static int base_skip_false_prob[QINDEX_RANGE][3];

// Tables relating active max Q to active min Q
static int kf_low_motion_minq[QINDEX_RANGE];
static int kf_high_motion_minq[QINDEX_RANGE];
static int gf_low_motion_minq[QINDEX_RANGE];
static int gf_high_motion_minq[QINDEX_RANGE];
static int inter_minq[QINDEX_RANGE];

// Functions to compute the active minq lookup table entries based on a
// formulaic approach to facilitate easier adjustment of the Q tables.
// The formulae were derived from computing a 3rd order polynomial best
// fit to the original data (after plotting real maxq vs minq (not q index))
static int calculate_minq_index(double maxq,
                                double x3, double x2, double x, double c) {
  int i;
  double minqtarget;
  double thisq;

  minqtarget = ((x3 * maxq * maxq * maxq) +
                (x2 * maxq * maxq) +
                (x * maxq) +
                c);

  if (minqtarget > maxq)
    minqtarget = maxq;

  for (i = 0; i < QINDEX_RANGE; i++) {
    thisq = vp9_convert_qindex_to_q(i);
    if (minqtarget <= vp9_convert_qindex_to_q(i))
      return i;
  }
  return QINDEX_RANGE - 1;
}

static void init_minq_luts(void) {
  int i;
  double maxq;

  for (i = 0; i < QINDEX_RANGE; i++) {
    maxq = vp9_convert_qindex_to_q(i);


    kf_low_motion_minq[i] = calculate_minq_index(maxq,
                                                 0.0000003,
                                                 -0.000015,
                                                 0.074,
                                                 0.0);
    kf_high_motion_minq[i] = calculate_minq_index(maxq,
                                                  0.0000004,
                                                  -0.000125,
                                                  0.14,
                                                  0.0);
    gf_low_motion_minq[i] = calculate_minq_index(maxq,
                                                 0.0000015,
                                                 -0.0009,
                                                 0.33,
                                                 0.0);
    gf_high_motion_minq[i] = calculate_minq_index(maxq,
                                                  0.0000021,
                                                  -0.00125,
                                                  0.45,
                                                  0.0);
    inter_minq[i] = calculate_minq_index(maxq,
                                         0.00000271,
                                         -0.00113,
                                         0.697,
                                         0.0);

  }
}

static void init_base_skip_probs(void) {
  int i;
  double q;
  int skip_prob, t;

  for (i = 0; i < QINDEX_RANGE; i++) {
    q = vp9_convert_qindex_to_q(i);

    // Exponential decay caluclation of baseline skip prob with clamping
    // Based on crude best fit of old table.
    t = (int)(564.25 * pow(2.71828, (-0.012 * q)));

    skip_prob = t;
    if (skip_prob < 1)
      skip_prob = 1;
    else if (skip_prob > 255)
      skip_prob = 255;
    base_skip_false_prob[i][1] = skip_prob;

    skip_prob = t * 0.75;
    if (skip_prob < 1)
      skip_prob = 1;
    else if (skip_prob > 255)
      skip_prob = 255;
    base_skip_false_prob[i][2] = skip_prob;

    skip_prob = t * 1.25;
    if (skip_prob < 1)
      skip_prob = 1;
    else if (skip_prob > 255)
      skip_prob = 255;
    base_skip_false_prob[i][0] = skip_prob;
  }
}

static void update_base_skip_probs(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;

  if (cm->frame_type != KEY_FRAME) {
    vp9_update_skip_probs(cpi);

    if (cm->refresh_alt_ref_frame) {
      int k;
      for (k = 0; k < MBSKIP_CONTEXTS; ++k)
        cpi->last_skip_false_probs[2][k] = cm->mbskip_pred_probs[k];
      cpi->last_skip_probs_q[2] = cm->base_qindex;
    } else if (cpi->common.refresh_golden_frame) {
      int k;
      for (k = 0; k < MBSKIP_CONTEXTS; ++k)
        cpi->last_skip_false_probs[1][k] = cm->mbskip_pred_probs[k];
      cpi->last_skip_probs_q[1] = cm->base_qindex;
    } else {
      int k;
      for (k = 0; k < MBSKIP_CONTEXTS; ++k)
        cpi->last_skip_false_probs[0][k] = cm->mbskip_pred_probs[k];
      cpi->last_skip_probs_q[0] = cm->base_qindex;

      // update the baseline table for the current q
      for (k = 0; k < MBSKIP_CONTEXTS; ++k)
        cpi->base_skip_false_prob[cm->base_qindex][k] =
          cm->mbskip_pred_probs[k];
    }
  }

}

void vp9_initialize_enc() {
  static int init_done = 0;

  if (!init_done) {
    vp9_initialize_common();
    vp9_tokenize_initialize();
    vp9_init_quant_tables();
    vp9_init_me_luts();
    init_minq_luts();
    init_base_skip_probs();
    init_done = 1;
  }
}
#ifdef PACKET_TESTING
extern FILE *vpxlogc;
#endif

static void setup_features(VP9_COMP *cpi) {
  MACROBLOCKD *xd = &cpi->mb.e_mbd;

  // Set up default state for MB feature flags

  xd->segmentation_enabled = 0;   // Default segmentation disabled

  xd->update_mb_segmentation_map = 0;
  xd->update_mb_segmentation_data = 0;
  vpx_memset(xd->mb_segment_tree_probs, 255, sizeof(xd->mb_segment_tree_probs));

  vp9_clearall_segfeatures(xd);

  xd->mode_ref_lf_delta_enabled = 0;
  xd->mode_ref_lf_delta_update = 0;
  vpx_memset(xd->ref_lf_deltas, 0, sizeof(xd->ref_lf_deltas));
  vpx_memset(xd->mode_lf_deltas, 0, sizeof(xd->mode_lf_deltas));
  vpx_memset(xd->last_ref_lf_deltas, 0, sizeof(xd->ref_lf_deltas));
  vpx_memset(xd->last_mode_lf_deltas, 0, sizeof(xd->mode_lf_deltas));

  set_default_lf_deltas(cpi);

}


static void dealloc_compressor_data(VP9_COMP *cpi) {
  vpx_free(cpi->tplist);
  cpi->tplist = NULL;

  // Delete last frame MV storage buffers
  vpx_free(cpi->lfmv);
  cpi->lfmv = 0;

  vpx_free(cpi->lf_ref_frame_sign_bias);
  cpi->lf_ref_frame_sign_bias = 0;

  vpx_free(cpi->lf_ref_frame);
  cpi->lf_ref_frame = 0;

  // Delete sementation map
  vpx_free(cpi->segmentation_map);
  cpi->segmentation_map = 0;
  vpx_free(cpi->common.last_frame_seg_map);
  cpi->common.last_frame_seg_map = 0;
  vpx_free(cpi->coding_context.last_frame_seg_map_copy);
  cpi->coding_context.last_frame_seg_map_copy = 0;

  vpx_free(cpi->active_map);
  cpi->active_map = 0;

  vp9_de_alloc_frame_buffers(&cpi->common);

  vp8_yv12_de_alloc_frame_buffer(&cpi->last_frame_uf);
  vp8_yv12_de_alloc_frame_buffer(&cpi->scaled_source);
#if VP9_TEMPORAL_ALT_REF
  vp8_yv12_de_alloc_frame_buffer(&cpi->alt_ref_buffer);
#endif
  vp9_lookahead_destroy(cpi->lookahead);

  vpx_free(cpi->tok);
  cpi->tok = 0;

  // Structure used to monitor GF usage
  vpx_free(cpi->gf_active_flags);
  cpi->gf_active_flags = 0;

  // Activity mask based per mb zbin adjustments
  vpx_free(cpi->mb_activity_map);
  cpi->mb_activity_map = 0;
  vpx_free(cpi->mb_norm_activity_map);
  cpi->mb_norm_activity_map = 0;

  vpx_free(cpi->mb.pip);
  cpi->mb.pip = 0;

  vpx_free(cpi->twopass.total_stats);
  cpi->twopass.total_stats = 0;

  vpx_free(cpi->twopass.total_left_stats);
  cpi->twopass.total_left_stats = 0;

  vpx_free(cpi->twopass.this_frame_stats);
  cpi->twopass.this_frame_stats = 0;
}

// Computes a q delta (in "q index" terms) to get from a starting q value
// to a target value
// target q value
static int compute_qdelta(VP9_COMP *cpi, double qstart, double qtarget) {
  int i;
  int start_index = cpi->worst_quality;
  int target_index = cpi->worst_quality;

  // Convert the average q value to an index.
  for (i = cpi->best_quality; i < cpi->worst_quality; i++) {
    start_index = i;
    if (vp9_convert_qindex_to_q(i) >= qstart)
      break;
  }

  // Convert the q target to an index
  for (i = cpi->best_quality; i < cpi->worst_quality; i++) {
    target_index = i;
    if (vp9_convert_qindex_to_q(i) >= qtarget)
      break;
  }

  return target_index - start_index;
}

static void init_seg_features(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;
  MACROBLOCKD *xd = &cpi->mb.e_mbd;

  int high_q = (int)(cpi->avg_q > 48.0);
  int qi_delta;

  // Disable and clear down for KF
  if (cm->frame_type == KEY_FRAME) {
    // Clear down the global segmentation map
    vpx_memset(cpi->segmentation_map, 0, (cm->mb_rows * cm->mb_cols));
    xd->update_mb_segmentation_map = 0;
    xd->update_mb_segmentation_data = 0;
    cpi->static_mb_pct = 0;

    // Disable segmentation
    vp9_disable_segmentation((VP9_PTR)cpi);

    // Clear down the segment features.
    vp9_clearall_segfeatures(xd);
  }

  // If this is an alt ref frame
  else if (cm->refresh_alt_ref_frame) {
    // Clear down the global segmentation map
    vpx_memset(cpi->segmentation_map, 0, (cm->mb_rows * cm->mb_cols));
    xd->update_mb_segmentation_map = 0;
    xd->update_mb_segmentation_data = 0;
    cpi->static_mb_pct = 0;

    // Disable segmentation and individual segment features by default
    vp9_disable_segmentation((VP9_PTR)cpi);
    vp9_clearall_segfeatures(xd);

    // Scan frames from current to arf frame.
    // This function re-enables segmentation if appropriate.
    vp9_update_mbgraph_stats(cpi);

    // If segmentation was enabled set those features needed for the
    // arf itself.
    if (xd->segmentation_enabled) {
      xd->update_mb_segmentation_map = 1;
      xd->update_mb_segmentation_data = 1;

      qi_delta = compute_qdelta(cpi, cpi->avg_q, (cpi->avg_q * 0.875));
      vp9_set_segdata(xd, 1, SEG_LVL_ALT_Q, (qi_delta - 2));
      vp9_set_segdata(xd, 1, SEG_LVL_ALT_LF, -2);

      vp9_enable_segfeature(xd, 1, SEG_LVL_ALT_Q);
      vp9_enable_segfeature(xd, 1, SEG_LVL_ALT_LF);

      // Where relevant assume segment data is delta data
      xd->mb_segment_abs_delta = SEGMENT_DELTADATA;

    }
  }
  // All other frames if segmentation has been enabled
  else if (xd->segmentation_enabled) {
    // First normal frame in a valid gf or alt ref group
    if (cpi->common.frames_since_golden == 0) {
      // Set up segment features for normal frames in an af group
      if (cpi->source_alt_ref_active) {
        xd->update_mb_segmentation_map = 0;
        xd->update_mb_segmentation_data = 1;
        xd->mb_segment_abs_delta = SEGMENT_DELTADATA;

        qi_delta = compute_qdelta(cpi, cpi->avg_q,
                                  (cpi->avg_q * 1.125));
        vp9_set_segdata(xd, 1, SEG_LVL_ALT_Q, (qi_delta + 2));
        vp9_set_segdata(xd, 1, SEG_LVL_ALT_Q, 0);
        vp9_enable_segfeature(xd, 1, SEG_LVL_ALT_Q);

        vp9_set_segdata(xd, 1, SEG_LVL_ALT_LF, -2);
        vp9_enable_segfeature(xd, 1, SEG_LVL_ALT_LF);

        // Segment coding disabled for compred testing
        if (high_q || (cpi->static_mb_pct == 100)) {
          // set_segref(xd, 1, LAST_FRAME);
          vp9_set_segref(xd, 1, ALTREF_FRAME);
          vp9_enable_segfeature(xd, 1, SEG_LVL_REF_FRAME);

          vp9_set_segdata(xd, 1, SEG_LVL_MODE, ZEROMV);
          vp9_enable_segfeature(xd, 1, SEG_LVL_MODE);

          // EOB segment coding not fixed for 8x8 yet
          vp9_set_segdata(xd, 1, SEG_LVL_EOB, 0);
          vp9_enable_segfeature(xd, 1, SEG_LVL_EOB);
        }
      }
      // Disable segmentation and clear down features if alt ref
      // is not active for this group
      else {
        vp9_disable_segmentation((VP9_PTR)cpi);

        vpx_memset(cpi->segmentation_map, 0,
                   (cm->mb_rows * cm->mb_cols));

        xd->update_mb_segmentation_map = 0;
        xd->update_mb_segmentation_data = 0;

        vp9_clearall_segfeatures(xd);
      }
    }

    // Special case where we are coding over the top of a previous
    // alt ref frame
    // Segment coding disabled for compred testing
    else if (cpi->is_src_frame_alt_ref) {
      // Enable mode and ref frame features for segment 0 as well
      vp9_enable_segfeature(xd, 0, SEG_LVL_REF_FRAME);
      vp9_enable_segfeature(xd, 0, SEG_LVL_MODE);
      vp9_enable_segfeature(xd, 1, SEG_LVL_REF_FRAME);
      vp9_enable_segfeature(xd, 1, SEG_LVL_MODE);

      // All mbs should use ALTREF_FRAME, ZEROMV exclusively
      vp9_clear_segref(xd, 0);
      vp9_set_segref(xd, 0, ALTREF_FRAME);
      vp9_clear_segref(xd, 1);
      vp9_set_segref(xd, 1, ALTREF_FRAME);
      vp9_set_segdata(xd, 0, SEG_LVL_MODE, ZEROMV);
      vp9_set_segdata(xd, 1, SEG_LVL_MODE, ZEROMV);

      // Skip all MBs if high Q
      if (high_q) {
        vp9_enable_segfeature(xd, 0, SEG_LVL_EOB);
        vp9_set_segdata(xd, 0, SEG_LVL_EOB, 0);
        vp9_enable_segfeature(xd, 1, SEG_LVL_EOB);
        vp9_set_segdata(xd, 1, SEG_LVL_EOB, 0);
      }
      // Enable data udpate
      xd->update_mb_segmentation_data = 1;
    }
    // All other frames.
    else {
      // No updates.. leave things as they are.
      xd->update_mb_segmentation_map = 0;
      xd->update_mb_segmentation_data = 0;
    }
  }
}

// DEBUG: Print out the segment id of each MB in the current frame.
static void print_seg_map(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;
  int row, col;
  int map_index = 0;
  FILE *statsfile;

  statsfile = fopen("segmap.stt", "a");

  fprintf(statsfile, "%10d\n",
          cm->current_video_frame);

  for (row = 0; row < cpi->common.mb_rows; row++) {
    for (col = 0; col < cpi->common.mb_cols; col++) {
      fprintf(statsfile, "%10d",
              cpi->segmentation_map[map_index]);
      map_index++;
    }
    fprintf(statsfile, "\n");
  }
  fprintf(statsfile, "\n");

  fclose(statsfile);
}

static void update_reference_segmentation_map(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;
  int row, col, sb_rows = (cm->mb_rows + 1) >> 1, sb_cols = (cm->mb_cols + 1) >> 1;
  MODE_INFO *mi = cm->mi;
  uint8_t *segmap = cpi->segmentation_map;
  uint8_t *segcache = cm->last_frame_seg_map;

  for (row = 0; row < sb_rows; row++) {
    for (col = 0; col < sb_cols; col++) {
      MODE_INFO *miptr = mi + col * 2;
      uint8_t *cache = segcache + col * 2;
#if CONFIG_SUPERBLOCKS
      if (miptr->mbmi.encoded_as_sb) {
        cache[0] = miptr->mbmi.segment_id;
        if (!(cm->mb_cols & 1) || col < sb_cols - 1)
          cache[1] = miptr->mbmi.segment_id;
        if (!(cm->mb_rows & 1) || row < sb_rows - 1) {
          cache[cm->mb_cols] = miptr->mbmi.segment_id;
          if (!(cm->mb_cols & 1) || col < sb_cols - 1)
            cache[cm->mb_cols + 1] = miptr->mbmi.segment_id;
        }
      } else
#endif
      {
        cache[0] = miptr[0].mbmi.segment_id;
        if (!(cm->mb_cols & 1) || col < sb_cols - 1)
          cache[1] = miptr[1].mbmi.segment_id;
        if (!(cm->mb_rows & 1) || row < sb_rows - 1) {
          cache[cm->mb_cols] = miptr[cm->mode_info_stride].mbmi.segment_id;
          if (!(cm->mb_cols & 1) || col < sb_cols - 1)
            cache[1] = miptr[1].mbmi.segment_id;
          cache[cm->mb_cols + 1] = miptr[cm->mode_info_stride + 1].mbmi.segment_id;
        }
      }
    }
    segmap += 2 * cm->mb_cols;
    segcache += 2 * cm->mb_cols;
    mi += 2 * cm->mode_info_stride;
  }
}

static void set_default_lf_deltas(VP9_COMP *cpi) {
  cpi->mb.e_mbd.mode_ref_lf_delta_enabled = 1;
  cpi->mb.e_mbd.mode_ref_lf_delta_update = 1;

  vpx_memset(cpi->mb.e_mbd.ref_lf_deltas, 0, sizeof(cpi->mb.e_mbd.ref_lf_deltas));
  vpx_memset(cpi->mb.e_mbd.mode_lf_deltas, 0, sizeof(cpi->mb.e_mbd.mode_lf_deltas));

  // Test of ref frame deltas
  cpi->mb.e_mbd.ref_lf_deltas[INTRA_FRAME] = 2;
  cpi->mb.e_mbd.ref_lf_deltas[LAST_FRAME] = 0;
  cpi->mb.e_mbd.ref_lf_deltas[GOLDEN_FRAME] = -2;
  cpi->mb.e_mbd.ref_lf_deltas[ALTREF_FRAME] = -2;

  cpi->mb.e_mbd.mode_lf_deltas[0] = 4;               // BPRED
  cpi->mb.e_mbd.mode_lf_deltas[1] = -2;              // Zero
  cpi->mb.e_mbd.mode_lf_deltas[2] = 2;               // New mv
  cpi->mb.e_mbd.mode_lf_deltas[3] = 4;               // Split mv
}

void vp9_set_speed_features(VP9_COMP *cpi) {
  SPEED_FEATURES *sf = &cpi->sf;
  int Mode = cpi->compressor_speed;
  int Speed = cpi->Speed;
  int i;
  VP9_COMMON *cm = &cpi->common;

  // Only modes 0 and 1 supported for now in experimental code basae
  if (Mode > 1)
    Mode = 1;

  // Initialise default mode frequency sampling variables
  for (i = 0; i < MAX_MODES; i ++) {
    cpi->mode_check_freq[i] = 0;
    cpi->mode_test_hit_counts[i] = 0;
    cpi->mode_chosen_counts[i] = 0;
  }

  // best quality defaults
  sf->RD = 1;
  sf->search_method = NSTEP;
  sf->improved_dct = 1;
  sf->auto_filter = 1;
  sf->recode_loop = 1;
  sf->quarter_pixel_search = 1;
  sf->half_pixel_search = 1;
  sf->iterative_sub_pixel = 1;
#if CONFIG_LOSSLESS
  sf->optimize_coefficients = 0;
#else
  sf->optimize_coefficients = 1;
#endif
  sf->no_skip_block4x4_search = 1;

  sf->first_step = 0;
  sf->max_step_search_steps = MAX_MVSEARCH_STEPS;
  sf->improved_mv_pred = 1;

  // default thresholds to 0
  for (i = 0; i < MAX_MODES; i++)
    sf->thresh_mult[i] = 0;

  switch (Mode) {
    case 0: // best quality mode
#if CONFIG_PRED_FILTER
      sf->thresh_mult[THR_ZEROMV        ] = 0;
      sf->thresh_mult[THR_ZEROMV_FILT   ] = 0;
      sf->thresh_mult[THR_ZEROG         ] = 0;
      sf->thresh_mult[THR_ZEROG_FILT    ] = 0;
      sf->thresh_mult[THR_ZEROA         ] = 0;
      sf->thresh_mult[THR_ZEROA_FILT    ] = 0;
      sf->thresh_mult[THR_NEARESTMV     ] = 0;
      sf->thresh_mult[THR_NEARESTMV_FILT] = 0;
      sf->thresh_mult[THR_NEARESTG      ] = 0;
      sf->thresh_mult[THR_NEARESTG_FILT ] = 0;
      sf->thresh_mult[THR_NEARESTA      ] = 0;
      sf->thresh_mult[THR_NEARESTA_FILT ] = 0;
      sf->thresh_mult[THR_NEARMV        ] = 0;
      sf->thresh_mult[THR_NEARMV_FILT   ] = 0;
      sf->thresh_mult[THR_NEARG         ] = 0;
      sf->thresh_mult[THR_NEARG_FILT    ] = 0;
      sf->thresh_mult[THR_NEARA         ] = 0;
      sf->thresh_mult[THR_NEARA_FILT    ] = 0;

      sf->thresh_mult[THR_DC       ] = 0;

      sf->thresh_mult[THR_V_PRED   ] = 1000;
      sf->thresh_mult[THR_H_PRED   ] = 1000;
      sf->thresh_mult[THR_D45_PRED ] = 1000;
      sf->thresh_mult[THR_D135_PRED] = 1000;
      sf->thresh_mult[THR_D117_PRED] = 1000;
      sf->thresh_mult[THR_D153_PRED] = 1000;
      sf->thresh_mult[THR_D27_PRED ] = 1000;
      sf->thresh_mult[THR_D63_PRED ] = 1000;
      sf->thresh_mult[THR_B_PRED   ] = 2000;
      sf->thresh_mult[THR_I8X8_PRED] = 2000;
      sf->thresh_mult[THR_TM       ] = 1000;

      sf->thresh_mult[THR_NEWMV    ] = 1000;
      sf->thresh_mult[THR_NEWG     ] = 1000;
      sf->thresh_mult[THR_NEWA     ] = 1000;
      sf->thresh_mult[THR_NEWMV_FILT    ] = 1000;
      sf->thresh_mult[THR_NEWG_FILT     ] = 1000;
      sf->thresh_mult[THR_NEWA_FILT     ] = 1000;
#else
      sf->thresh_mult[THR_ZEROMV   ] = 0;
      sf->thresh_mult[THR_ZEROG    ] = 0;
      sf->thresh_mult[THR_ZEROA    ] = 0;
      sf->thresh_mult[THR_NEARESTMV] = 0;
      sf->thresh_mult[THR_NEARESTG ] = 0;
      sf->thresh_mult[THR_NEARESTA ] = 0;
      sf->thresh_mult[THR_NEARMV   ] = 0;
      sf->thresh_mult[THR_NEARG    ] = 0;
      sf->thresh_mult[THR_NEARA    ] = 0;

      sf->thresh_mult[THR_DC       ] = 0;

      sf->thresh_mult[THR_V_PRED   ] = 1000;
      sf->thresh_mult[THR_H_PRED   ] = 1000;
      sf->thresh_mult[THR_D45_PRED ] = 1000;
      sf->thresh_mult[THR_D135_PRED] = 1000;
      sf->thresh_mult[THR_D117_PRED] = 1000;
      sf->thresh_mult[THR_D153_PRED] = 1000;
      sf->thresh_mult[THR_D27_PRED ] = 1000;
      sf->thresh_mult[THR_D63_PRED ] = 1000;
      sf->thresh_mult[THR_B_PRED   ] = 2000;
      sf->thresh_mult[THR_I8X8_PRED] = 2000;
      sf->thresh_mult[THR_TM       ] = 1000;

      sf->thresh_mult[THR_NEWMV    ] = 1000;
      sf->thresh_mult[THR_NEWG     ] = 1000;
      sf->thresh_mult[THR_NEWA     ] = 1000;
#endif
      sf->thresh_mult[THR_SPLITMV  ] = 2500;
      sf->thresh_mult[THR_SPLITG   ] = 5000;
      sf->thresh_mult[THR_SPLITA   ] = 5000;

      sf->thresh_mult[THR_COMP_ZEROLG   ] = 0;
      sf->thresh_mult[THR_COMP_NEARESTLG] = 0;
      sf->thresh_mult[THR_COMP_NEARLG   ] = 0;
      sf->thresh_mult[THR_COMP_ZEROLA   ] = 0;
      sf->thresh_mult[THR_COMP_NEARESTLA] = 0;
      sf->thresh_mult[THR_COMP_NEARLA   ] = 0;
      sf->thresh_mult[THR_COMP_ZEROGA   ] = 0;
      sf->thresh_mult[THR_COMP_NEARESTGA] = 0;
      sf->thresh_mult[THR_COMP_NEARGA   ] = 0;

      sf->thresh_mult[THR_COMP_NEWLG    ] = 1000;
      sf->thresh_mult[THR_COMP_NEWLA    ] = 1000;
      sf->thresh_mult[THR_COMP_NEWGA    ] = 1000;

      sf->thresh_mult[THR_COMP_SPLITLA  ] = 2500;
      sf->thresh_mult[THR_COMP_SPLITGA  ] = 5000;
      sf->thresh_mult[THR_COMP_SPLITLG  ] = 5000;

      sf->first_step = 0;
      sf->max_step_search_steps = MAX_MVSEARCH_STEPS;
      sf->search_best_filter = SEARCH_BEST_FILTER;
      break;
    case 1:
#if CONFIG_PRED_FILTER
      sf->thresh_mult[THR_NEARESTMV] = 0;
      sf->thresh_mult[THR_NEARESTMV_FILT] = 0;
      sf->thresh_mult[THR_ZEROMV   ] = 0;
      sf->thresh_mult[THR_ZEROMV_FILT   ] = 0;
      sf->thresh_mult[THR_DC       ] = 0;
      sf->thresh_mult[THR_NEARMV   ] = 0;
      sf->thresh_mult[THR_NEARMV_FILT   ] = 0;
      sf->thresh_mult[THR_V_PRED   ] = 1000;
      sf->thresh_mult[THR_H_PRED   ] = 1000;
      sf->thresh_mult[THR_D45_PRED ] = 1000;
      sf->thresh_mult[THR_D135_PRED] = 1000;
      sf->thresh_mult[THR_D117_PRED] = 1000;
      sf->thresh_mult[THR_D153_PRED] = 1000;
      sf->thresh_mult[THR_D27_PRED ] = 1000;
      sf->thresh_mult[THR_D63_PRED ] = 1000;
      sf->thresh_mult[THR_B_PRED   ] = 2500;
      sf->thresh_mult[THR_I8X8_PRED] = 2500;
      sf->thresh_mult[THR_TM       ] = 1000;

      sf->thresh_mult[THR_NEARESTG ] = 1000;
      sf->thresh_mult[THR_NEARESTG_FILT ] = 1000;
      sf->thresh_mult[THR_NEARESTA ] = 1000;
      sf->thresh_mult[THR_NEARESTA_FILT ] = 1000;

      sf->thresh_mult[THR_ZEROG    ] = 1000;
      sf->thresh_mult[THR_ZEROA    ] = 1000;
      sf->thresh_mult[THR_NEARG    ] = 1000;
      sf->thresh_mult[THR_NEARA    ] = 1000;
      sf->thresh_mult[THR_ZEROG_FILT    ] = 1000;
      sf->thresh_mult[THR_ZEROA_FILT    ] = 1000;
      sf->thresh_mult[THR_NEARG_FILT    ] = 1000;
      sf->thresh_mult[THR_NEARA_FILT    ] = 1000;

      sf->thresh_mult[THR_ZEROMV   ] = 0;
      sf->thresh_mult[THR_ZEROG    ] = 0;
      sf->thresh_mult[THR_ZEROA    ] = 0;
      sf->thresh_mult[THR_NEARESTMV] = 0;
      sf->thresh_mult[THR_NEARESTG ] = 0;
      sf->thresh_mult[THR_NEARESTA ] = 0;
      sf->thresh_mult[THR_NEARMV   ] = 0;
      sf->thresh_mult[THR_NEARG    ] = 0;
      sf->thresh_mult[THR_NEARA    ] = 0;
      sf->thresh_mult[THR_ZEROMV_FILT   ] = 0;
      sf->thresh_mult[THR_ZEROG_FILT    ] = 0;
      sf->thresh_mult[THR_ZEROA_FILT    ] = 0;
      sf->thresh_mult[THR_NEARESTMV_FILT] = 0;
      sf->thresh_mult[THR_NEARESTG_FILT ] = 0;
      sf->thresh_mult[THR_NEARESTA_FILT ] = 0;
      sf->thresh_mult[THR_NEARMV_FILT   ] = 0;
      sf->thresh_mult[THR_NEARG_FILT    ] = 0;
      sf->thresh_mult[THR_NEARA_FILT    ] = 0;

      sf->thresh_mult[THR_NEWMV    ] = 1000;
      sf->thresh_mult[THR_NEWG     ] = 1000;
      sf->thresh_mult[THR_NEWA     ] = 1000;
      sf->thresh_mult[THR_NEWMV_FILT    ] = 1000;
      sf->thresh_mult[THR_NEWG_FILT     ] = 1000;
      sf->thresh_mult[THR_NEWA_FILT     ] = 1000;
#else
      sf->thresh_mult[THR_NEARESTMV] = 0;
      sf->thresh_mult[THR_ZEROMV   ] = 0;
      sf->thresh_mult[THR_DC       ] = 0;
      sf->thresh_mult[THR_NEARMV   ] = 0;
      sf->thresh_mult[THR_V_PRED   ] = 1000;
      sf->thresh_mult[THR_H_PRED   ] = 1000;
      sf->thresh_mult[THR_D45_PRED ] = 1000;
      sf->thresh_mult[THR_D135_PRED] = 1000;
      sf->thresh_mult[THR_D117_PRED] = 1000;
      sf->thresh_mult[THR_D153_PRED] = 1000;
      sf->thresh_mult[THR_D27_PRED ] = 1000;
      sf->thresh_mult[THR_D63_PRED ] = 1000;
      sf->thresh_mult[THR_B_PRED   ] = 2500;
      sf->thresh_mult[THR_I8X8_PRED] = 2500;
      sf->thresh_mult[THR_TM       ] = 1000;

      sf->thresh_mult[THR_NEARESTG ] = 1000;
      sf->thresh_mult[THR_NEARESTA ] = 1000;

      sf->thresh_mult[THR_ZEROG    ] = 1000;
      sf->thresh_mult[THR_ZEROA    ] = 1000;
      sf->thresh_mult[THR_NEARG    ] = 1000;
      sf->thresh_mult[THR_NEARA    ] = 1000;

      sf->thresh_mult[THR_ZEROMV   ] = 0;
      sf->thresh_mult[THR_ZEROG    ] = 0;
      sf->thresh_mult[THR_ZEROA    ] = 0;
      sf->thresh_mult[THR_NEARESTMV] = 0;
      sf->thresh_mult[THR_NEARESTG ] = 0;
      sf->thresh_mult[THR_NEARESTA ] = 0;
      sf->thresh_mult[THR_NEARMV   ] = 0;
      sf->thresh_mult[THR_NEARG    ] = 0;
      sf->thresh_mult[THR_NEARA    ] = 0;

      sf->thresh_mult[THR_NEWMV    ] = 1000;
      sf->thresh_mult[THR_NEWG     ] = 1000;
      sf->thresh_mult[THR_NEWA     ] = 1000;
#endif
      sf->thresh_mult[THR_SPLITMV  ] = 1700;
      sf->thresh_mult[THR_SPLITG   ] = 4500;
      sf->thresh_mult[THR_SPLITA   ] = 4500;

      sf->thresh_mult[THR_COMP_ZEROLG   ] = 0;
      sf->thresh_mult[THR_COMP_NEARESTLG] = 0;
      sf->thresh_mult[THR_COMP_NEARLG   ] = 0;
      sf->thresh_mult[THR_COMP_ZEROLA   ] = 0;
      sf->thresh_mult[THR_COMP_NEARESTLA] = 0;
      sf->thresh_mult[THR_COMP_NEARLA   ] = 0;
      sf->thresh_mult[THR_COMP_ZEROGA   ] = 0;
      sf->thresh_mult[THR_COMP_NEARESTGA] = 0;
      sf->thresh_mult[THR_COMP_NEARGA   ] = 0;

      sf->thresh_mult[THR_COMP_NEWLG    ] = 1000;
      sf->thresh_mult[THR_COMP_NEWLA    ] = 1000;
      sf->thresh_mult[THR_COMP_NEWGA    ] = 1000;

      sf->thresh_mult[THR_COMP_SPLITLA  ] = 1700;
      sf->thresh_mult[THR_COMP_SPLITGA  ] = 4500;
      sf->thresh_mult[THR_COMP_SPLITLG  ] = 4500;

      if (Speed > 0) {
        /* Disable coefficient optimization above speed 0 */
        sf->optimize_coefficients = 0;
        sf->no_skip_block4x4_search = 0;

        sf->first_step = 1;

        cpi->mode_check_freq[THR_SPLITG] = 2;
        cpi->mode_check_freq[THR_SPLITA] = 2;
        cpi->mode_check_freq[THR_SPLITMV] = 0;

        cpi->mode_check_freq[THR_COMP_SPLITGA] = 2;
        cpi->mode_check_freq[THR_COMP_SPLITLG] = 2;
        cpi->mode_check_freq[THR_COMP_SPLITLA] = 0;
      }

      if (Speed > 1) {
        cpi->mode_check_freq[THR_SPLITG] = 4;
        cpi->mode_check_freq[THR_SPLITA] = 4;
        cpi->mode_check_freq[THR_SPLITMV] = 2;

        cpi->mode_check_freq[THR_COMP_SPLITGA] = 4;
        cpi->mode_check_freq[THR_COMP_SPLITLG] = 4;
        cpi->mode_check_freq[THR_COMP_SPLITLA] = 2;

        sf->thresh_mult[THR_TM       ] = 1500;
        sf->thresh_mult[THR_V_PRED   ] = 1500;
        sf->thresh_mult[THR_H_PRED   ] = 1500;
        sf->thresh_mult[THR_D45_PRED ] = 1500;
        sf->thresh_mult[THR_D135_PRED] = 1500;
        sf->thresh_mult[THR_D117_PRED] = 1500;
        sf->thresh_mult[THR_D153_PRED] = 1500;
        sf->thresh_mult[THR_D27_PRED ] = 1500;
        sf->thresh_mult[THR_D63_PRED ] = 1500;
        sf->thresh_mult[THR_B_PRED   ] = 5000;
        sf->thresh_mult[THR_I8X8_PRED] = 5000;

        if (cpi->ref_frame_flags & VP9_LAST_FLAG) {
          sf->thresh_mult[THR_NEWMV    ] = 2000;
#if CONFIG_PRED_FILTER
          sf->thresh_mult[THR_NEWMV_FILT    ] = 2000;
#endif
          sf->thresh_mult[THR_SPLITMV  ] = 10000;
          sf->thresh_mult[THR_COMP_SPLITLG  ] = 20000;
        }

        if (cpi->ref_frame_flags & VP9_GOLD_FLAG) {
          sf->thresh_mult[THR_NEARESTG ] = 1500;
          sf->thresh_mult[THR_ZEROG    ] = 1500;
          sf->thresh_mult[THR_NEARG    ] = 1500;
          sf->thresh_mult[THR_NEWG     ] = 2000;
#if CONFIG_PRED_FILTER
          sf->thresh_mult[THR_NEARESTG_FILT ] = 1500;
          sf->thresh_mult[THR_ZEROG_FILT    ] = 1500;
          sf->thresh_mult[THR_NEARG_FILT    ] = 1500;
          sf->thresh_mult[THR_NEWG_FILT     ] = 2000;
#endif
          sf->thresh_mult[THR_SPLITG   ] = 20000;
          sf->thresh_mult[THR_COMP_SPLITGA  ] = 20000;
        }

        if (cpi->ref_frame_flags & VP9_ALT_FLAG) {
          sf->thresh_mult[THR_NEARESTA ] = 1500;
          sf->thresh_mult[THR_ZEROA    ] = 1500;
          sf->thresh_mult[THR_NEARA    ] = 1500;
          sf->thresh_mult[THR_NEWA     ] = 2000;
#if CONFIG_PRED_FILTER
          sf->thresh_mult[THR_NEARESTA_FILT ] = 1500;
          sf->thresh_mult[THR_ZEROA_FILT    ] = 1500;
          sf->thresh_mult[THR_NEARA_FILT    ] = 1500;
          sf->thresh_mult[THR_NEWA_FILT     ] = 2000;
#endif
          sf->thresh_mult[THR_SPLITA   ] = 20000;
          sf->thresh_mult[THR_COMP_SPLITLA  ] = 10000;
        }

        sf->thresh_mult[THR_COMP_ZEROLG   ] = 1500;
        sf->thresh_mult[THR_COMP_NEARESTLG] = 1500;
        sf->thresh_mult[THR_COMP_NEARLG   ] = 1500;
        sf->thresh_mult[THR_COMP_ZEROLA   ] = 1500;
        sf->thresh_mult[THR_COMP_NEARESTLA] = 1500;
        sf->thresh_mult[THR_COMP_NEARLA   ] = 1500;
        sf->thresh_mult[THR_COMP_ZEROGA   ] = 1500;
        sf->thresh_mult[THR_COMP_NEARESTGA] = 1500;
        sf->thresh_mult[THR_COMP_NEARGA   ] = 1500;

        sf->thresh_mult[THR_COMP_NEWLG    ] = 2000;
        sf->thresh_mult[THR_COMP_NEWLA    ] = 2000;
        sf->thresh_mult[THR_COMP_NEWGA    ] = 2000;
      }

      if (Speed > 2) {
        cpi->mode_check_freq[THR_SPLITG] = 15;
        cpi->mode_check_freq[THR_SPLITA] = 15;
        cpi->mode_check_freq[THR_SPLITMV] = 7;

        cpi->mode_check_freq[THR_COMP_SPLITGA] = 15;
        cpi->mode_check_freq[THR_COMP_SPLITLG] = 15;
        cpi->mode_check_freq[THR_COMP_SPLITLA] = 7;

        sf->thresh_mult[THR_TM       ] = 2000;
        sf->thresh_mult[THR_V_PRED   ] = 2000;
        sf->thresh_mult[THR_H_PRED   ] = 2000;
        sf->thresh_mult[THR_D45_PRED ] = 2000;
        sf->thresh_mult[THR_D135_PRED] = 2000;
        sf->thresh_mult[THR_D117_PRED] = 2000;
        sf->thresh_mult[THR_D153_PRED] = 2000;
        sf->thresh_mult[THR_D27_PRED ] = 2000;
        sf->thresh_mult[THR_D63_PRED ] = 2000;
        sf->thresh_mult[THR_B_PRED   ] = 7500;
        sf->thresh_mult[THR_I8X8_PRED] = 7500;

        if (cpi->ref_frame_flags & VP9_LAST_FLAG) {
          sf->thresh_mult[THR_NEWMV    ] = 2000;
#if CONFIG_PRED_FILTER
          sf->thresh_mult[THR_NEWMV_FILT    ] = 2000;
#endif
          sf->thresh_mult[THR_SPLITMV  ] = 25000;
          sf->thresh_mult[THR_COMP_SPLITLG  ] = 50000;
        }

        if (cpi->ref_frame_flags & VP9_GOLD_FLAG) {
          sf->thresh_mult[THR_NEARESTG ] = 2000;
          sf->thresh_mult[THR_ZEROG    ] = 2000;
          sf->thresh_mult[THR_NEARG    ] = 2000;
          sf->thresh_mult[THR_NEWG     ] = 2500;
#if CONFIG_PRED_FILTER
          sf->thresh_mult[THR_NEARESTG_FILT ] = 2000;
          sf->thresh_mult[THR_ZEROG_FILT    ] = 2000;
          sf->thresh_mult[THR_NEARG_FILT    ] = 2000;
          sf->thresh_mult[THR_NEWG_FILT     ] = 2500;
#endif
          sf->thresh_mult[THR_SPLITG   ] = 50000;
          sf->thresh_mult[THR_COMP_SPLITGA  ] = 50000;
        }

        if (cpi->ref_frame_flags & VP9_ALT_FLAG) {
          sf->thresh_mult[THR_NEARESTA ] = 2000;
          sf->thresh_mult[THR_ZEROA    ] = 2000;
          sf->thresh_mult[THR_NEARA    ] = 2000;
          sf->thresh_mult[THR_NEWA     ] = 2500;
#if CONFIG_PRED_FILTER
          sf->thresh_mult[THR_NEARESTA_FILT ] = 2000;
          sf->thresh_mult[THR_ZEROA_FILT    ] = 2000;
          sf->thresh_mult[THR_NEARA_FILT    ] = 2000;
          sf->thresh_mult[THR_NEWA_FILT     ] = 2500;
#endif
          sf->thresh_mult[THR_SPLITA   ] = 50000;
          sf->thresh_mult[THR_COMP_SPLITLA  ] = 25000;
        }

        sf->thresh_mult[THR_COMP_ZEROLG   ] = 2000;
        sf->thresh_mult[THR_COMP_NEARESTLG] = 2000;
        sf->thresh_mult[THR_COMP_NEARLG   ] = 2000;
        sf->thresh_mult[THR_COMP_ZEROLA   ] = 2000;
        sf->thresh_mult[THR_COMP_NEARESTLA] = 2000;
        sf->thresh_mult[THR_COMP_NEARLA   ] = 2000;
        sf->thresh_mult[THR_COMP_ZEROGA   ] = 2000;
        sf->thresh_mult[THR_COMP_NEARESTGA] = 2000;
        sf->thresh_mult[THR_COMP_NEARGA   ] = 2000;

        sf->thresh_mult[THR_COMP_NEWLG    ] = 2500;
        sf->thresh_mult[THR_COMP_NEWLA    ] = 2500;
        sf->thresh_mult[THR_COMP_NEWGA    ] = 2500;

        sf->improved_dct = 0;

        // Only do recode loop on key frames, golden frames and
        // alt ref frames
        sf->recode_loop = 2;

      }

      break;

  }; /* switch */

  /* disable frame modes if flags not set */
  if (!(cpi->ref_frame_flags & VP9_LAST_FLAG)) {
    sf->thresh_mult[THR_NEWMV    ] = INT_MAX;
    sf->thresh_mult[THR_NEARESTMV] = INT_MAX;
    sf->thresh_mult[THR_ZEROMV   ] = INT_MAX;
    sf->thresh_mult[THR_NEARMV   ] = INT_MAX;
#if CONFIG_PRED_FILTER
    sf->thresh_mult[THR_NEWMV_FILT    ] = INT_MAX;
    sf->thresh_mult[THR_NEARESTMV_FILT] = INT_MAX;
    sf->thresh_mult[THR_ZEROMV_FILT   ] = INT_MAX;
    sf->thresh_mult[THR_NEARMV_FILT   ] = INT_MAX;
#endif
    sf->thresh_mult[THR_SPLITMV  ] = INT_MAX;
  }

  if (!(cpi->ref_frame_flags & VP9_GOLD_FLAG)) {
    sf->thresh_mult[THR_NEARESTG ] = INT_MAX;
    sf->thresh_mult[THR_ZEROG    ] = INT_MAX;
    sf->thresh_mult[THR_NEARG    ] = INT_MAX;
    sf->thresh_mult[THR_NEWG     ] = INT_MAX;
#if CONFIG_PRED_FILTER
    sf->thresh_mult[THR_NEARESTG_FILT ] = INT_MAX;
    sf->thresh_mult[THR_ZEROG_FILT    ] = INT_MAX;
    sf->thresh_mult[THR_NEARG_FILT    ] = INT_MAX;
    sf->thresh_mult[THR_NEWG_FILT     ] = INT_MAX;
#endif
    sf->thresh_mult[THR_SPLITG   ] = INT_MAX;
  }

  if (!(cpi->ref_frame_flags & VP9_ALT_FLAG)) {
    sf->thresh_mult[THR_NEARESTA ] = INT_MAX;
    sf->thresh_mult[THR_ZEROA    ] = INT_MAX;
    sf->thresh_mult[THR_NEARA    ] = INT_MAX;
    sf->thresh_mult[THR_NEWA     ] = INT_MAX;
#if CONFIG_PRED_FILTER
    sf->thresh_mult[THR_NEARESTA_FILT ] = INT_MAX;
    sf->thresh_mult[THR_ZEROA_FILT    ] = INT_MAX;
    sf->thresh_mult[THR_NEARA_FILT    ] = INT_MAX;
    sf->thresh_mult[THR_NEWA_FILT     ] = INT_MAX;
#endif
    sf->thresh_mult[THR_SPLITA   ] = INT_MAX;
  }

  if ((cpi->ref_frame_flags & (VP9_LAST_FLAG | VP9_GOLD_FLAG)) != (VP9_LAST_FLAG | VP9_GOLD_FLAG)) {
    sf->thresh_mult[THR_COMP_ZEROLG   ] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEARESTLG] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEARLG   ] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEWLG    ] = INT_MAX;
    sf->thresh_mult[THR_COMP_SPLITLG  ] = INT_MAX;
  }

  if ((cpi->ref_frame_flags & (VP9_LAST_FLAG | VP9_ALT_FLAG)) != (VP9_LAST_FLAG | VP9_ALT_FLAG)) {
    sf->thresh_mult[THR_COMP_ZEROLA   ] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEARESTLA] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEARLA   ] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEWLA    ] = INT_MAX;
    sf->thresh_mult[THR_COMP_SPLITLA  ] = INT_MAX;
  }

  if ((cpi->ref_frame_flags & (VP9_GOLD_FLAG | VP9_ALT_FLAG)) != (VP9_GOLD_FLAG | VP9_ALT_FLAG)) {
    sf->thresh_mult[THR_COMP_ZEROGA   ] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEARESTGA] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEARGA   ] = INT_MAX;
    sf->thresh_mult[THR_COMP_NEWGA    ] = INT_MAX;
    sf->thresh_mult[THR_COMP_SPLITGA  ] = INT_MAX;
  }

  // Slow quant, dct and trellis not worthwhile for first pass
  // so make sure they are always turned off.
  if (cpi->pass == 1) {
    sf->optimize_coefficients = 0;
    sf->improved_dct = 0;
  }

  if (cpi->sf.search_method == NSTEP) {
    vp9_init3smotion_compensation(&cpi->mb,
                                  cm->yv12_fb[cm->lst_fb_idx].y_stride);
  } else if (cpi->sf.search_method == DIAMOND) {
    vp9_init_dsmotion_compensation(&cpi->mb,
                                   cm->yv12_fb[cm->lst_fb_idx].y_stride);
  }

  cpi->mb.vp9_short_fdct16x16 = vp9_short_fdct16x16;
  cpi->mb.vp9_short_fdct8x8 = vp9_short_fdct8x8;
  cpi->mb.vp9_short_fdct8x4 = vp9_short_fdct8x4;
  cpi->mb.vp9_short_fdct4x4 = vp9_short_fdct4x4;
  cpi->mb.short_walsh4x4 = vp9_short_walsh4x4;
  cpi->mb.short_fhaar2x2 = vp9_short_fhaar2x2;

#if CONFIG_LOSSLESS
  if (cpi->oxcf.lossless) {
    cpi->mb.vp9_short_fdct8x4 = vp9_short_walsh8x4_x8;
    cpi->mb.vp9_short_fdct4x4 = vp9_short_walsh4x4_x8;
    cpi->mb.short_walsh4x4 = vp9_short_walsh4x4;
    cpi->mb.short_fhaar2x2 = vp9_short_fhaar2x2;
    cpi->mb.short_walsh4x4 = vp9_short_walsh4x4_lossless;
  }
#endif


  cpi->mb.quantize_b_4x4      = vp9_regular_quantize_b_4x4;
  cpi->mb.quantize_b_4x4_pair = vp9_regular_quantize_b_4x4_pair;
  cpi->mb.quantize_b_8x8      = vp9_regular_quantize_b_8x8;
  cpi->mb.quantize_b_16x16    = vp9_regular_quantize_b_16x16;
  cpi->mb.quantize_b_2x2      = vp9_regular_quantize_b_2x2;

  vp9_init_quantizer(cpi);

#if CONFIG_RUNTIME_CPU_DETECT
  cpi->mb.e_mbd.rtcd = &cpi->common.rtcd;
#endif

  if (cpi->sf.iterative_sub_pixel == 1) {
    cpi->find_fractional_mv_step = vp9_find_best_sub_pixel_step_iteratively;
  } else if (cpi->sf.quarter_pixel_search) {
    cpi->find_fractional_mv_step = vp9_find_best_sub_pixel_step;
  } else if (cpi->sf.half_pixel_search) {
    cpi->find_fractional_mv_step = vp9_find_best_half_pixel_step;
  }

  if (cpi->sf.optimize_coefficients == 1 && cpi->pass != 1)
    cpi->mb.optimize = 1;
  else
    cpi->mb.optimize = 0;

#ifdef SPEEDSTATS
  frames_at_speed[cpi->Speed]++;
#endif
}
static void alloc_raw_frame_buffers(VP9_COMP *cpi) {
  int width = (cpi->oxcf.Width + 15) & ~15;
  int height = (cpi->oxcf.Height + 15) & ~15;

  cpi->lookahead = vp9_lookahead_init(cpi->oxcf.Width, cpi->oxcf.Height,
                                      cpi->oxcf.lag_in_frames);
  if (!cpi->lookahead)
    vpx_internal_error(&cpi->common.error, VPX_CODEC_MEM_ERROR,
                       "Failed to allocate lag buffers");

#if VP9_TEMPORAL_ALT_REF

  if (vp8_yv12_alloc_frame_buffer(&cpi->alt_ref_buffer,
                                  width, height, VP9BORDERINPIXELS))
    vpx_internal_error(&cpi->common.error, VPX_CODEC_MEM_ERROR,
                       "Failed to allocate altref buffer");

#endif
}

static int alloc_partition_data(VP9_COMP *cpi) {
  vpx_free(cpi->mb.pip);

  cpi->mb.pip = vpx_calloc((cpi->common.mb_cols + 1) *
                           (cpi->common.mb_rows + 1),
                           sizeof(PARTITION_INFO));
  if (!cpi->mb.pip)
    return 1;

  cpi->mb.pi = cpi->mb.pip + cpi->common.mode_info_stride + 1;

  return 0;
}

void vp9_alloc_compressor_data(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;

  int width = cm->Width;
  int height = cm->Height;

  if (vp9_alloc_frame_buffers(cm, width, height))
    vpx_internal_error(&cpi->common.error, VPX_CODEC_MEM_ERROR,
                       "Failed to allocate frame buffers");

  if (alloc_partition_data(cpi))
    vpx_internal_error(&cpi->common.error, VPX_CODEC_MEM_ERROR,
                       "Failed to allocate partition data");


  if ((width & 0xf) != 0)
    width += 16 - (width & 0xf);

  if ((height & 0xf) != 0)
    height += 16 - (height & 0xf);


  if (vp8_yv12_alloc_frame_buffer(&cpi->last_frame_uf,
                                  width, height, VP9BORDERINPIXELS))
    vpx_internal_error(&cpi->common.error, VPX_CODEC_MEM_ERROR,
                       "Failed to allocate last frame buffer");

  if (vp8_yv12_alloc_frame_buffer(&cpi->scaled_source,
                                  width, height, VP9BORDERINPIXELS))
    vpx_internal_error(&cpi->common.error, VPX_CODEC_MEM_ERROR,
                       "Failed to allocate scaled source buffer");


  vpx_free(cpi->tok);

  {
    unsigned int tokens = cm->mb_rows * cm->mb_cols * 24 * 16;

    CHECK_MEM_ERROR(cpi->tok, vpx_calloc(tokens, sizeof(*cpi->tok)));
  }

  // Data used for real time vc mode to see if gf needs refreshing
  cpi->inter_zz_count = 0;
  cpi->gf_bad_count = 0;
  cpi->gf_update_recommended = 0;


  // Structures used to minitor GF usage
  vpx_free(cpi->gf_active_flags);
  CHECK_MEM_ERROR(cpi->gf_active_flags,
                  vpx_calloc(1, cm->mb_rows * cm->mb_cols));
  cpi->gf_active_count = cm->mb_rows * cm->mb_cols;

  vpx_free(cpi->mb_activity_map);
  CHECK_MEM_ERROR(cpi->mb_activity_map,
                  vpx_calloc(sizeof(unsigned int),
                             cm->mb_rows * cm->mb_cols));

  vpx_free(cpi->mb_norm_activity_map);
  CHECK_MEM_ERROR(cpi->mb_norm_activity_map,
                  vpx_calloc(sizeof(unsigned int),
                             cm->mb_rows * cm->mb_cols));

  vpx_free(cpi->twopass.total_stats);

  cpi->twopass.total_stats = vpx_calloc(1, sizeof(FIRSTPASS_STATS));

  vpx_free(cpi->twopass.total_left_stats);
  cpi->twopass.total_left_stats = vpx_calloc(1, sizeof(FIRSTPASS_STATS));

  vpx_free(cpi->twopass.this_frame_stats);

  cpi->twopass.this_frame_stats = vpx_calloc(1, sizeof(FIRSTPASS_STATS));

  if (!cpi->twopass.total_stats ||
      !cpi->twopass.total_left_stats ||
      !cpi->twopass.this_frame_stats)
    vpx_internal_error(&cpi->common.error, VPX_CODEC_MEM_ERROR,
                       "Failed to allocate firstpass stats");

  vpx_free(cpi->tplist);

  CHECK_MEM_ERROR(cpi->tplist,
                  vpx_malloc(sizeof(TOKENLIST) * (cpi->common.mb_rows)));
}


// TODO perhaps change number of steps expose to outside world when setting
// max and min limits. Also this will likely want refining for the extended Q
// range.
//
// Table that converts 0-63 Q range values passed in outside to the Qindex
// range used internally.
static const int q_trans[] = {
  0,    4,   8,  12,  16,  20,  24,  28,
  32,   36,  40,  44,  48,  52,  56,  60,
  64,   68,  72,  76,  80,  84,  88,  92,
  96,  100, 104, 108, 112, 116, 120, 124,
  128, 132, 136, 140, 144, 148, 152, 156,
  160, 164, 168, 172, 176, 180, 184, 188,
  192, 196, 200, 204, 208, 212, 216, 220,
  224, 228, 232, 236, 240, 244, 249, 255,
};

int vp9_reverse_trans(int x) {
  int i;

  for (i = 0; i < 64; i++)
    if (q_trans[i] >= x)
      return i;

  return 63;
};
void vp9_new_frame_rate(VP9_COMP *cpi, double framerate) {
  if (framerate < .1)
    framerate = 30;

  cpi->oxcf.frame_rate             = framerate;
  cpi->output_frame_rate            = cpi->oxcf.frame_rate;
  cpi->per_frame_bandwidth          = (int)(cpi->oxcf.target_bandwidth / cpi->output_frame_rate);
  cpi->av_per_frame_bandwidth        = (int)(cpi->oxcf.target_bandwidth / cpi->output_frame_rate);
  cpi->min_frame_bandwidth          = (int)(cpi->av_per_frame_bandwidth * cpi->oxcf.two_pass_vbrmin_section / 100);

  if (cpi->min_frame_bandwidth < FRAME_OVERHEAD_BITS)
    cpi->min_frame_bandwidth = FRAME_OVERHEAD_BITS;

  // Set Maximum gf/arf interval
  cpi->max_gf_interval = ((int)(cpi->output_frame_rate / 2.0) + 2);

  if (cpi->max_gf_interval < 12)
    cpi->max_gf_interval = 12;

  // Extended interval for genuinely static scenes
  cpi->twopass.static_scene_max_gf_interval = cpi->key_frame_frequency >> 1;

  // Special conditions when altr ref frame enabled in lagged compress mode
  if (cpi->oxcf.play_alternate && cpi->oxcf.lag_in_frames) {
    if (cpi->max_gf_interval > cpi->oxcf.lag_in_frames - 1)
      cpi->max_gf_interval = cpi->oxcf.lag_in_frames - 1;

    if (cpi->twopass.static_scene_max_gf_interval > cpi->oxcf.lag_in_frames - 1)
      cpi->twopass.static_scene_max_gf_interval = cpi->oxcf.lag_in_frames - 1;
  }

  if (cpi->max_gf_interval > cpi->twopass.static_scene_max_gf_interval)
    cpi->max_gf_interval = cpi->twopass.static_scene_max_gf_interval;
}


static int
rescale(int val, int num, int denom) {
  int64_t llnum = num;
  int64_t llden = denom;
  int64_t llval = val;

  return llval * llnum / llden;
}


static void init_config(VP9_PTR ptr, VP9_CONFIG *oxcf) {
  VP9_COMP *cpi = (VP9_COMP *)(ptr);
  VP9_COMMON *cm = &cpi->common;

  cpi->oxcf = *oxcf;

  cpi->goldfreq = 7;

  cm->version = oxcf->Version;
  vp9_setup_version(cm);

  // change includes all joint functionality
  vp9_change_config(ptr, oxcf);

  // Initialize active best and worst q and average q values.
  cpi->active_worst_quality         = cpi->oxcf.worst_allowed_q;
  cpi->active_best_quality          = cpi->oxcf.best_allowed_q;
  cpi->avg_frame_qindex             = cpi->oxcf.worst_allowed_q;

  // Initialise the starting buffer levels
  cpi->buffer_level                 = cpi->oxcf.starting_buffer_level;
  cpi->bits_off_target              = cpi->oxcf.starting_buffer_level;

  cpi->rolling_target_bits          = cpi->av_per_frame_bandwidth;
  cpi->rolling_actual_bits          = cpi->av_per_frame_bandwidth;
  cpi->long_rolling_target_bits     = cpi->av_per_frame_bandwidth;
  cpi->long_rolling_actual_bits     = cpi->av_per_frame_bandwidth;

  cpi->total_actual_bits            = 0;
  cpi->total_target_vs_actual       = 0;

  cpi->static_mb_pct = 0;

#if VP9_TEMPORAL_ALT_REF
  {
    int i;

    cpi->fixed_divide[0] = 0;

    for (i = 1; i < 512; i++)
      cpi->fixed_divide[i] = 0x80000 / i;
  }
#endif
}


void vp9_change_config(VP9_PTR ptr, VP9_CONFIG *oxcf) {
  VP9_COMP *cpi = (VP9_COMP *)(ptr);
  VP9_COMMON *cm = &cpi->common;

  if (!cpi)
    return;

  if (!oxcf)
    return;

  if (cm->version != oxcf->Version) {
    cm->version = oxcf->Version;
    vp9_setup_version(cm);
  }

  cpi->oxcf = *oxcf;

  switch (cpi->oxcf.Mode) {
      // Real time and one pass deprecated in test code base
    case MODE_FIRSTPASS:
      cpi->pass = 1;
      cpi->compressor_speed = 1;
      break;

    case MODE_SECONDPASS:
      cpi->pass = 2;
      cpi->compressor_speed = 1;

      if (cpi->oxcf.cpu_used < -5) {
        cpi->oxcf.cpu_used = -5;
      }

      if (cpi->oxcf.cpu_used > 5)
        cpi->oxcf.cpu_used = 5;

      break;

    case MODE_SECONDPASS_BEST:
      cpi->pass = 2;
      cpi->compressor_speed = 0;
      break;
  }

  cpi->oxcf.worst_allowed_q = q_trans[oxcf->worst_allowed_q];
  cpi->oxcf.best_allowed_q = q_trans[oxcf->best_allowed_q];
  cpi->oxcf.cq_level = q_trans[cpi->oxcf.cq_level];

#if CONFIG_LOSSLESS
  cpi->oxcf.lossless = oxcf->lossless;
  if (cpi->oxcf.lossless) {
    cpi->common.rtcd.idct.idct1        = vp9_short_inv_walsh4x4_1_x8_c;
    cpi->common.rtcd.idct.idct16       = vp9_short_inv_walsh4x4_x8_c;
    cpi->common.rtcd.idct.idct1_scalar_add  = vp9_dc_only_inv_walsh_add_c;
    cpi->common.rtcd.idct.iwalsh1      = vp9_short_inv_walsh4x4_1_c;
    cpi->common.rtcd.idct.iwalsh16     = vp9_short_inv_walsh4x4_lossless_c;
  }
#endif

  cpi->baseline_gf_interval = DEFAULT_GF_INTERVAL;

  cpi->ref_frame_flags = VP9_ALT_FLAG | VP9_GOLD_FLAG | VP9_LAST_FLAG;

  // cpi->use_golden_frame_only = 0;
  // cpi->use_last_frame_only = 0;
  cm->refresh_golden_frame = 0;
  cm->refresh_last_frame = 1;
  cm->refresh_entropy_probs = 1;

  setup_features(cpi);
  cpi->mb.e_mbd.allow_high_precision_mv = 0;   // Default mv precision adaptation

  {
    int i;

    for (i = 0; i < MAX_MB_SEGMENTS; i++)
      cpi->segment_encode_breakout[i] = cpi->oxcf.encode_breakout;
  }

  // At the moment the first order values may not be > MAXQ
  if (cpi->oxcf.fixed_q > MAXQ)
    cpi->oxcf.fixed_q = MAXQ;

  // local file playback mode == really big buffer
  if (cpi->oxcf.end_usage == USAGE_LOCAL_FILE_PLAYBACK) {
    cpi->oxcf.starting_buffer_level   = 60000;
    cpi->oxcf.optimal_buffer_level    = 60000;
    cpi->oxcf.maximum_buffer_size     = 240000;
  }

  // Convert target bandwidth from Kbit/s to Bit/s
  cpi->oxcf.target_bandwidth       *= 1000;

  cpi->oxcf.starting_buffer_level =
    rescale(cpi->oxcf.starting_buffer_level,
            cpi->oxcf.target_bandwidth, 1000);

  // Set or reset optimal and maximum buffer levels.
  if (cpi->oxcf.optimal_buffer_level == 0)
    cpi->oxcf.optimal_buffer_level = cpi->oxcf.target_bandwidth / 8;
  else
    cpi->oxcf.optimal_buffer_level =
      rescale(cpi->oxcf.optimal_buffer_level,
              cpi->oxcf.target_bandwidth, 1000);

  if (cpi->oxcf.maximum_buffer_size == 0)
    cpi->oxcf.maximum_buffer_size = cpi->oxcf.target_bandwidth / 8;
  else
    cpi->oxcf.maximum_buffer_size =
      rescale(cpi->oxcf.maximum_buffer_size,
              cpi->oxcf.target_bandwidth, 1000);

  // Set up frame rate and related parameters rate control values.
  vp9_new_frame_rate(cpi, cpi->oxcf.frame_rate);

  // Set absolute upper and lower quality limits
  cpi->worst_quality               = cpi->oxcf.worst_allowed_q;
  cpi->best_quality                = cpi->oxcf.best_allowed_q;

  // active values should only be modified if out of new range
  if (cpi->active_worst_quality > cpi->oxcf.worst_allowed_q) {
    cpi->active_worst_quality = cpi->oxcf.worst_allowed_q;
  }
  // less likely
  else if (cpi->active_worst_quality < cpi->oxcf.best_allowed_q) {
    cpi->active_worst_quality = cpi->oxcf.best_allowed_q;
  }
  if (cpi->active_best_quality < cpi->oxcf.best_allowed_q) {
    cpi->active_best_quality = cpi->oxcf.best_allowed_q;
  }
  // less likely
  else if (cpi->active_best_quality > cpi->oxcf.worst_allowed_q) {
    cpi->active_best_quality = cpi->oxcf.worst_allowed_q;
  }

  cpi->buffered_mode = (cpi->oxcf.optimal_buffer_level > 0) ? TRUE : FALSE;

  cpi->cq_target_quality = cpi->oxcf.cq_level;

  if (!cm->use_bilinear_mc_filter)
    cm->mcomp_filter_type = DEFAULT_INTERP_FILTER;
  else
    cm->mcomp_filter_type = BILINEAR;

  cpi->target_bandwidth = cpi->oxcf.target_bandwidth;

  cm->Width       = cpi->oxcf.Width;
  cm->Height      = cpi->oxcf.Height;

  cm->horiz_scale  = cpi->horiz_scale;
  cm->vert_scale   = cpi->vert_scale;

  // VP8 sharpness level mapping 0-7 (vs 0-10 in general VPx dialogs)
  if (cpi->oxcf.Sharpness > 7)
    cpi->oxcf.Sharpness = 7;

  cm->sharpness_level = cpi->oxcf.Sharpness;

  if (cm->horiz_scale != NORMAL || cm->vert_scale != NORMAL) {
    int UNINITIALIZED_IS_SAFE(hr), UNINITIALIZED_IS_SAFE(hs);
    int UNINITIALIZED_IS_SAFE(vr), UNINITIALIZED_IS_SAFE(vs);

    Scale2Ratio(cm->horiz_scale, &hr, &hs);
    Scale2Ratio(cm->vert_scale, &vr, &vs);

    // always go to the next whole number
    cm->Width = (hs - 1 + cpi->oxcf.Width * hr) / hs;
    cm->Height = (vs - 1 + cpi->oxcf.Height * vr) / vs;
  }

  if (((cm->Width + 15) & 0xfffffff0) !=
      cm->yv12_fb[cm->lst_fb_idx].y_width ||
      ((cm->Height + 15) & 0xfffffff0) !=
      cm->yv12_fb[cm->lst_fb_idx].y_height ||
      cm->yv12_fb[cm->lst_fb_idx].y_width == 0) {
    alloc_raw_frame_buffers(cpi);
    vp9_alloc_compressor_data(cpi);
  }

  if (cpi->oxcf.fixed_q >= 0) {
    cpi->last_q[0] = cpi->oxcf.fixed_q;
    cpi->last_q[1] = cpi->oxcf.fixed_q;
    cpi->last_boosted_qindex = cpi->oxcf.fixed_q;
  }

  cpi->Speed = cpi->oxcf.cpu_used;

  // force to allowlag to 0 if lag_in_frames is 0;
  if (cpi->oxcf.lag_in_frames == 0) {
    cpi->oxcf.allow_lag = 0;
  }
  // Limit on lag buffers as these are not currently dynamically allocated
  else if (cpi->oxcf.lag_in_frames > MAX_LAG_BUFFERS)
    cpi->oxcf.lag_in_frames = MAX_LAG_BUFFERS;

  // YX Temp
  cpi->alt_ref_source = NULL;
  cpi->is_src_frame_alt_ref = 0;

#if 0
  // Experimental RD Code
  cpi->frame_distortion = 0;
  cpi->last_frame_distortion = 0;
#endif

}

#define M_LOG2_E 0.693147180559945309417
#define log2f(x) (log (x) / (float) M_LOG2_E)

static void cal_nmvjointsadcost(int *mvjointsadcost) {
  mvjointsadcost[0] = 600;
  mvjointsadcost[1] = 300;
  mvjointsadcost[2] = 300;
  mvjointsadcost[0] = 300;
}

static void cal_nmvsadcosts(int *mvsadcost[2]) {
  int i = 1;

  mvsadcost [0] [0] = 0;
  mvsadcost [1] [0] = 0;

  do {
    double z = 256 * (2 * (log2f(8 * i) + .6));
    mvsadcost [0][i] = (int) z;
    mvsadcost [1][i] = (int) z;
    mvsadcost [0][-i] = (int) z;
    mvsadcost [1][-i] = (int) z;
  } while (++i <= MV_MAX);
}

static void cal_nmvsadcosts_hp(int *mvsadcost[2]) {
  int i = 1;

  mvsadcost [0] [0] = 0;
  mvsadcost [1] [0] = 0;

  do {
    double z = 256 * (2 * (log2f(8 * i) + .6));
    mvsadcost [0][i] = (int) z;
    mvsadcost [1][i] = (int) z;
    mvsadcost [0][-i] = (int) z;
    mvsadcost [1][-i] = (int) z;
  } while (++i <= MV_MAX);
}

VP9_PTR vp9_create_compressor(VP9_CONFIG *oxcf) {
  int i;
  volatile union {
    VP9_COMP *cpi;
    VP9_PTR   ptr;
  } ctx;

  VP9_COMP *cpi;
  VP9_COMMON *cm;

  cpi = ctx.cpi = vpx_memalign(32, sizeof(VP9_COMP));
  // Check that the CPI instance is valid
  if (!cpi)
    return 0;

  cm = &cpi->common;

  vpx_memset(cpi, 0, sizeof(VP9_COMP));

  if (setjmp(cm->error.jmp)) {
    VP9_PTR ptr = ctx.ptr;

    ctx.cpi->common.error.setjmp = 0;
    vp9_remove_compressor(&ptr);
    return 0;
  }

  cpi->common.error.setjmp = 1;

  CHECK_MEM_ERROR(cpi->mb.ss, vpx_calloc(sizeof(search_site), (MAX_MVSEARCH_STEPS * 8) + 1));

  vp9_create_common(&cpi->common);
  vp9_cmachine_specific_config(cpi);

  init_config((VP9_PTR)cpi, oxcf);

  memcpy(cpi->base_skip_false_prob, base_skip_false_prob, sizeof(base_skip_false_prob));
  cpi->common.current_video_frame   = 0;
  cpi->kf_overspend_bits            = 0;
  cpi->kf_bitrate_adjustment        = 0;
  cpi->frames_till_gf_update_due      = 0;
  cpi->gf_overspend_bits            = 0;
  cpi->non_gf_bitrate_adjustment     = 0;
  cm->prob_last_coded               = 128;
  cm->prob_gf_coded                 = 128;
  cm->prob_intra_coded              = 63;
#if CONFIG_SUPERBLOCKS
  cm->sb_coded                      = 200;
#endif
  for (i = 0; i < COMP_PRED_CONTEXTS; i++)
    cm->prob_comppred[i]         = 128;
  for (i = 0; i < TX_SIZE_MAX - 1; i++)
    cm->prob_tx[i]               = 128;

  // Prime the recent reference frame useage counters.
  // Hereafter they will be maintained as a sort of moving average
  cpi->recent_ref_frame_usage[INTRA_FRAME]  = 1;
  cpi->recent_ref_frame_usage[LAST_FRAME]   = 1;
  cpi->recent_ref_frame_usage[GOLDEN_FRAME] = 1;
  cpi->recent_ref_frame_usage[ALTREF_FRAME] = 1;

  // Set reference frame sign bias for ALTREF frame to 1 (for now)
  cpi->common.ref_frame_sign_bias[ALTREF_FRAME] = 1;

  cpi->baseline_gf_interval = DEFAULT_GF_INTERVAL;

  cpi->gold_is_last = 0;
  cpi->alt_is_last  = 0;
  cpi->gold_is_alt  = 0;

  // allocate memory for storing last frame's MVs for MV prediction.
  CHECK_MEM_ERROR(cpi->lfmv, vpx_calloc((cpi->common.mb_rows + 2) * (cpi->common.mb_cols + 2), sizeof(int_mv)));
  CHECK_MEM_ERROR(cpi->lf_ref_frame_sign_bias, vpx_calloc((cpi->common.mb_rows + 2) * (cpi->common.mb_cols + 2), sizeof(int)));
  CHECK_MEM_ERROR(cpi->lf_ref_frame, vpx_calloc((cpi->common.mb_rows + 2) * (cpi->common.mb_cols + 2), sizeof(int)));

  // Create the encoder segmentation map and set all entries to 0
  CHECK_MEM_ERROR(cpi->segmentation_map, vpx_calloc((cpi->common.mb_rows * cpi->common.mb_cols), 1));

  // And a copy in common for temporal coding
  CHECK_MEM_ERROR(cm->last_frame_seg_map,
                  vpx_calloc((cpi->common.mb_rows * cpi->common.mb_cols), 1));

  // And a place holder structure is the coding context
  // for use if we want to save and restore it
  CHECK_MEM_ERROR(cpi->coding_context.last_frame_seg_map_copy,
                  vpx_calloc((cpi->common.mb_rows * cpi->common.mb_cols), 1));

  CHECK_MEM_ERROR(cpi->active_map, vpx_calloc(cpi->common.mb_rows * cpi->common.mb_cols, 1));
  vpx_memset(cpi->active_map, 1, (cpi->common.mb_rows * cpi->common.mb_cols));
  cpi->active_map_enabled = 0;

  for (i = 0; i < (sizeof(cpi->mbgraph_stats) /
                   sizeof(cpi->mbgraph_stats[0])); i++) {
    CHECK_MEM_ERROR(cpi->mbgraph_stats[i].mb_stats,
                    vpx_calloc(cpi->common.mb_rows * cpi->common.mb_cols *
                               sizeof(*cpi->mbgraph_stats[i].mb_stats),
                               1));
  }

#ifdef ENTROPY_STATS
  if (cpi->pass != 1)
    init_context_counters();
#endif
#ifdef MODE_STATS
  vp9_zero(y_modes);
  vp9_zero(i8x8_modes);
  vp9_zero(uv_modes);
  vp9_zero(uv_modes_y);
  vp9_zero(b_modes);
  vp9_zero(inter_y_modes);
  vp9_zero(inter_uv_modes);
  vp9_zero(inter_b_modes);
#endif
#ifdef NMV_STATS
  init_nmvstats();
#endif

  /*Initialize the feed-forward activity masking.*/
  cpi->activity_avg = 90 << 12;

  cpi->frames_since_key = 8;        // Give a sensible default for the first frame.
  cpi->key_frame_frequency = cpi->oxcf.key_freq;
  cpi->this_key_frame_forced = FALSE;
  cpi->next_key_frame_forced = FALSE;

  cpi->source_alt_ref_pending = FALSE;
  cpi->source_alt_ref_active = FALSE;
  cpi->common.refresh_alt_ref_frame = 0;

  cpi->b_calculate_psnr = CONFIG_INTERNAL_STATS;
#if CONFIG_INTERNAL_STATS
  cpi->b_calculate_ssimg = 0;

  cpi->count = 0;
  cpi->bytes = 0;

  if (cpi->b_calculate_psnr) {
    cpi->total_sq_error = 0.0;
    cpi->total_sq_error2 = 0.0;
    cpi->total_y = 0.0;
    cpi->total_u = 0.0;
    cpi->total_v = 0.0;
    cpi->total = 0.0;
    cpi->totalp_y = 0.0;
    cpi->totalp_u = 0.0;
    cpi->totalp_v = 0.0;
    cpi->totalp = 0.0;
    cpi->tot_recode_hits = 0;
    cpi->summed_quality = 0;
    cpi->summed_weights = 0;
  }

  if (cpi->b_calculate_ssimg) {
    cpi->total_ssimg_y = 0;
    cpi->total_ssimg_u = 0;
    cpi->total_ssimg_v = 0;
    cpi->total_ssimg_all = 0;
  }

#endif

#ifndef LLONG_MAX
#define LLONG_MAX  9223372036854775807LL
#endif
  cpi->first_time_stamp_ever = LLONG_MAX;

  cpi->frames_till_gf_update_due      = 0;
  cpi->key_frame_count              = 1;

  cpi->ni_av_qi                     = cpi->oxcf.worst_allowed_q;
  cpi->ni_tot_qi                    = 0;
  cpi->ni_frames                   = 0;
  cpi->tot_q = 0.0;
  cpi->avg_q = vp9_convert_qindex_to_q(cpi->oxcf.worst_allowed_q);
  cpi->total_byte_count             = 0;

  cpi->rate_correction_factor         = 1.0;
  cpi->key_frame_rate_correction_factor = 1.0;
  cpi->gf_rate_correction_factor  = 1.0;
  cpi->twopass.est_max_qcorrection_factor  = 1.0;

  cal_nmvjointsadcost(cpi->mb.nmvjointsadcost);
  cpi->mb.nmvcost[0] = &cpi->mb.nmvcosts[0][MV_MAX];
  cpi->mb.nmvcost[1] = &cpi->mb.nmvcosts[1][MV_MAX];
  cpi->mb.nmvsadcost[0] = &cpi->mb.nmvsadcosts[0][MV_MAX];
  cpi->mb.nmvsadcost[1] = &cpi->mb.nmvsadcosts[1][MV_MAX];
  cal_nmvsadcosts(cpi->mb.nmvsadcost);

  cpi->mb.nmvcost_hp[0] = &cpi->mb.nmvcosts_hp[0][MV_MAX];
  cpi->mb.nmvcost_hp[1] = &cpi->mb.nmvcosts_hp[1][MV_MAX];
  cpi->mb.nmvsadcost_hp[0] = &cpi->mb.nmvsadcosts_hp[0][MV_MAX];
  cpi->mb.nmvsadcost_hp[1] = &cpi->mb.nmvsadcosts_hp[1][MV_MAX];
  cal_nmvsadcosts_hp(cpi->mb.nmvsadcost_hp);

  for (i = 0; i < KEY_FRAME_CONTEXT; i++) {
    cpi->prior_key_frame_distance[i] = (int)cpi->output_frame_rate;
  }

#ifdef OUTPUT_YUV_SRC
  yuv_file = fopen("bd.yuv", "ab");
#endif
#ifdef OUTPUT_YUV_REC
  yuv_rec_file = fopen("rec.yuv", "wb");
#endif

#if 0
  framepsnr = fopen("framepsnr.stt", "a");
  kf_list = fopen("kf_list.stt", "w");
#endif

  cpi->output_pkt_list = oxcf->output_pkt_list;

  if (cpi->pass == 1) {
    vp9_init_first_pass(cpi);
  } else if (cpi->pass == 2) {
    size_t packet_sz = sizeof(FIRSTPASS_STATS);
    int packets = oxcf->two_pass_stats_in.sz / packet_sz;

    cpi->twopass.stats_in_start = oxcf->two_pass_stats_in.buf;
    cpi->twopass.stats_in = cpi->twopass.stats_in_start;
    cpi->twopass.stats_in_end = (void *)((char *)cpi->twopass.stats_in
                                         + (packets - 1) * packet_sz);
    vp9_init_second_pass(cpi);
  }

  vp9_set_speed_features(cpi);

  // Set starting values of RD threshold multipliers (128 = *1)
  for (i = 0; i < MAX_MODES; i++) {
    cpi->rd_thresh_mult[i] = 128;
  }

#ifdef ENTROPY_STATS
  init_mv_ref_counts();
#endif

#define BFP(BT, SDF, VF, SVF, SVFHH, SVFHV, SVFHHV, SDX3F, SDX8F, SDX4DF) \
    cpi->fn_ptr[BT].sdf            = SDF; \
    cpi->fn_ptr[BT].vf             = VF; \
    cpi->fn_ptr[BT].svf            = SVF; \
    cpi->fn_ptr[BT].svf_halfpix_h  = SVFHH; \
    cpi->fn_ptr[BT].svf_halfpix_v  = SVFHV; \
    cpi->fn_ptr[BT].svf_halfpix_hv = SVFHHV; \
    cpi->fn_ptr[BT].sdx3f          = SDX3F; \
    cpi->fn_ptr[BT].sdx8f          = SDX8F; \
    cpi->fn_ptr[BT].sdx4df         = SDX4DF;


#if CONFIG_SUPERBLOCKS
  BFP(BLOCK_32X32, vp9_sad32x32, vp9_variance32x32, vp9_sub_pixel_variance32x32,
      vp9_variance_halfpixvar32x32_h, vp9_variance_halfpixvar32x32_v,
      vp9_variance_halfpixvar32x32_hv, vp9_sad32x32x3, vp9_sad32x32x8,
      vp9_sad32x32x4d)
#endif

  BFP(BLOCK_16X16, vp9_sad16x16, vp9_variance16x16, vp9_sub_pixel_variance16x16,
       vp9_variance_halfpixvar16x16_h, vp9_variance_halfpixvar16x16_v,
       vp9_variance_halfpixvar16x16_hv, vp9_sad16x16x3, vp9_sad16x16x8,
       vp9_sad16x16x4d)

  BFP(BLOCK_16X8, vp9_sad16x8, vp9_variance16x8, vp9_sub_pixel_variance16x8,
      NULL, NULL, NULL, vp9_sad16x8x3, vp9_sad16x8x8, vp9_sad16x8x4d)

  BFP(BLOCK_8X16, vp9_sad8x16, vp9_variance8x16, vp9_sub_pixel_variance8x16,
      NULL, NULL, NULL, vp9_sad8x16x3, vp9_sad8x16x8, vp9_sad8x16x4d)

  BFP(BLOCK_8X8, vp9_sad8x8, vp9_variance8x8, vp9_sub_pixel_variance8x8,
      NULL, NULL, NULL, vp9_sad8x8x3, vp9_sad8x8x8, vp9_sad8x8x4d)

  BFP(BLOCK_4X4, vp9_sad4x4, vp9_variance4x4, vp9_sub_pixel_variance4x4,
      NULL, NULL, NULL, vp9_sad4x4x3, vp9_sad4x4x8, vp9_sad4x4x4d)

#if ARCH_X86 || ARCH_X86_64
  cpi->fn_ptr[BLOCK_16X16].copymem  = vp9_copy32xn;
  cpi->fn_ptr[BLOCK_16X8].copymem   = vp9_copy32xn;
  cpi->fn_ptr[BLOCK_8X16].copymem   = vp9_copy32xn;
  cpi->fn_ptr[BLOCK_8X8].copymem    = vp9_copy32xn;
  cpi->fn_ptr[BLOCK_4X4].copymem    = vp9_copy32xn;
#endif

  cpi->full_search_sad = SEARCH_INVOKE(&cpi->rtcd.search, full_search);
  cpi->diamond_search_sad = SEARCH_INVOKE(&cpi->rtcd.search, diamond_search);
  cpi->refining_search_sad = SEARCH_INVOKE(&cpi->rtcd.search, refining_search);

  // make sure frame 1 is okay
  cpi->error_bins[0] = cpi->common.MBs;

  /* vp9_init_quantizer() is first called here. Add check in
   * vp9_frame_init_quantizer() so that vp9_init_quantizer is only
   * called later when needed. This will avoid unnecessary calls of
   * vp9_init_quantizer() for every frame.
   */
  vp9_init_quantizer(cpi);

  vp9_loop_filter_init(cm);

  cpi->common.error.setjmp = 0;

  vp9_zero(cpi->y_uv_mode_count)

  return (VP9_PTR) cpi;
}

void vp9_remove_compressor(VP9_PTR *ptr) {
  VP9_COMP *cpi = (VP9_COMP *)(*ptr);
  int i;

  if (!cpi)
    return;

  if (cpi && (cpi->common.current_video_frame > 0)) {
    if (cpi->pass == 2) {
      vp9_end_second_pass(cpi);
    }

#ifdef ENTROPY_STATS
    if (cpi->pass != 1) {
      print_context_counters();
      print_tree_update_probs();
      print_mode_context();
    }
#endif
#ifdef NMV_STATS
    if (cpi->pass != 1)
      print_nmvstats();
#endif

#if CONFIG_INTERNAL_STATS

    vp9_clear_system_state();

    // printf("\n8x8-4x4:%d-%d\n", cpi->t8x8_count, cpi->t4x4_count);
    if (cpi->pass != 1) {
      FILE *f = fopen("opsnr.stt", "a");
      double time_encoded = (cpi->last_end_time_stamp_seen
                             - cpi->first_time_stamp_ever) / 10000000.000;
      double total_encode_time = (cpi->time_receive_data + cpi->time_compress_data)   / 1000.000;
      double dr = (double)cpi->bytes * (double) 8 / (double)1000  / time_encoded;
#if defined(MODE_STATS)
      print_mode_contexts(&cpi->common);
#endif
      if (cpi->b_calculate_psnr) {
        YV12_BUFFER_CONFIG *lst_yv12 = &cpi->common.yv12_fb[cpi->common.lst_fb_idx];
        double samples = 3.0 / 2 * cpi->count * lst_yv12->y_width * lst_yv12->y_height;
        double total_psnr = vp9_mse2psnr(samples, 255.0, cpi->total_sq_error);
        double total_psnr2 = vp9_mse2psnr(samples, 255.0, cpi->total_sq_error2);
        double total_ssim = 100 * pow(cpi->summed_quality / cpi->summed_weights, 8.0);

        fprintf(f, "Bitrate\tAVGPsnr\tGLBPsnr\tAVPsnrP\tGLPsnrP\tVPXSSIM\t  Time(ms)\n");
        fprintf(f, "%7.2f\t%7.3f\t%7.3f\t%7.3f\t%7.3f\t%7.3f\t%8.0f\n",
                dr, cpi->total / cpi->count, total_psnr, cpi->totalp / cpi->count, total_psnr2, total_ssim,
                total_encode_time);
//                fprintf(f, "%7.3f\t%7.3f\t%7.3f\t%7.3f\t%7.3f\t%7.3f\t%8.0f %10ld\n",
//                        dr, cpi->total / cpi->count, total_psnr, cpi->totalp / cpi->count, total_psnr2, total_ssim,
//                        total_encode_time, cpi->tot_recode_hits);
      }

      if (cpi->b_calculate_ssimg) {
        fprintf(f, "BitRate\tSSIM_Y\tSSIM_U\tSSIM_V\tSSIM_A\t  Time(ms)\n");
        fprintf(f, "%7.2f\t%6.4f\t%6.4f\t%6.4f\t%6.4f\t%8.0f\n", dr,
                cpi->total_ssimg_y / cpi->count, cpi->total_ssimg_u / cpi->count,
                cpi->total_ssimg_v / cpi->count, cpi->total_ssimg_all / cpi->count, total_encode_time);
//                fprintf(f, "%7.3f\t%6.4f\t%6.4f\t%6.4f\t%6.4f\t%8.0f  %10ld\n", dr,
//                        cpi->total_ssimg_y / cpi->count, cpi->total_ssimg_u / cpi->count,
//                        cpi->total_ssimg_v / cpi->count, cpi->total_ssimg_all / cpi->count, total_encode_time, cpi->tot_recode_hits);
      }

      fclose(f);
    }

#endif


#ifdef MODE_STATS
    {
      extern int count_mb_seg[4];
      char modes_stats_file[250];
      FILE *f;
      double dr = (double)cpi->oxcf.frame_rate * (double)cpi->bytes * (double)8 / (double)cpi->count / (double)1000;
      sprintf(modes_stats_file, "modes_q%03d.stt", cpi->common.base_qindex);
      f = fopen(modes_stats_file, "w");
      fprintf(f, "intra_mode in Intra Frames:\n");
      {
        int i;
        fprintf(f, "Y: ");
        for (i = 0; i < VP9_YMODES; i++) fprintf(f, " %8d,", y_modes[i]);
        fprintf(f, "\n");
      }
      {
        int i;
        fprintf(f, "I8: ");
        for (i = 0; i < VP9_I8X8_MODES; i++) fprintf(f, " %8d,", i8x8_modes[i]);
        fprintf(f, "\n");
      }
      {
        int i;
        fprintf(f, "UV: ");
        for (i = 0; i < VP9_UV_MODES; i++) fprintf(f, " %8d,", uv_modes[i]);
        fprintf(f, "\n");
      }
      {
        int i, j;
        fprintf(f, "KeyFrame Y-UV:\n");
        for (i = 0; i < VP9_YMODES; i++) {
          fprintf(f, "%2d:", i);
          for (j = 0; j < VP9_UV_MODES; j++) fprintf(f, "%8d, ", uv_modes_y[i][j]);
          fprintf(f, "\n");
        }
      }
      {
        int i, j;
        fprintf(f, "Inter Y-UV:\n");
        for (i = 0; i < VP9_YMODES; i++) {
          fprintf(f, "%2d:", i);
          for (j = 0; j < VP9_UV_MODES; j++) fprintf(f, "%8d, ", cpi->y_uv_mode_count[i][j]);
          fprintf(f, "\n");
        }
      }
      {
        int i;

        fprintf(f, "B: ");
        for (i = 0; i < VP9_BINTRAMODES; i++)
          fprintf(f, "%8d, ", b_modes[i]);

        fprintf(f, "\n");

      }

      fprintf(f, "Modes in Inter Frames:\n");
      {
        int i;
        fprintf(f, "Y: ");
        for (i = 0; i < MB_MODE_COUNT; i++) fprintf(f, " %8d,", inter_y_modes[i]);
        fprintf(f, "\n");
      }
      {
        int i;
        fprintf(f, "UV: ");
        for (i = 0; i < VP9_UV_MODES; i++) fprintf(f, " %8d,", inter_uv_modes[i]);
        fprintf(f, "\n");
      }
      {
        int i;
        fprintf(f, "B: ");
        for (i = 0; i < B_MODE_COUNT; i++) fprintf(f, "%8d, ", inter_b_modes[i]);
        fprintf(f, "\n");
      }
      fprintf(f, "P:%8d, %8d, %8d, %8d\n", count_mb_seg[0], count_mb_seg[1], count_mb_seg[2], count_mb_seg[3]);
      fprintf(f, "PB:%8d, %8d, %8d, %8d\n", inter_b_modes[LEFT4X4], inter_b_modes[ABOVE4X4], inter_b_modes[ZERO4X4], inter_b_modes[NEW4X4]);
      fclose(f);
    }
#endif

#ifdef ENTROPY_STATS
    {
      int i, j, k;
      FILE *fmode = fopen("modecontext.c", "w");

      fprintf(fmode, "\n#include \"entropymode.h\"\n\n");
      fprintf(fmode, "const unsigned int vp9_kf_default_bmode_counts ");
      fprintf(fmode, "[VP9_BINTRAMODES] [VP9_BINTRAMODES] [VP9_BINTRAMODES] =\n{\n");

      for (i = 0; i < 10; i++) {

        fprintf(fmode, "    { // Above Mode :  %d\n", i);

        for (j = 0; j < 10; j++) {

          fprintf(fmode, "        {");

          for (k = 0; k < VP9_BINTRAMODES; k++) {
            if (!intra_mode_stats[i][j][k])
              fprintf(fmode, " %5d, ", 1);
            else
              fprintf(fmode, " %5d, ", intra_mode_stats[i][j][k]);
          }

          fprintf(fmode, "}, // left_mode %d\n", j);

        }

        fprintf(fmode, "    },\n");

      }

      fprintf(fmode, "};\n");
      fclose(fmode);
    }
#endif


#if defined(SECTIONBITS_OUTPUT)

    if (0) {
      int i;
      FILE *f = fopen("tokenbits.stt", "a");

      for (i = 0; i < 28; i++)
        fprintf(f, "%8d", (int)(Sectionbits[i] / 256));

      fprintf(f, "\n");
      fclose(f);
    }

#endif

#if 0
    {
      printf("\n_pick_loop_filter_level:%d\n", cpi->time_pick_lpf / 1000);
      printf("\n_frames recive_data encod_mb_row compress_frame  Total\n");
      printf("%6d %10ld %10ld %10ld %10ld\n", cpi->common.current_video_frame, cpi->time_receive_data / 1000, cpi->time_encode_mb_row / 1000, cpi->time_compress_data / 1000, (cpi->time_receive_data + cpi->time_compress_data) / 1000);
    }
#endif

  }

  dealloc_compressor_data(cpi);
  vpx_free(cpi->mb.ss);
  vpx_free(cpi->tok);

  for (i = 0; i < sizeof(cpi->mbgraph_stats) / sizeof(cpi->mbgraph_stats[0]); i++) {
    vpx_free(cpi->mbgraph_stats[i].mb_stats);
  }

  vp9_remove_common(&cpi->common);
  vpx_free(cpi);
  *ptr = 0;

#ifdef OUTPUT_YUV_SRC
  fclose(yuv_file);
#endif
#ifdef OUTPUT_YUV_REC
  fclose(yuv_rec_file);
#endif

#if 0

  if (keyfile)
    fclose(keyfile);

  if (framepsnr)
    fclose(framepsnr);

  if (kf_list)
    fclose(kf_list);

#endif

}


static uint64_t calc_plane_error(unsigned char *orig, int orig_stride,
                                 unsigned char *recon, int recon_stride,
                                 unsigned int cols, unsigned int rows) {
  unsigned int row, col;
  uint64_t total_sse = 0;
  int diff;

  for (row = 0; row + 16 <= rows; row += 16) {
    for (col = 0; col + 16 <= cols; col += 16) {
      unsigned int sse;

      vp9_mse16x16(orig + col, orig_stride, recon + col, recon_stride, &sse);
      total_sse += sse;
    }

    /* Handle odd-sized width */
    if (col < cols) {
      unsigned int   border_row, border_col;
      unsigned char *border_orig = orig;
      unsigned char *border_recon = recon;

      for (border_row = 0; border_row < 16; border_row++) {
        for (border_col = col; border_col < cols; border_col++) {
          diff = border_orig[border_col] - border_recon[border_col];
          total_sse += diff * diff;
        }

        border_orig += orig_stride;
        border_recon += recon_stride;
      }
    }

    orig += orig_stride * 16;
    recon += recon_stride * 16;
  }

  /* Handle odd-sized height */
  for (; row < rows; row++) {
    for (col = 0; col < cols; col++) {
      diff = orig[col] - recon[col];
      total_sse += diff * diff;
    }

    orig += orig_stride;
    recon += recon_stride;
  }

  return total_sse;
}


static void generate_psnr_packet(VP9_COMP *cpi) {
  YV12_BUFFER_CONFIG      *orig = cpi->Source;
  YV12_BUFFER_CONFIG      *recon = cpi->common.frame_to_show;
  struct vpx_codec_cx_pkt  pkt;
  uint64_t                 sse;
  int                      i;
  unsigned int             width = cpi->common.Width;
  unsigned int             height = cpi->common.Height;

  pkt.kind = VPX_CODEC_PSNR_PKT;
  sse = calc_plane_error(orig->y_buffer, orig->y_stride,
                         recon->y_buffer, recon->y_stride,
                         width, height);
  pkt.data.psnr.sse[0] = sse;
  pkt.data.psnr.sse[1] = sse;
  pkt.data.psnr.samples[0] = width * height;
  pkt.data.psnr.samples[1] = width * height;

  width = (width + 1) / 2;
  height = (height + 1) / 2;

  sse = calc_plane_error(orig->u_buffer, orig->uv_stride,
                         recon->u_buffer, recon->uv_stride,
                         width, height);
  pkt.data.psnr.sse[0] += sse;
  pkt.data.psnr.sse[2] = sse;
  pkt.data.psnr.samples[0] += width * height;
  pkt.data.psnr.samples[2] = width * height;

  sse = calc_plane_error(orig->v_buffer, orig->uv_stride,
                         recon->v_buffer, recon->uv_stride,
                         width, height);
  pkt.data.psnr.sse[0] += sse;
  pkt.data.psnr.sse[3] = sse;
  pkt.data.psnr.samples[0] += width * height;
  pkt.data.psnr.samples[3] = width * height;

  for (i = 0; i < 4; i++)
    pkt.data.psnr.psnr[i] = vp9_mse2psnr(pkt.data.psnr.samples[i], 255.0,
                                         pkt.data.psnr.sse[i]);

  vpx_codec_pkt_list_add(cpi->output_pkt_list, &pkt);
}


int vp9_use_as_reference(VP9_PTR ptr, int ref_frame_flags) {
  VP9_COMP *cpi = (VP9_COMP *)(ptr);

  if (ref_frame_flags > 7)
    return -1;

  cpi->ref_frame_flags = ref_frame_flags;
  return 0;
}
int vp9_update_reference(VP9_PTR ptr, int ref_frame_flags) {
  VP9_COMP *cpi = (VP9_COMP *)(ptr);

  if (ref_frame_flags > 7)
    return -1;

  cpi->common.refresh_golden_frame = 0;
  cpi->common.refresh_alt_ref_frame = 0;
  cpi->common.refresh_last_frame   = 0;

  if (ref_frame_flags & VP9_LAST_FLAG)
    cpi->common.refresh_last_frame = 1;

  if (ref_frame_flags & VP9_GOLD_FLAG)
    cpi->common.refresh_golden_frame = 1;

  if (ref_frame_flags & VP9_ALT_FLAG)
    cpi->common.refresh_alt_ref_frame = 1;

  return 0;
}

int vp9_get_reference_enc(VP9_PTR ptr, VP9_REFFRAME ref_frame_flag,
                          YV12_BUFFER_CONFIG *sd) {
  VP9_COMP *cpi = (VP9_COMP *)(ptr);
  VP9_COMMON *cm = &cpi->common;
  int ref_fb_idx;

  if (ref_frame_flag == VP9_LAST_FLAG)
    ref_fb_idx = cm->lst_fb_idx;
  else if (ref_frame_flag == VP9_GOLD_FLAG)
    ref_fb_idx = cm->gld_fb_idx;
  else if (ref_frame_flag == VP9_ALT_FLAG)
    ref_fb_idx = cm->alt_fb_idx;
  else
    return -1;

  vp8_yv12_copy_frame(&cm->yv12_fb[ref_fb_idx], sd);

  return 0;
}

int vp9_set_reference_enc(VP9_PTR ptr, VP9_REFFRAME ref_frame_flag,
                          YV12_BUFFER_CONFIG *sd) {
  VP9_COMP *cpi = (VP9_COMP *)(ptr);
  VP9_COMMON *cm = &cpi->common;

  int ref_fb_idx;

  if (ref_frame_flag == VP9_LAST_FLAG)
    ref_fb_idx = cm->lst_fb_idx;
  else if (ref_frame_flag == VP9_GOLD_FLAG)
    ref_fb_idx = cm->gld_fb_idx;
  else if (ref_frame_flag == VP9_ALT_FLAG)
    ref_fb_idx = cm->alt_fb_idx;
  else
    return -1;

  vp8_yv12_copy_frame(sd, &cm->yv12_fb[ref_fb_idx]);

  return 0;
}
int vp9_update_entropy(VP9_PTR comp, int update) {
  VP9_COMP *cpi = (VP9_COMP *) comp;
  VP9_COMMON *cm = &cpi->common;
  cm->refresh_entropy_probs = update;

  return 0;
}


#ifdef OUTPUT_YUV_SRC
void vp9_write_yuv_frame(YV12_BUFFER_CONFIG *s) {
  unsigned char *src = s->y_buffer;
  int h = s->y_height;

  do {
    fwrite(src, s->y_width, 1,  yuv_file);
    src += s->y_stride;
  } while (--h);

  src = s->u_buffer;
  h = s->uv_height;

  do {
    fwrite(src, s->uv_width, 1,  yuv_file);
    src += s->uv_stride;
  } while (--h);

  src = s->v_buffer;
  h = s->uv_height;

  do {
    fwrite(src, s->uv_width, 1, yuv_file);
    src += s->uv_stride;
  } while (--h);
}
#endif

#ifdef OUTPUT_YUV_REC
void vp9_write_yuv_rec_frame(VP9_COMMON *cm) {
  YV12_BUFFER_CONFIG *s = cm->frame_to_show;
  unsigned char *src = s->y_buffer;
  int h = cm->Height;

  do {
    fwrite(src, s->y_width, 1,  yuv_rec_file);
    src += s->y_stride;
  } while (--h);

  src = s->u_buffer;
  h = (cm->Height + 1) / 2;

  do {
    fwrite(src, s->uv_width, 1,  yuv_rec_file);
    src += s->uv_stride;
  } while (--h);

  src = s->v_buffer;
  h = (cm->Height + 1) / 2;

  do {
    fwrite(src, s->uv_width, 1, yuv_rec_file);
    src += s->uv_stride;
  } while (--h);
}
#endif

static void update_alt_ref_frame_stats(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;

  // Update data structure that monitors level of reference to last GF
  vpx_memset(cpi->gf_active_flags, 1, (cm->mb_rows * cm->mb_cols));
  cpi->gf_active_count = cm->mb_rows * cm->mb_cols;

  // this frame refreshes means next frames don't unless specified by user
  cpi->common.frames_since_golden = 0;

  // Clear the alternate reference update pending flag.
  cpi->source_alt_ref_pending = FALSE;

  // Set the alternate refernce frame active flag
  cpi->source_alt_ref_active = TRUE;


}
static void update_golden_frame_stats(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;

  // Update the Golden frame usage counts.
  if (cm->refresh_golden_frame) {
    // Update data structure that monitors level of reference to last GF
    vpx_memset(cpi->gf_active_flags, 1, (cm->mb_rows * cm->mb_cols));
    cpi->gf_active_count = cm->mb_rows * cm->mb_cols;

    // this frame refreshes means next frames don't unless specified by user
    cm->refresh_golden_frame = 0;
    cpi->common.frames_since_golden = 0;

    // if ( cm->frame_type == KEY_FRAME )
    // {
    cpi->recent_ref_frame_usage[INTRA_FRAME] = 1;
    cpi->recent_ref_frame_usage[LAST_FRAME] = 1;
    cpi->recent_ref_frame_usage[GOLDEN_FRAME] = 1;
    cpi->recent_ref_frame_usage[ALTREF_FRAME] = 1;
    // }
    // else
    // {
    //  // Carry a potrtion of count over to begining of next gf sequence
    //  cpi->recent_ref_frame_usage[INTRA_FRAME] >>= 5;
    //  cpi->recent_ref_frame_usage[LAST_FRAME] >>= 5;
    //  cpi->recent_ref_frame_usage[GOLDEN_FRAME] >>= 5;
    //  cpi->recent_ref_frame_usage[ALTREF_FRAME] >>= 5;
    // }

    // ******** Fixed Q test code only ************
    // If we are going to use the ALT reference for the next group of frames set a flag to say so.
    if (cpi->oxcf.fixed_q >= 0 &&
        cpi->oxcf.play_alternate && !cpi->common.refresh_alt_ref_frame) {
      cpi->source_alt_ref_pending = TRUE;
      cpi->frames_till_gf_update_due = cpi->baseline_gf_interval;
    }

    if (!cpi->source_alt_ref_pending)
      cpi->source_alt_ref_active = FALSE;

    // Decrement count down till next gf
    if (cpi->frames_till_gf_update_due > 0)
      cpi->frames_till_gf_update_due--;

  } else if (!cpi->common.refresh_alt_ref_frame) {
    // Decrement count down till next gf
    if (cpi->frames_till_gf_update_due > 0)
      cpi->frames_till_gf_update_due--;

    if (cpi->common.frames_till_alt_ref_frame)
      cpi->common.frames_till_alt_ref_frame--;

    cpi->common.frames_since_golden++;

    if (cpi->common.frames_since_golden > 1) {
      cpi->recent_ref_frame_usage[INTRA_FRAME] += cpi->count_mb_ref_frame_usage[INTRA_FRAME];
      cpi->recent_ref_frame_usage[LAST_FRAME] += cpi->count_mb_ref_frame_usage[LAST_FRAME];
      cpi->recent_ref_frame_usage[GOLDEN_FRAME] += cpi->count_mb_ref_frame_usage[GOLDEN_FRAME];
      cpi->recent_ref_frame_usage[ALTREF_FRAME] += cpi->count_mb_ref_frame_usage[ALTREF_FRAME];
    }
  }
}

static int find_fp_qindex() {
  int i;

  for (i = 0; i < QINDEX_RANGE; i++) {
    if (vp9_convert_qindex_to_q(i) >= 30.0) {
      break;
    }
  }

  if (i == QINDEX_RANGE)
    i--;

  return i;
}

static void Pass1Encode(VP9_COMP *cpi, unsigned long *size, unsigned char *dest, unsigned int *frame_flags) {
  (void) size;
  (void) dest;
  (void) frame_flags;


  vp9_set_quantizer(cpi, find_fp_qindex());
  vp9_first_pass(cpi);
}

#define WRITE_RECON_BUFFER 0
#if WRITE_RECON_BUFFER
void write_cx_frame_to_file(YV12_BUFFER_CONFIG *frame, int this_frame) {

  // write the frame
  FILE *yframe;
  int i;
  char filename[255];

  sprintf(filename, "cx\\y%04d.raw", this_frame);
  yframe = fopen(filename, "wb");

  for (i = 0; i < frame->y_height; i++)
    fwrite(frame->y_buffer + i * frame->y_stride,
           frame->y_width, 1, yframe);

  fclose(yframe);
  sprintf(filename, "cx\\u%04d.raw", this_frame);
  yframe = fopen(filename, "wb");

  for (i = 0; i < frame->uv_height; i++)
    fwrite(frame->u_buffer + i * frame->uv_stride,
           frame->uv_width, 1, yframe);

  fclose(yframe);
  sprintf(filename, "cx\\v%04d.raw", this_frame);
  yframe = fopen(filename, "wb");

  for (i = 0; i < frame->uv_height; i++)
    fwrite(frame->v_buffer + i * frame->uv_stride,
           frame->uv_width, 1, yframe);

  fclose(yframe);
}
#endif

static double compute_edge_pixel_proportion(YV12_BUFFER_CONFIG *frame) {
#define EDGE_THRESH 128
  int i, j;
  int num_edge_pels = 0;
  int num_pels = (frame->y_height - 2) * (frame->y_width - 2);
  unsigned char *prev = frame->y_buffer + 1;
  unsigned char *curr = frame->y_buffer + 1 + frame->y_stride;
  unsigned char *next = frame->y_buffer + 1 + 2 * frame->y_stride;
  for (i = 1; i < frame->y_height - 1; i++) {
    for (j = 1; j < frame->y_width - 1; j++) {
      /* Sobel hor and ver gradients */
      int v = 2 * (curr[1] - curr[-1]) + (prev[1] - prev[-1]) + (next[1] - next[-1]);
      int h = 2 * (prev[0] - next[0]) + (prev[1] - next[1]) + (prev[-1] - next[-1]);
      h = (h < 0 ? -h : h);
      v = (v < 0 ? -v : v);
      if (h > EDGE_THRESH || v > EDGE_THRESH) num_edge_pels++;
      curr++;
      prev++;
      next++;
    }
    curr += frame->y_stride - frame->y_width + 2;
    prev += frame->y_stride - frame->y_width + 2;
    next += frame->y_stride - frame->y_width + 2;
  }
  return (double)num_edge_pels / (double)num_pels;
}

// Function to test for conditions that indicate we should loop
// back and recode a frame.
static BOOL recode_loop_test(VP9_COMP *cpi,
                             int high_limit, int low_limit,
                             int q, int maxq, int minq) {
  BOOL    force_recode = FALSE;
  VP9_COMMON *cm = &cpi->common;

  // Is frame recode allowed at all
  // Yes if either recode mode 1 is selected or mode two is selcted
  // and the frame is a key frame. golden frame or alt_ref_frame
  if ((cpi->sf.recode_loop == 1) ||
      ((cpi->sf.recode_loop == 2) &&
       ((cm->frame_type == KEY_FRAME) ||
        cm->refresh_golden_frame ||
        cm->refresh_alt_ref_frame))) {
    // General over and under shoot tests
    if (((cpi->projected_frame_size > high_limit) && (q < maxq)) ||
        ((cpi->projected_frame_size < low_limit) && (q > minq))) {
      force_recode = TRUE;
    }
    // Special Constrained quality tests
    else if (cpi->oxcf.end_usage == USAGE_CONSTRAINED_QUALITY) {
      // Undershoot and below auto cq level
      if ((q > cpi->cq_target_quality) &&
          (cpi->projected_frame_size <
           ((cpi->this_frame_target * 7) >> 3))) {
        force_recode = TRUE;
      }
      // Severe undershoot and between auto and user cq level
      else if ((q > cpi->oxcf.cq_level) &&
               (cpi->projected_frame_size < cpi->min_frame_bandwidth) &&
               (cpi->active_best_quality > cpi->oxcf.cq_level)) {
        force_recode = TRUE;
        cpi->active_best_quality = cpi->oxcf.cq_level;
      }
    }
  }

  return force_recode;
}

static void update_reference_frames(VP9_COMMON *cm) {
  YV12_BUFFER_CONFIG *yv12_fb = cm->yv12_fb;

  // At this point the new frame has been encoded.
  // If any buffer copy / swapping is signaled it should be done here.

  if (cm->frame_type == KEY_FRAME) {
    yv12_fb[cm->new_fb_idx].flags |= VP9_GOLD_FLAG | VP9_ALT_FLAG;

    yv12_fb[cm->gld_fb_idx].flags &= ~VP9_GOLD_FLAG;
    yv12_fb[cm->alt_fb_idx].flags &= ~VP9_ALT_FLAG;

    cm->alt_fb_idx = cm->gld_fb_idx = cm->new_fb_idx;
  } else { /* For non key frames */
    if (cm->refresh_alt_ref_frame) {
      assert(!cm->copy_buffer_to_arf);

      cm->yv12_fb[cm->new_fb_idx].flags |= VP9_ALT_FLAG;
      cm->yv12_fb[cm->alt_fb_idx].flags &= ~VP9_ALT_FLAG;
      cm->alt_fb_idx = cm->new_fb_idx;
    } else if (cm->copy_buffer_to_arf) {
      assert(!(cm->copy_buffer_to_arf & ~0x3));

      if (cm->copy_buffer_to_arf == 1) {
        if (cm->alt_fb_idx != cm->lst_fb_idx) {
          yv12_fb[cm->lst_fb_idx].flags |= VP9_ALT_FLAG;
          yv12_fb[cm->alt_fb_idx].flags &= ~VP9_ALT_FLAG;
          cm->alt_fb_idx = cm->lst_fb_idx;
        }
      } else { /* if (cm->copy_buffer_to_arf == 2) */
        if (cm->alt_fb_idx != cm->gld_fb_idx) {
          yv12_fb[cm->gld_fb_idx].flags |= VP9_ALT_FLAG;
          yv12_fb[cm->alt_fb_idx].flags &= ~VP9_ALT_FLAG;
          cm->alt_fb_idx = cm->gld_fb_idx;
        }
      }
    }

    if (cm->refresh_golden_frame) {
      assert(!cm->copy_buffer_to_gf);

      cm->yv12_fb[cm->new_fb_idx].flags |= VP9_GOLD_FLAG;
      cm->yv12_fb[cm->gld_fb_idx].flags &= ~VP9_GOLD_FLAG;
      cm->gld_fb_idx = cm->new_fb_idx;
    } else if (cm->copy_buffer_to_gf) {
      assert(!(cm->copy_buffer_to_arf & ~0x3));

      if (cm->copy_buffer_to_gf == 1) {
        if (cm->gld_fb_idx != cm->lst_fb_idx) {
          yv12_fb[cm->lst_fb_idx].flags |= VP9_GOLD_FLAG;
          yv12_fb[cm->gld_fb_idx].flags &= ~VP9_GOLD_FLAG;
          cm->gld_fb_idx = cm->lst_fb_idx;
        }
      } else { /* if (cm->copy_buffer_to_gf == 2) */
        if (cm->alt_fb_idx != cm->gld_fb_idx) {
          yv12_fb[cm->alt_fb_idx].flags |= VP9_GOLD_FLAG;
          yv12_fb[cm->gld_fb_idx].flags &= ~VP9_GOLD_FLAG;
          cm->gld_fb_idx = cm->alt_fb_idx;
        }
      }
    }
  }

  if (cm->refresh_last_frame) {
    cm->yv12_fb[cm->new_fb_idx].flags |= VP9_LAST_FLAG;
    cm->yv12_fb[cm->lst_fb_idx].flags &= ~VP9_LAST_FLAG;
    cm->lst_fb_idx = cm->new_fb_idx;
  }
}

static void loopfilter_frame(VP9_COMP *cpi, VP9_COMMON *cm) {
  if (cm->no_lpf) {
    cm->filter_level = 0;
  }
#if CONFIG_LOSSLESS
  else if (cpi->oxcf.lossless) {
    cm->filter_level = 0;
  }
#endif
  else {
    struct vpx_usec_timer timer;

    vp9_clear_system_state();

    vpx_usec_timer_start(&timer);
    if (cpi->sf.auto_filter == 0)
      vp9_pick_filter_level_fast(cpi->Source, cpi);
    else
      vp9_pick_filter_level(cpi->Source, cpi);

    vpx_usec_timer_mark(&timer);
    cpi->time_pick_lpf += vpx_usec_timer_elapsed(&timer);
  }

  if (cm->filter_level > 0) {
    vp9_set_alt_lf_level(cpi, cm->filter_level);
    vp9_loop_filter_frame(cm, &cpi->mb.e_mbd);
  }

  vp8_yv12_extend_frame_borders(cm->frame_to_show);

}

#if CONFIG_PRED_FILTER
void select_pred_filter_mode(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;

  int prob_pred_filter_off = cm->prob_pred_filter_off;

  // Force filter on/off if probability is extreme
  if (prob_pred_filter_off >= 255 * 0.95)
    cm->pred_filter_mode = 0;   // Off at the frame level
  else if (prob_pred_filter_off <= 255 * 0.05)
    cm->pred_filter_mode = 1;   // On at the frame level
  else
    cm->pred_filter_mode = 2;   // Selectable at the MB level
}

void update_pred_filt_prob(VP9_COMP *cpi) {
  VP9_COMMON *cm = &cpi->common;
  int prob_pred_filter_off;

  // Based on the selection in the previous frame determine what mode
  // to use for the current frame and work out the signaling probability
  if (cpi->pred_filter_on_count + cpi->pred_filter_off_count) {
    prob_pred_filter_off = cpi->pred_filter_off_count * 256 /
                           (cpi->pred_filter_on_count + cpi->pred_filter_off_count);

    if (prob_pred_filter_off < 1)
      prob_pred_filter_off = 1;

    if (prob_pred_filter_off > 255)
      prob_pred_filter_off = 255;

    cm->prob_pred_filter_off = prob_pred_filter_off;
  } else
    cm->prob_pred_filter_off = 128;
  /*
      {
        FILE *fp = fopen("filt_use.txt", "a");
        fprintf (fp, "%d %d prob=%d\n", cpi->pred_filter_off_count,
                 cpi->pred_filter_on_count, cm->prob_pred_filter_off);
        fclose(fp);
      }
  */
}
#endif

static void encode_frame_to_data_rate
(
  VP9_COMP *cpi,
  unsigned long *size,
  unsigned char *dest,
  unsigned int *frame_flags
) {
  VP9_COMMON *cm = &cpi->common;
  MACROBLOCKD *xd = &cpi->mb.e_mbd;

  int Q;
  int frame_over_shoot_limit;
  int frame_under_shoot_limit;

  int Loop = FALSE;
  int loop_count;
  int this_q;
  int last_zbin_oq;

  int q_low;
  int q_high;
  int zbin_oq_high;
  int zbin_oq_low = 0;

  int top_index;
  int bottom_index;
  int active_worst_qchanged = FALSE;

  int overshoot_seen = FALSE;
  int undershoot_seen = FALSE;

  int loop_size_estimate = 0;

  SPEED_FEATURES *sf = &cpi->sf;
#if RESET_FOREACH_FILTER
  int q_low0;
  int q_high0;
  int zbin_oq_high0;
  int zbin_oq_low0 = 0;
  int Q0;
  int last_zbin_oq0;
  int active_best_quality0;
  int active_worst_quality0;
  double rate_correction_factor0;
  double gf_rate_correction_factor0;
#endif

  /* list of filters to search over */
  int mcomp_filters_to_search[] = {
    EIGHTTAP, EIGHTTAP_SHARP, SIXTAP, SWITCHABLE
  };
  int mcomp_filters = sizeof(mcomp_filters_to_search) /
      sizeof(*mcomp_filters_to_search);
  int mcomp_filter_index = 0;
  INT64 mcomp_filter_cost[4];

  // Clear down mmx registers to allow floating point in what follows
  vp9_clear_system_state();


  // For an alt ref frame in 2 pass we skip the call to the second
  // pass function that sets the target bandwidth so must set it here
  if (cpi->common.refresh_alt_ref_frame) {
    cpi->per_frame_bandwidth = cpi->twopass.gf_bits;                           // Per frame bit target for the alt ref frame
    cpi->target_bandwidth = cpi->twopass.gf_bits * cpi->output_frame_rate;      // per second target bitrate
  }

  // Default turn off buffer to buffer copying
  cm->copy_buffer_to_gf = 0;
  cm->copy_buffer_to_arf = 0;

  // Clear zbin over-quant value and mode boost values.
  cpi->zbin_over_quant = 0;
  cpi->zbin_mode_boost = 0;

  // Enable or disable mode based tweaking of the zbin
  // For 2 Pass Only used where GF/ARF prediction quality
  // is above a threshold
  cpi->zbin_mode_boost = 0;
#if CONFIG_LOSSLESS
  cpi->zbin_mode_boost_enabled = FALSE;
#else
  cpi->zbin_mode_boost_enabled = TRUE;
#endif
  if (cpi->gfu_boost <= 400) {
    cpi->zbin_mode_boost_enabled = FALSE;
  }

  // Current default encoder behaviour for the altref sign bias
  if (cpi->source_alt_ref_active)
    cpi->common.ref_frame_sign_bias[ALTREF_FRAME] = 1;
  else
    cpi->common.ref_frame_sign_bias[ALTREF_FRAME] = 0;

  // Check to see if a key frame is signalled
  // For two pass with auto key frame enabled cm->frame_type may already be set, but not for one pass.
  if ((cm->current_video_frame == 0) ||
      (cm->frame_flags & FRAMEFLAGS_KEY) ||
      (cpi->oxcf.auto_key && (cpi->frames_since_key % cpi->key_frame_frequency == 0))) {
    // Key frame from VFW/auto-keyframe/first frame
    cm->frame_type = KEY_FRAME;
  }

  // Set default state for segment based loop filter update flags
  xd->mode_ref_lf_delta_update = 0;

  // Set various flags etc to special state if it is a key frame
  if (cm->frame_type == KEY_FRAME) {
    int i;

    // Reset the loop filter deltas and segmentation map
    setup_features(cpi);

    // If segmentation is enabled force a map update for key frames
    if (xd->segmentation_enabled) {
      xd->update_mb_segmentation_map = 1;
      xd->update_mb_segmentation_data = 1;
    }

    // The alternate reference frame cannot be active for a key frame
    cpi->source_alt_ref_active = FALSE;

    // Reset the RD threshold multipliers to default of * 1 (128)
    for (i = 0; i < MAX_MODES; i++) {
      cpi->rd_thresh_mult[i] = 128;
    }
  }

  // Test code for new segment features
  init_seg_features(cpi);

  // Decide how big to make the frame
  vp9_pick_frame_size(cpi);

  vp9_clear_system_state();

  // Set an active best quality and if necessary active worst quality
  Q = cpi->active_worst_quality;

  if (cm->frame_type == KEY_FRAME) {
    int high = 2000;
    int low = 400;

    if (cpi->kf_boost > high)
      cpi->active_best_quality = kf_low_motion_minq[Q];
    else if (cpi->kf_boost < low)
      cpi->active_best_quality = kf_high_motion_minq[Q];
    else {
      int gap = high - low;
      int offset = high - cpi->kf_boost;
      int qdiff = kf_high_motion_minq[Q] - kf_low_motion_minq[Q];
      int adjustment = ((offset * qdiff) + (gap >> 1)) / gap;

      cpi->active_best_quality = kf_low_motion_minq[Q] + adjustment;
    }

    // Make an adjustment based on the %s static
    // The main impact of this is at lower Q to prevent overly large key
    // frames unless a lot of the image is static.
    if (cpi->kf_zeromotion_pct < 64)
      cpi->active_best_quality += 4 - (cpi->kf_zeromotion_pct >> 4);

    // Special case for key frames forced because we have reached
    // the maximum key frame interval. Here force the Q to a range
    // based on the ambient Q to reduce the risk of popping
    if (cpi->this_key_frame_forced) {
      int delta_qindex;
      int qindex = cpi->last_boosted_qindex;

      delta_qindex = compute_qdelta(cpi, qindex,
                                    (qindex * 0.75));

      cpi->active_best_quality = qindex + delta_qindex;
      if (cpi->active_best_quality < cpi->best_quality)
        cpi->active_best_quality = cpi->best_quality;
    }
  }

  else if (cm->refresh_golden_frame || cpi->common.refresh_alt_ref_frame) {
    int high = 2000;
    int low = 400;

    // Use the lower of cpi->active_worst_quality and recent
    // average Q as basis for GF/ARF Q limit unless last frame was
    // a key frame.
    if ((cpi->frames_since_key > 1) &&
        (cpi->avg_frame_qindex < cpi->active_worst_quality)) {
      Q = cpi->avg_frame_qindex;
    }

    // For constrained quality dont allow Q less than the cq level
    if ((cpi->oxcf.end_usage == USAGE_CONSTRAINED_QUALITY) &&
        (Q < cpi->cq_target_quality)) {
      Q = cpi->cq_target_quality;
    }

    if (cpi->gfu_boost > high)
      cpi->active_best_quality = gf_low_motion_minq[Q];
    else if (cpi->gfu_boost < low)
      cpi->active_best_quality = gf_high_motion_minq[Q];
    else {
      int gap = high - low;
      int offset = high - cpi->gfu_boost;
      int qdiff = gf_high_motion_minq[Q] - gf_low_motion_minq[Q];
      int adjustment = ((offset * qdiff) + (gap >> 1)) / gap;

      cpi->active_best_quality = gf_low_motion_minq[Q] + adjustment;
    }

    // Constrained quality use slightly lower active best.
    if (cpi->oxcf.end_usage == USAGE_CONSTRAINED_QUALITY) {
      cpi->active_best_quality =
        cpi->active_best_quality * 15 / 16;
    }
  } else {
    cpi->active_best_quality = inter_minq[Q];

    // For the constant/constrained quality mode we dont want
    // q to fall below the cq level.
    if ((cpi->oxcf.end_usage == USAGE_CONSTRAINED_QUALITY) &&
        (cpi->active_best_quality < cpi->cq_target_quality)) {
      // If we are strongly undershooting the target rate in the last
      // frames then use the user passed in cq value not the auto
      // cq value.
      if (cpi->rolling_actual_bits < cpi->min_frame_bandwidth)
        cpi->active_best_quality = cpi->oxcf.cq_level;
      else
        cpi->active_best_quality = cpi->cq_target_quality;
    }
  }

  // Clip the active best and worst quality values to limits
  if (cpi->active_worst_quality > cpi->worst_quality)
    cpi->active_worst_quality = cpi->worst_quality;

  if (cpi->active_best_quality < cpi->best_quality)
    cpi->active_best_quality = cpi->best_quality;

  if (cpi->active_best_quality > cpi->worst_quality)
    cpi->active_best_quality = cpi->worst_quality;

  if (cpi->active_worst_quality < cpi->active_best_quality)
    cpi->active_worst_quality = cpi->active_best_quality;

  // Specuial case code to try and match quality with forced key frames
  if ((cm->frame_type == KEY_FRAME) && cpi->this_key_frame_forced) {
    Q = cpi->last_boosted_qindex;
  } else {
    // Determine initial Q to try
    Q = vp9_regulate_q(cpi, cpi->this_frame_target);
  }
  last_zbin_oq = cpi->zbin_over_quant;

  // Set highest allowed value for Zbin over quant
  if (cm->frame_type == KEY_FRAME)
    zbin_oq_high = 0; // ZBIN_OQ_MAX/16
  else if (cm->refresh_alt_ref_frame || (cm->refresh_golden_frame && !cpi->source_alt_ref_active))
    zbin_oq_high = 16;
  else
    zbin_oq_high = ZBIN_OQ_MAX;

  vp9_compute_frame_size_bounds(cpi, &frame_under_shoot_limit,
                                &frame_over_shoot_limit);

  // Limit Q range for the adaptive loop.
  bottom_index = cpi->active_best_quality;
  top_index    = cpi->active_worst_quality;
  q_low  = cpi->active_best_quality;
  q_high = cpi->active_worst_quality;

  loop_count = 0;

  if (cm->frame_type != KEY_FRAME) {
    /* TODO: Decide this more intelligently */
    if (sf->search_best_filter) {
      cm->mcomp_filter_type = mcomp_filters_to_search[0];
      mcomp_filter_index = 0;
    } else {
      cm->mcomp_filter_type = DEFAULT_INTERP_FILTER;
    }
    /* TODO: Decide this more intelligently */
    xd->allow_high_precision_mv = (Q < HIGH_PRECISION_MV_QTHRESH);
  }

#if CONFIG_POSTPROC

  if (cpi->oxcf.noise_sensitivity > 0) {
    unsigned char *src;
    int l = 0;

    switch (cpi->oxcf.noise_sensitivity) {
      case 1:
        l = 20;
        break;
      case 2:
        l = 40;
        break;
      case 3:
        l = 60;
        break;
      case 4:

      case 5:
        l = 100;
        break;
      case 6:
        l = 150;
        break;
    }


    if (cm->frame_type == KEY_FRAME) {
      vp9_de_noise(cpi->Source, cpi->Source, l, 1,  0, RTCD(postproc));
    } else {
      vp9_de_noise(cpi->Source, cpi->Source, l, 1,  0, RTCD(postproc));

      src = cpi->Source->y_buffer;

      if (cpi->Source->y_stride < 0) {
        src += cpi->Source->y_stride * (cpi->Source->y_height - 1);
      }
    }
  }

#endif

#ifdef OUTPUT_YUV_SRC
  vp9_write_yuv_frame(cpi->Source);
#endif

#if RESET_FOREACH_FILTER
  if (sf->search_best_filter) {
    q_low0 = q_low;
    q_high0 = q_high;
    Q0 = Q;
    zbin_oq_low0 = zbin_oq_low;
    zbin_oq_high0 = zbin_oq_high;
    last_zbin_oq0 = last_zbin_oq;
    rate_correction_factor0 = cpi->rate_correction_factor;
    gf_rate_correction_factor0 = cpi->gf_rate_correction_factor;
    active_best_quality0 = cpi->active_best_quality;
    active_worst_quality0 = cpi->active_worst_quality;
  }
#endif
  do {
    vp9_clear_system_state();  // __asm emms;

    vp9_set_quantizer(cpi, Q);
    this_q = Q;

    if (loop_count == 0) {

      // setup skip prob for costing in mode/mv decision
      if (cpi->common.mb_no_coeff_skip) {
        int k;
        for (k = 0; k < MBSKIP_CONTEXTS; k++)
          cm->mbskip_pred_probs[k] = cpi->base_skip_false_prob[Q][k];

        if (cm->frame_type != KEY_FRAME) {
          if (cpi->common.refresh_alt_ref_frame) {
            for (k = 0; k < MBSKIP_CONTEXTS; k++) {
              if (cpi->last_skip_false_probs[2][k] != 0)
                cm->mbskip_pred_probs[k] = cpi->last_skip_false_probs[2][k];
            }
          } else if (cpi->common.refresh_golden_frame) {
            for (k = 0; k < MBSKIP_CONTEXTS; k++) {
              if (cpi->last_skip_false_probs[1][k] != 0)
                cm->mbskip_pred_probs[k] = cpi->last_skip_false_probs[1][k];
            }
          } else {
            int k;
            for (k = 0; k < MBSKIP_CONTEXTS; k++) {
              if (cpi->last_skip_false_probs[0][k] != 0)
                cm->mbskip_pred_probs[k] = cpi->last_skip_false_probs[0][k];
            }
          }

          // as this is for cost estimate, let's make sure it does not
          // get extreme either way
          {
            int k;
            for (k = 0; k < MBSKIP_CONTEXTS; ++k) {
              if (cm->mbskip_pred_probs[k] < 5)
                cm->mbskip_pred_probs[k] = 5;

              if (cm->mbskip_pred_probs[k] > 250)
                cm->mbskip_pred_probs[k] = 250;

              if (cpi->is_src_frame_alt_ref)
                cm->mbskip_pred_probs[k] = 1;
            }
          }
        }
      }

      // Set up entropy depending on frame type.
      if (cm->frame_type == KEY_FRAME)
        vp9_setup_key_frame(cpi);
      else
        vp9_setup_inter_frame(cpi);
    }

    // transform / motion compensation build reconstruction frame

    vp9_encode_frame(cpi);

    // Update the skip mb flag probabilities based on the distribution
    // seen in the last encoder iteration.
    update_base_skip_probs(cpi);

    vp9_clear_system_state();  // __asm emms;

#if CONFIG_PRED_FILTER
    // Update prediction filter on/off probability based on
    // selection made for the current frame
    if (cm->frame_type != KEY_FRAME)
      update_pred_filt_prob(cpi);
#endif

    // Dummy pack of the bitstream using up to date stats to get an
    // accurate estimate of output frame size to determine if we need
    // to recode.
    vp9_save_coding_context(cpi);
    cpi->dummy_packing = 1;
    vp9_pack_bitstream(cpi, dest, size);
    cpi->projected_frame_size = (*size) << 3;
    vp9_restore_coding_context(cpi);

    if (frame_over_shoot_limit == 0)
      frame_over_shoot_limit = 1;
    active_worst_qchanged = FALSE;

    // Special case handling for forced key frames
    if ((cm->frame_type == KEY_FRAME) && cpi->this_key_frame_forced) {
      int last_q = Q;
      int kf_err = vp9_calc_ss_err(cpi->Source,
                                   &cm->yv12_fb[cm->new_fb_idx]);

      int high_err_target = cpi->ambient_err;
      int low_err_target = (cpi->ambient_err >> 1);

      // Prevent possible divide by zero error below for perfect KF
      kf_err += (!kf_err);

      // The key frame is not good enough or we can afford
      // to make it better without undue risk of popping.
      if (((kf_err > high_err_target) &&
           (cpi->projected_frame_size <= frame_over_shoot_limit)) ||
          ((kf_err > low_err_target) &&
           (cpi->projected_frame_size <= frame_under_shoot_limit))) {
        // Lower q_high
        q_high = (Q > q_low) ? (Q - 1) : q_low;

        // Adjust Q
        Q = (Q * high_err_target) / kf_err;
        if (Q < ((q_high + q_low) >> 1))
          Q = (q_high + q_low) >> 1;
      }
      // The key frame is much better than the previous frame
      else if ((kf_err < low_err_target) &&
               (cpi->projected_frame_size >= frame_under_shoot_limit)) {
        // Raise q_low
        q_low = (Q < q_high) ? (Q + 1) : q_high;

        // Adjust Q
        Q = (Q * low_err_target) / kf_err;
        if (Q > ((q_high + q_low + 1) >> 1))
          Q = (q_high + q_low + 1) >> 1;
      }

      // Clamp Q to upper and lower limits:
      if (Q > q_high)
        Q = q_high;
      else if (Q < q_low)
        Q = q_low;

      Loop = ((Q != last_q)) ? TRUE : FALSE;
    }

    // Is the projected frame size out of range and are we allowed to attempt to recode.
    else if (recode_loop_test(cpi,
                              frame_over_shoot_limit, frame_under_shoot_limit,
                              Q, top_index, bottom_index)) {
      int last_q = Q;
      int Retries = 0;

      // Frame size out of permitted range:
      // Update correction factor & compute new Q to try...

      // Frame is too large
      if (cpi->projected_frame_size > cpi->this_frame_target) {
        q_low = (Q < q_high) ? (Q + 1) : q_high; // Raise Qlow as to at least the current value

        if (cpi->zbin_over_quant > 0)            // If we are using over quant do the same for zbin_oq_low
          zbin_oq_low = (cpi->zbin_over_quant < zbin_oq_high) ? (cpi->zbin_over_quant + 1) : zbin_oq_high;

        if (undershoot_seen || (loop_count > 1)) {
          // Update rate_correction_factor unless cpi->active_worst_quality has changed.
          if (!active_worst_qchanged)
            vp9_update_rate_correction_factors(cpi, 1);

          Q = (q_high + q_low + 1) / 2;

          // Adjust cpi->zbin_over_quant (only allowed when Q is max)
          if (Q < MAXQ)
            cpi->zbin_over_quant = 0;
          else {
            zbin_oq_low = (cpi->zbin_over_quant < zbin_oq_high) ? (cpi->zbin_over_quant + 1) : zbin_oq_high;
            cpi->zbin_over_quant = (zbin_oq_high + zbin_oq_low) / 2;
          }
        } else {
          // Update rate_correction_factor unless cpi->active_worst_quality has changed.
          if (!active_worst_qchanged)
            vp9_update_rate_correction_factors(cpi, 0);

          Q = vp9_regulate_q(cpi, cpi->this_frame_target);

          while (((Q < q_low) || (cpi->zbin_over_quant < zbin_oq_low)) && (Retries < 10)) {
            vp9_update_rate_correction_factors(cpi, 0);
            Q = vp9_regulate_q(cpi, cpi->this_frame_target);
            Retries++;
          }
        }

        overshoot_seen = TRUE;
      }
      // Frame is too small
      else {
        if (cpi->zbin_over_quant == 0)
          q_high = (Q > q_low) ? (Q - 1) : q_low; // Lower q_high if not using over quant
        else                                    // else lower zbin_oq_high
          zbin_oq_high = (cpi->zbin_over_quant > zbin_oq_low) ? (cpi->zbin_over_quant - 1) : zbin_oq_low;

        if (overshoot_seen || (loop_count > 1)) {
          // Update rate_correction_factor unless cpi->active_worst_quality has changed.
          if (!active_worst_qchanged)
            vp9_update_rate_correction_factors(cpi, 1);

          Q = (q_high + q_low) / 2;

          // Adjust cpi->zbin_over_quant (only allowed when Q is max)
          if (Q < MAXQ)
            cpi->zbin_over_quant = 0;
          else
            cpi->zbin_over_quant = (zbin_oq_high + zbin_oq_low) / 2;
        } else {
          // Update rate_correction_factor unless cpi->active_worst_quality has changed.
          if (!active_worst_qchanged)
            vp9_update_rate_correction_factors(cpi, 0);

          Q = vp9_regulate_q(cpi, cpi->this_frame_target);

          // Special case reset for qlow for constrained quality.
          // This should only trigger where there is very substantial
          // undershoot on a frame and the auto cq level is above
          // the user passsed in value.
          if ((cpi->oxcf.end_usage == USAGE_CONSTRAINED_QUALITY) &&
              (Q < q_low)) {
            q_low = Q;
          }

          while (((Q > q_high) || (cpi->zbin_over_quant > zbin_oq_high)) && (Retries < 10)) {
            vp9_update_rate_correction_factors(cpi, 0);
            Q = vp9_regulate_q(cpi, cpi->this_frame_target);
            Retries++;
          }
        }

        undershoot_seen = TRUE;
      }

      // Clamp Q to upper and lower limits:
      if (Q > q_high)
        Q = q_high;
      else if (Q < q_low)
        Q = q_low;

      // Clamp cpi->zbin_over_quant
      cpi->zbin_over_quant = (cpi->zbin_over_quant < zbin_oq_low) ?
          zbin_oq_low : (cpi->zbin_over_quant > zbin_oq_high) ?
          zbin_oq_high : cpi->zbin_over_quant;

      // Loop = ((Q != last_q) || (last_zbin_oq != cpi->zbin_over_quant)) ? TRUE : FALSE;
      Loop = ((Q != last_q)) ? TRUE : FALSE;
      last_zbin_oq = cpi->zbin_over_quant;
    } else
      Loop = FALSE;

    if (cpi->is_src_frame_alt_ref)
      Loop = FALSE;

    if (cm->frame_type != KEY_FRAME &&
        !sf->search_best_filter &&
        cm->mcomp_filter_type == SWITCHABLE) {
      int interp_factor = Q / 3;  /* denominator is 256 */
      int count[VP9_SWITCHABLE_FILTERS];
      int tot_count = 0, c = 0, thr;
      int i, j;
      for (i = 0; i < VP9_SWITCHABLE_FILTERS; ++i) {
        count[i] = 0;
        for (j = 0; j <= VP9_SWITCHABLE_FILTERS; ++j) {
          count[i] += cpi->switchable_interp_count[j][i];
        }
        tot_count += count[i];
      }

      thr = ((tot_count * interp_factor + 128) >> 8);
      for (i = 0; i < VP9_SWITCHABLE_FILTERS; ++i) {
        c += (count[i] >= thr);
      }
      if (c == 1) {
        /* Mostly one filter is used. So set the filter at frame level */
        for (i = 0; i < VP9_SWITCHABLE_FILTERS; ++i) {
          if (count[i]) {
            cm->mcomp_filter_type = vp9_switchable_interp[i];
            Loop = TRUE;  /* Make sure to loop since the filter changed */
            break;
          }
        }
      }
    }

    if (Loop == FALSE && cm->frame_type != KEY_FRAME && sf->search_best_filter) {
      if (mcomp_filter_index < mcomp_filters) {
        INT64 err = vp9_calc_ss_err(cpi->Source,
                                    &cm->yv12_fb[cm->new_fb_idx]);
        INT64 rate = cpi->projected_frame_size << 8;
        mcomp_filter_cost[mcomp_filter_index] =
          (RDCOST(cpi->RDMULT, cpi->RDDIV, rate, err));
        mcomp_filter_index++;
        if (mcomp_filter_index < mcomp_filters) {
          cm->mcomp_filter_type = mcomp_filters_to_search[mcomp_filter_index];
          loop_count = -1;
          Loop = TRUE;
        } else {
          int f;
          INT64 best_cost = mcomp_filter_cost[0];
          int mcomp_best_filter = mcomp_filters_to_search[0];
          for (f = 1; f < mcomp_filters; f++) {
            if (mcomp_filter_cost[f] < best_cost) {
              mcomp_best_filter = mcomp_filters_to_search[f];
              best_cost = mcomp_filter_cost[f];
            }
          }
          if (mcomp_best_filter != mcomp_filters_to_search[mcomp_filters - 1]) {
            loop_count = -1;
            Loop = TRUE;
            cm->mcomp_filter_type = mcomp_best_filter;
          }
          /*
          printf("  best filter = %d, ( ", mcomp_best_filter);
          for (f=0;f<mcomp_filters; f++) printf("%d ",  mcomp_filter_cost[f]);
          printf(")\n");
          */
        }
#if RESET_FOREACH_FILTER
        if (Loop == TRUE) {
          overshoot_seen = FALSE;
          undershoot_seen = FALSE;
          zbin_oq_low = zbin_oq_low0;
          zbin_oq_high = zbin_oq_high0;
          q_low = q_low0;
          q_high = q_high0;
          Q = Q0;
          cpi->zbin_over_quant = last_zbin_oq = last_zbin_oq0;
          cpi->rate_correction_factor = rate_correction_factor0;
          cpi->gf_rate_correction_factor = gf_rate_correction_factor0;
          cpi->active_best_quality = active_best_quality0;
          cpi->active_worst_quality = active_worst_quality0;
        }
#endif
      }
    }

    if (Loop == TRUE) {
      loop_count++;
#if CONFIG_INTERNAL_STATS
      cpi->tot_recode_hits++;
#endif
    }
  } while (Loop == TRUE);

  // Special case code to reduce pulsing when key frames are forced at a
  // fixed interval. Note the reconstruction error if it is the frame before
  // the force key frame
  if (cpi->next_key_frame_forced && (cpi->twopass.frames_to_key == 0)) {
    cpi->ambient_err = vp9_calc_ss_err(cpi->Source,
                                       &cm->yv12_fb[cm->new_fb_idx]);
  }

  // This frame's MVs are saved and will be used in next frame's MV
  // prediction. Last frame has one more line(add to bottom) and one
  // more column(add to right) than cm->mip. The edge elements are
  // initialized to 0.
  if (cm->show_frame) { // do not save for altref frame
    int mb_row;
    int mb_col;
    MODE_INFO *tmp = cm->mip;

    if (cm->frame_type != KEY_FRAME) {
      for (mb_row = 0; mb_row < cm->mb_rows + 1; mb_row ++) {
        for (mb_col = 0; mb_col < cm->mb_cols + 1; mb_col ++) {
          if (tmp->mbmi.ref_frame != INTRA_FRAME)
            cpi->lfmv[mb_col + mb_row * (cm->mode_info_stride + 1)].as_int = tmp->mbmi.mv[0].as_int;

          cpi->lf_ref_frame_sign_bias[mb_col + mb_row * (cm->mode_info_stride + 1)] = cm->ref_frame_sign_bias[tmp->mbmi.ref_frame];
          cpi->lf_ref_frame[mb_col + mb_row * (cm->mode_info_stride + 1)] = tmp->mbmi.ref_frame;
          tmp++;
        }
      }
    }
  }

  // Update the GF useage maps.
  // This is done after completing the compression of a frame when all modes
  // etc. are finalized but before loop filter
  vp9_update_gf_useage_maps(cpi, cm, &cpi->mb);

  if (cm->frame_type == KEY_FRAME)
    cm->refresh_last_frame = 1;

#if 0
  {
    FILE *f = fopen("gfactive.stt", "a");
    fprintf(f, "%8d %8d %8d %8d %8d\n", cm->current_video_frame, (100 * cpi->gf_active_count) / (cpi->common.mb_rows * cpi->common.mb_cols), cpi->this_iiratio, cpi->next_iiratio, cm->refresh_golden_frame);
    fclose(f);
  }
#endif

  cm->frame_to_show = &cm->yv12_fb[cm->new_fb_idx];

#if WRITE_RECON_BUFFER
  if (cm->show_frame)
    write_cx_frame_to_file(cm->frame_to_show,
                           cm->current_video_frame);
  else
    write_cx_frame_to_file(cm->frame_to_show,
                           cm->current_video_frame + 1000);
#endif

  // Pick the loop filter level for the frame.
  loopfilter_frame(cpi, cm);

  // build the bitstream
  cpi->dummy_packing = 0;
  vp9_pack_bitstream(cpi, dest, size);

  if (cpi->mb.e_mbd.update_mb_segmentation_map) {
    update_reference_segmentation_map(cpi);
  }

#if CONFIG_PRED_FILTER
  // Select the prediction filtering mode to use for the
  // next frame based on the current frame selections
  if (cm->frame_type != KEY_FRAME)
    select_pred_filter_mode(cpi);
#endif

  update_reference_frames(cm);
  vp9_copy(cpi->common.fc.coef_counts, cpi->coef_counts);
  vp9_copy(cpi->common.fc.hybrid_coef_counts, cpi->hybrid_coef_counts);
  vp9_copy(cpi->common.fc.coef_counts_8x8, cpi->coef_counts_8x8);
  vp9_copy(cpi->common.fc.hybrid_coef_counts_8x8, cpi->hybrid_coef_counts_8x8);
  vp9_copy(cpi->common.fc.coef_counts_16x16, cpi->coef_counts_16x16);
  vp9_copy(cpi->common.fc.hybrid_coef_counts_16x16,
           cpi->hybrid_coef_counts_16x16);
  vp9_adapt_coef_probs(&cpi->common);
  if (cpi->common.frame_type != KEY_FRAME) {
    vp9_copy(cpi->common.fc.ymode_counts, cpi->ymode_count);
    vp9_copy(cpi->common.fc.uv_mode_counts, cpi->y_uv_mode_count);
    vp9_copy(cpi->common.fc.bmode_counts, cpi->bmode_count);
    vp9_copy(cpi->common.fc.i8x8_mode_counts, cpi->i8x8_mode_count);
    vp9_copy(cpi->common.fc.sub_mv_ref_counts, cpi->sub_mv_ref_count);
    vp9_copy(cpi->common.fc.mbsplit_counts, cpi->mbsplit_count);
    vp9_adapt_mode_probs(&cpi->common);

    cpi->common.fc.NMVcount = cpi->NMVcount;
    vp9_adapt_nmv_probs(&cpi->common, cpi->mb.e_mbd.allow_high_precision_mv);
    vp9_update_mode_context(&cpi->common);
  }

  /* Move storing frame_type out of the above loop since it is also
   * needed in motion search besides loopfilter */
  cm->last_frame_type = cm->frame_type;

  // Keep a copy of the size estimate used in the loop
  loop_size_estimate = cpi->projected_frame_size;

  // Update rate control heuristics
  cpi->total_byte_count += (*size);
  cpi->projected_frame_size = (*size) << 3;

  if (!active_worst_qchanged)
    vp9_update_rate_correction_factors(cpi, 2);

  cpi->last_q[cm->frame_type] = cm->base_qindex;

  // Keep record of last boosted (KF/KF/ARF) Q value.
  // If the current frame is coded at a lower Q then we also update it.
  // If all mbs in this group are skipped only update if the Q value is
  // better than that already stored.
  // This is used to help set quality in forced key frames to reduce popping
  if ((cm->base_qindex < cpi->last_boosted_qindex) ||
      ((cpi->static_mb_pct < 100) &&
       ((cm->frame_type == KEY_FRAME) ||
        cm->refresh_alt_ref_frame ||
        (cm->refresh_golden_frame && !cpi->is_src_frame_alt_ref)))) {
    cpi->last_boosted_qindex = cm->base_qindex;
  }

  if (cm->frame_type == KEY_FRAME) {
    vp9_adjust_key_frame_context(cpi);
  }

  // Keep a record of ambient average Q.
  if (cm->frame_type != KEY_FRAME)
    cpi->avg_frame_qindex = (2 + 3 * cpi->avg_frame_qindex + cm->base_qindex) >> 2;

  // Keep a record from which we can calculate the average Q excluding GF updates and key frames
  if ((cm->frame_type != KEY_FRAME) && !cm->refresh_golden_frame && !cm->refresh_alt_ref_frame) {
    cpi->ni_frames++;
    cpi->tot_q += vp9_convert_qindex_to_q(Q);
    cpi->avg_q = cpi->tot_q / (double)cpi->ni_frames;

    // Calculate the average Q for normal inter frames (not key or GFU
    // frames).
    cpi->ni_tot_qi += Q;
    cpi->ni_av_qi = (cpi->ni_tot_qi / cpi->ni_frames);
  }

  // Update the buffer level variable.
  // Non-viewable frames are a special case and are treated as pure overhead.
  if (!cm->show_frame)
    cpi->bits_off_target -= cpi->projected_frame_size;
  else
    cpi->bits_off_target += cpi->av_per_frame_bandwidth - cpi->projected_frame_size;

  // Clip the buffer level at the maximum buffer size
  if (cpi->bits_off_target > cpi->oxcf.maximum_buffer_size)
    cpi->bits_off_target = cpi->oxcf.maximum_buffer_size;

  // Rolling monitors of whether we are over or underspending used to help regulate min and Max Q in two pass.
  cpi->rolling_target_bits = ((cpi->rolling_target_bits * 3) + cpi->this_frame_target + 2) / 4;
  cpi->rolling_actual_bits = ((cpi->rolling_actual_bits * 3) + cpi->projected_frame_size + 2) / 4;
  cpi->long_rolling_target_bits = ((cpi->long_rolling_target_bits * 31) + cpi->this_frame_target + 16) / 32;
  cpi->long_rolling_actual_bits = ((cpi->long_rolling_actual_bits * 31) + cpi->projected_frame_size + 16) / 32;

  // Actual bits spent
  cpi->total_actual_bits    += cpi->projected_frame_size;

  // Debug stats
  cpi->total_target_vs_actual += (cpi->this_frame_target - cpi->projected_frame_size);

  cpi->buffer_level = cpi->bits_off_target;

  // Update bits left to the kf and gf groups to account for overshoot or undershoot on these frames
  if (cm->frame_type == KEY_FRAME) {
    cpi->twopass.kf_group_bits += cpi->this_frame_target - cpi->projected_frame_size;

    if (cpi->twopass.kf_group_bits < 0)
      cpi->twopass.kf_group_bits = 0;
  } else if (cm->refresh_golden_frame || cm->refresh_alt_ref_frame) {
    cpi->twopass.gf_group_bits += cpi->this_frame_target - cpi->projected_frame_size;

    if (cpi->twopass.gf_group_bits < 0)
      cpi->twopass.gf_group_bits = 0;
  }

  // Update the skip mb flag probabilities based on the distribution seen
  // in this frame.
  update_base_skip_probs(cpi);

#if 0 //CONFIG_NEW_MVREF && CONFIG_INTERNAL_STATS
  {
    FILE *f = fopen("mv_ref_dist.stt", "a");
    unsigned int i;
    for (i = 0; i < MAX_MV_REFS; ++i) {
      fprintf(f, "%10d", cpi->best_ref_index_counts[0][i]);
    }
    fprintf(f, "\n" );

    fclose(f);
  }
#endif

#if 0// 1 && CONFIG_INTERNAL_STATS
  {
    FILE *f = fopen("tmp.stt", "a");
    int recon_err;

    vp9_clear_system_state();  // __asm emms;

    recon_err = vp9_calc_ss_err(cpi->Source,
                                &cm->yv12_fb[cm->new_fb_idx]);

    if (cpi->twopass.total_left_stats->coded_error != 0.0)
      fprintf(f, "%10d %10d %10d %10d %10d %10d %10d %10d"
              "%7.2f %7.2f %7.2f %7.2f %7.2f %7.2f %7.2f"
              "%6d %5d %5d %5d %8d %8.2f %10d %10.3f"
              "%10.3f %8d %10d %10d %10d\n",
              cpi->common.current_video_frame, cpi->this_frame_target,
              cpi->projected_frame_size, loop_size_estimate,
              (cpi->projected_frame_size - cpi->this_frame_target),
              (int)cpi->total_target_vs_actual,
              (cpi->oxcf.starting_buffer_level - cpi->bits_off_target),
              (int)cpi->total_actual_bits,
              vp9_convert_qindex_to_q(cm->base_qindex),
              (double)vp9_dc_quant(cm->base_qindex, 0) / 4.0,
              vp9_convert_qindex_to_q(cpi->active_best_quality),
              vp9_convert_qindex_to_q(cpi->active_worst_quality),
              cpi->avg_q,
              vp9_convert_qindex_to_q(cpi->ni_av_qi),
              vp9_convert_qindex_to_q(cpi->cq_target_quality),
              cpi->zbin_over_quant,
              // cpi->avg_frame_qindex, cpi->zbin_over_quant,
              cm->refresh_golden_frame, cm->refresh_alt_ref_frame,
              cm->frame_type, cpi->gfu_boost,
              cpi->twopass.est_max_qcorrection_factor,
              (int)cpi->twopass.bits_left,
              cpi->twopass.total_left_stats->coded_error,
              (double)cpi->twopass.bits_left /
              cpi->twopass.total_left_stats->coded_error,
              cpi->tot_recode_hits, recon_err, cpi->kf_boost,
              cpi->kf_zeromotion_pct);
    else
      fprintf(f, "%10d %10d %10d %10d %10d %10d %10d %10d"
              "%7.2f %7.2f %7.2f %7.2f %7.2f %7.2f %7.2f"
              "%6d %5d %5d %5d %8d %8.2f %10d %10.3f"
              "%8d %10d %10d %10d\n",
              cpi->common.current_video_frame,
              cpi->this_frame_target, cpi->projected_frame_size,
              loop_size_estimate,
              (cpi->projected_frame_size - cpi->this_frame_target),
              (int)cpi->total_target_vs_actual,
              (cpi->oxcf.starting_buffer_level - cpi->bits_off_target),
              (int)cpi->total_actual_bits,
              vp9_convert_qindex_to_q(cm->base_qindex),
              (double)vp9_dc_quant(cm->base_qindex, 0) / 4.0,
              vp9_convert_qindex_to_q(cpi->active_best_quality),
              vp9_convert_qindex_to_q(cpi->active_worst_quality),
              cpi->avg_q,
              vp9_convert_qindex_to_q(cpi->ni_av_qi),
              vp9_convert_qindex_to_q(cpi->cq_target_quality),
              cpi->zbin_over_quant,
              // cpi->avg_frame_qindex, cpi->zbin_over_quant,
              cm->refresh_golden_frame, cm->refresh_alt_ref_frame,
              cm->frame_type, cpi->gfu_boost,
              cpi->twopass.est_max_qcorrection_factor,
              (int)cpi->twopass.bits_left,
              cpi->twopass.total_left_stats->coded_error,
              cpi->tot_recode_hits, recon_err, cpi->kf_boost,
              cpi->kf_zeromotion_pct);

    fclose(f);

    if (0) {
      FILE *fmodes = fopen("Modes.stt", "a");
      int i;

      fprintf(fmodes, "%6d:%1d:%1d:%1d ",
              cpi->common.current_video_frame,
              cm->frame_type, cm->refresh_golden_frame,
              cm->refresh_alt_ref_frame);

      for (i = 0; i < MAX_MODES; i++)
        fprintf(fmodes, "%5d ", cpi->mode_chosen_counts[i]);

      fprintf(fmodes, "\n");

      fclose(fmodes);
    }
  }

#endif

#if 0
  // Debug stats for segment feature experiments.
  print_seg_map(cpi);
#endif

  // If this was a kf or Gf note the Q
  if ((cm->frame_type == KEY_FRAME) || cm->refresh_golden_frame || cm->refresh_alt_ref_frame)
    cm->last_kf_gf_q = cm->base_qindex;

  if (cm->refresh_golden_frame == 1)
    cm->frame_flags = cm->frame_flags | FRAMEFLAGS_GOLDEN;
  else
    cm->frame_flags = cm->frame_flags&~FRAMEFLAGS_GOLDEN;

  if (cm->refresh_alt_ref_frame == 1)
    cm->frame_flags = cm->frame_flags | FRAMEFLAGS_ALTREF;
  else
    cm->frame_flags = cm->frame_flags&~FRAMEFLAGS_ALTREF;


  if (cm->refresh_last_frame & cm->refresh_golden_frame) // both refreshed
    cpi->gold_is_last = 1;
  else if (cm->refresh_last_frame ^ cm->refresh_golden_frame) // 1 refreshed but not the other
    cpi->gold_is_last = 0;

  if (cm->refresh_last_frame & cm->refresh_alt_ref_frame) // both refreshed
    cpi->alt_is_last = 1;
  else if (cm->refresh_last_frame ^ cm->refresh_alt_ref_frame) // 1 refreshed but not the other
    cpi->alt_is_last = 0;

  if (cm->refresh_alt_ref_frame & cm->refresh_golden_frame) // both refreshed
    cpi->gold_is_alt = 1;
  else if (cm->refresh_alt_ref_frame ^ cm->refresh_golden_frame) // 1 refreshed but not the other
    cpi->gold_is_alt = 0;

  cpi->ref_frame_flags = VP9_ALT_FLAG | VP9_GOLD_FLAG | VP9_LAST_FLAG;

  if (cpi->gold_is_last)
    cpi->ref_frame_flags &= ~VP9_GOLD_FLAG;

  if (cpi->alt_is_last)
    cpi->ref_frame_flags &= ~VP9_ALT_FLAG;

  if (cpi->gold_is_alt)
    cpi->ref_frame_flags &= ~VP9_ALT_FLAG;

  if (cpi->oxcf.play_alternate && cm->refresh_alt_ref_frame && (cm->frame_type != KEY_FRAME))
    // Update the alternate reference frame stats as appropriate.
    update_alt_ref_frame_stats(cpi);
  else
    // Update the Golden frame stats as appropriate.
    update_golden_frame_stats(cpi);

  if (cm->frame_type == KEY_FRAME) {
    // Tell the caller that the frame was coded as a key frame
    *frame_flags = cm->frame_flags | FRAMEFLAGS_KEY;

    // As this frame is a key frame  the next defaults to an inter frame.
    cm->frame_type = INTER_FRAME;
  } else {
    *frame_flags = cm->frame_flags&~FRAMEFLAGS_KEY;
  }

  // Clear the one shot update flags for segmentation map and mode/ref loop filter deltas.
  xd->update_mb_segmentation_map = 0;
  xd->update_mb_segmentation_data = 0;
  xd->mode_ref_lf_delta_update = 0;


  // Dont increment frame counters if this was an altref buffer update not a real frame
  if (cm->show_frame) {
    cm->current_video_frame++;
    cpi->frames_since_key++;
  }

  // reset to normal state now that we are done.


#if 0
  {
    char filename[512];
    FILE *recon_file;
    sprintf(filename, "enc%04d.yuv", (int) cm->current_video_frame);
    recon_file = fopen(filename, "wb");
    fwrite(cm->yv12_fb[cm->lst_fb_idx].buffer_alloc,
           cm->yv12_fb[cm->lst_fb_idx].frame_size, 1, recon_file);
    fclose(recon_file);
  }
#endif
#ifdef OUTPUT_YUV_REC
  vp9_write_yuv_rec_frame(cm);
#endif

  if (cm->show_frame) {
    vpx_memcpy(cm->prev_mip, cm->mip,
               (cm->mb_cols + 1) * (cm->mb_rows + 1)* sizeof(MODE_INFO));
  } else {
    vpx_memset(cm->prev_mip, 0,
               (cm->mb_cols + 1) * (cm->mb_rows + 1)* sizeof(MODE_INFO));
  }
}

static void Pass2Encode(VP9_COMP *cpi, unsigned long *size,
                        unsigned char *dest, unsigned int *frame_flags) {

  if (!cpi->common.refresh_alt_ref_frame)
    vp9_second_pass(cpi);

  encode_frame_to_data_rate(cpi, size, dest, frame_flags);
  cpi->twopass.bits_left -= 8 * *size;

  if (!cpi->common.refresh_alt_ref_frame) {
    double lower_bounds_min_rate = FRAME_OVERHEAD_BITS * cpi->oxcf.frame_rate;
    double two_pass_min_rate = (double)(cpi->oxcf.target_bandwidth
                                        * cpi->oxcf.two_pass_vbrmin_section / 100);

    if (two_pass_min_rate < lower_bounds_min_rate)
      two_pass_min_rate = lower_bounds_min_rate;

    cpi->twopass.bits_left += (int64_t)(two_pass_min_rate / cpi->oxcf.frame_rate);
  }
}

// For ARM NEON, d8-d15 are callee-saved registers, and need to be saved by us.
#if HAVE_ARMV7
extern void vp9_push_neon(int64_t *store);
extern void vp9_pop_neon(int64_t *store);
#endif


int vp9_receive_raw_frame(VP9_PTR ptr, unsigned int frame_flags,
                          YV12_BUFFER_CONFIG *sd, int64_t time_stamp,
                          int64_t end_time) {
#if HAVE_ARMV7
  int64_t store_reg[8];
#endif
  VP9_COMP              *cpi = (VP9_COMP *) ptr;
  VP9_COMMON            *cm = &cpi->common;
  struct vpx_usec_timer  timer;
  int                    res = 0;

#if HAVE_ARMV7
#if CONFIG_RUNTIME_CPU_DETECT
  if (cm->rtcd.flags & HAS_NEON)
#endif
  {
    vp9_push_neon(store_reg);
  }
#endif

  vpx_usec_timer_start(&timer);
  if (vp9_lookahead_push(cpi->lookahead, sd, time_stamp, end_time, frame_flags,
                         cpi->active_map_enabled ? cpi->active_map : NULL))
    res = -1;
  cm->clr_type = sd->clrtype;
  vpx_usec_timer_mark(&timer);
  cpi->time_receive_data += vpx_usec_timer_elapsed(&timer);

#if HAVE_ARMV7
#if CONFIG_RUNTIME_CPU_DETECT
  if (cm->rtcd.flags & HAS_NEON)
#endif
  {
    vp9_pop_neon(store_reg);
  }
#endif

  return res;
}


static int frame_is_reference(const VP9_COMP *cpi) {
  const VP9_COMMON *cm = &cpi->common;
  const MACROBLOCKD *xd = &cpi->mb.e_mbd;

  return cm->frame_type == KEY_FRAME || cm->refresh_last_frame
         || cm->refresh_golden_frame || cm->refresh_alt_ref_frame
         || cm->copy_buffer_to_gf || cm->copy_buffer_to_arf
         || cm->refresh_entropy_probs
         || xd->mode_ref_lf_delta_update
         || xd->update_mb_segmentation_map || xd->update_mb_segmentation_data;
}


int vp9_get_compressed_data(VP9_PTR ptr, unsigned int *frame_flags,
                            unsigned long *size, unsigned char *dest,
                            int64_t *time_stamp, int64_t *time_end, int flush) {
#if HAVE_ARMV7
  int64_t store_reg[8];
#endif
  VP9_COMP *cpi = (VP9_COMP *) ptr;
  VP9_COMMON *cm = &cpi->common;
  struct vpx_usec_timer  cmptimer;
  YV12_BUFFER_CONFIG    *force_src_buffer = NULL;

  if (!cpi)
    return -1;

#if HAVE_ARMV7
#if CONFIG_RUNTIME_CPU_DETECT
  if (cm->rtcd.flags & HAS_NEON)
#endif
  {
    vp9_push_neon(store_reg);
  }
#endif

  vpx_usec_timer_start(&cmptimer);

  cpi->source = NULL;

  cpi->mb.e_mbd.allow_high_precision_mv = ALTREF_HIGH_PRECISION_MV;
  // Should we code an alternate reference frame
  if (cpi->oxcf.play_alternate &&
      cpi->source_alt_ref_pending) {
    if ((cpi->source = vp9_lookahead_peek(cpi->lookahead,
                                          cpi->frames_till_gf_update_due))) {
      cpi->alt_ref_source = cpi->source;
      if (cpi->oxcf.arnr_max_frames > 0) {
        vp9_temporal_filter_prepare_c(cpi,
                                      cpi->frames_till_gf_update_due);
        force_src_buffer = &cpi->alt_ref_buffer;
      }
      cm->frames_till_alt_ref_frame = cpi->frames_till_gf_update_due;
      cm->refresh_alt_ref_frame = 1;
      cm->refresh_golden_frame = 0;
      cm->refresh_last_frame = 0;
      cm->show_frame = 0;
      cpi->source_alt_ref_pending = FALSE;   // Clear Pending altf Ref flag.
      cpi->is_src_frame_alt_ref = 0;
    }
  }

  if (!cpi->source) {
    if ((cpi->source = vp9_lookahead_pop(cpi->lookahead, flush))) {
      cm->show_frame = 1;

      cpi->is_src_frame_alt_ref = cpi->alt_ref_source
                                  && (cpi->source == cpi->alt_ref_source);

      if (cpi->is_src_frame_alt_ref)
        cpi->alt_ref_source = NULL;
    }
  }

  if (cpi->source) {
    cpi->un_scaled_source =
      cpi->Source = force_src_buffer ? force_src_buffer : &cpi->source->img;
    *time_stamp = cpi->source->ts_start;
    *time_end = cpi->source->ts_end;
    *frame_flags = cpi->source->flags;
  } else {
    *size = 0;
    if (flush && cpi->pass == 1 && !cpi->twopass.first_pass_done) {
      vp9_end_first_pass(cpi);    /* get last stats packet */
      cpi->twopass.first_pass_done = 1;
    }

#if HAVE_ARMV7
#if CONFIG_RUNTIME_CPU_DETECT
    if (cm->rtcd.flags & HAS_NEON)
#endif
    {
      vp9_pop_neon(store_reg);
    }
#endif
    return -1;
  }

  if (cpi->source->ts_start < cpi->first_time_stamp_ever) {
    cpi->first_time_stamp_ever = cpi->source->ts_start;
    cpi->last_end_time_stamp_seen = cpi->source->ts_start;
  }

  // adjust frame rates based on timestamps given
  if (!cm->refresh_alt_ref_frame) {
    int64_t this_duration;
    int step = 0;

    if (cpi->source->ts_start == cpi->first_time_stamp_ever) {
      this_duration = cpi->source->ts_end - cpi->source->ts_start;
      step = 1;
    } else {
      int64_t last_duration;

      this_duration = cpi->source->ts_end - cpi->last_end_time_stamp_seen;
      last_duration = cpi->last_end_time_stamp_seen
                      - cpi->last_time_stamp_seen;
      // do a step update if the duration changes by 10%
      if (last_duration)
        step = ((this_duration - last_duration) * 10 / last_duration);
    }

    if (this_duration) {
      if (step)
        vp9_new_frame_rate(cpi, 10000000.0 / this_duration);
      else {
        double avg_duration, interval;

        /* Average this frame's rate into the last second's average
         * frame rate. If we haven't seen 1 second yet, then average
         * over the whole interval seen.
         */
        interval = cpi->source->ts_end - cpi->first_time_stamp_ever;
        if (interval > 10000000.0)
          interval = 10000000;

        avg_duration = 10000000.0 / cpi->oxcf.frame_rate;
        avg_duration *= (interval - avg_duration + this_duration);
        avg_duration /= interval;

        vp9_new_frame_rate(cpi, 10000000.0 / avg_duration);
      }
    }

    cpi->last_time_stamp_seen = cpi->source->ts_start;
    cpi->last_end_time_stamp_seen = cpi->source->ts_end;
  }

  // start with a 0 size frame
  *size = 0;

  // Clear down mmx registers
  vp9_clear_system_state();  // __asm emms;

  cm->frame_type = INTER_FRAME;
  cm->frame_flags = *frame_flags;

#if 0

  if (cm->refresh_alt_ref_frame) {
    // cm->refresh_golden_frame = 1;
    cm->refresh_golden_frame = 0;
    cm->refresh_last_frame = 0;
  } else {
    cm->refresh_golden_frame = 0;
    cm->refresh_last_frame = 1;
  }

#endif
  /* find a free buffer for the new frame */
  {
    int i = 0;
    for (; i < NUM_YV12_BUFFERS; i++) {
      if (!cm->yv12_fb[i].flags) {
        cm->new_fb_idx = i;
        break;
      }
    }

    assert(i < NUM_YV12_BUFFERS);
  }
  if (cpi->pass == 1) {
    Pass1Encode(cpi, size, dest, frame_flags);
  } else if (cpi->pass == 2) {
    Pass2Encode(cpi, size, dest, frame_flags);
  } else {
    encode_frame_to_data_rate(cpi, size, dest, frame_flags);
  }

  if (cm->refresh_entropy_probs) {
    if (cm->refresh_alt_ref_frame)
      vpx_memcpy(&cm->lfc_a, &cm->fc, sizeof(cm->fc));
    else
      vpx_memcpy(&cm->lfc, &cm->fc, sizeof(cm->fc));
  }

  // if its a dropped frame honor the requests on subsequent frames
  if (*size > 0) {
    cpi->droppable = !frame_is_reference(cpi);

    // return to normal state
    cm->refresh_entropy_probs = 1;
    cm->refresh_alt_ref_frame = 0;
    cm->refresh_golden_frame = 0;
    cm->refresh_last_frame = 1;
    cm->frame_type = INTER_FRAME;

  }

  vpx_usec_timer_mark(&cmptimer);
  cpi->time_compress_data += vpx_usec_timer_elapsed(&cmptimer);

  if (cpi->b_calculate_psnr && cpi->pass != 1 && cm->show_frame) {
    generate_psnr_packet(cpi);
  }

#if CONFIG_INTERNAL_STATS

  if (cpi->pass != 1) {
    cpi->bytes += *size;

    if (cm->show_frame) {

      cpi->count++;

      if (cpi->b_calculate_psnr) {
        double ye, ue, ve;
        double frame_psnr;
        YV12_BUFFER_CONFIG      *orig = cpi->Source;
        YV12_BUFFER_CONFIG      *recon = cpi->common.frame_to_show;
        YV12_BUFFER_CONFIG      *pp = &cm->post_proc_buffer;
        int y_samples = orig->y_height * orig->y_width;
        int uv_samples = orig->uv_height * orig->uv_width;
        int t_samples = y_samples + 2 * uv_samples;
        int64_t sq_error;

        ye = calc_plane_error(orig->y_buffer, orig->y_stride,
                              recon->y_buffer, recon->y_stride, orig->y_width,
                              orig->y_height);

        ue = calc_plane_error(orig->u_buffer, orig->uv_stride,
                              recon->u_buffer, recon->uv_stride, orig->uv_width,
                              orig->uv_height);

        ve = calc_plane_error(orig->v_buffer, orig->uv_stride,
                              recon->v_buffer, recon->uv_stride, orig->uv_width,
                              orig->uv_height);

        sq_error = ye + ue + ve;

        frame_psnr = vp9_mse2psnr(t_samples, 255.0, sq_error);

        cpi->total_y += vp9_mse2psnr(y_samples, 255.0, ye);
        cpi->total_u += vp9_mse2psnr(uv_samples, 255.0, ue);
        cpi->total_v += vp9_mse2psnr(uv_samples, 255.0, ve);
        cpi->total_sq_error += sq_error;
        cpi->total  += frame_psnr;
        {
          double frame_psnr2, frame_ssim2 = 0;
          double weight = 0;
#if CONFIG_POSTPROC
          vp9_deblock(cm->frame_to_show, &cm->post_proc_buffer, cm->filter_level * 10 / 6, 1, 0, IF_RTCD(&cm->rtcd.postproc));
#endif
          vp9_clear_system_state();

          ye = calc_plane_error(orig->y_buffer, orig->y_stride,
                                pp->y_buffer, pp->y_stride, orig->y_width,
                                orig->y_height);

          ue = calc_plane_error(orig->u_buffer, orig->uv_stride,
                                pp->u_buffer, pp->uv_stride, orig->uv_width,
                                orig->uv_height);

          ve = calc_plane_error(orig->v_buffer, orig->uv_stride,
                                pp->v_buffer, pp->uv_stride, orig->uv_width,
                                orig->uv_height);

          sq_error = ye + ue + ve;

          frame_psnr2 = vp9_mse2psnr(t_samples, 255.0, sq_error);

          cpi->totalp_y += vp9_mse2psnr(y_samples, 255.0, ye);
          cpi->totalp_u += vp9_mse2psnr(uv_samples, 255.0, ue);
          cpi->totalp_v += vp9_mse2psnr(uv_samples, 255.0, ve);
          cpi->total_sq_error2 += sq_error;
          cpi->totalp  += frame_psnr2;

          frame_ssim2 = vp9_calc_ssim(cpi->Source,
                                      &cm->post_proc_buffer, 1, &weight);

          cpi->summed_quality += frame_ssim2 * weight;
          cpi->summed_weights += weight;
#if 0
          {
            FILE *f = fopen("q_used.stt", "a");
            fprintf(f, "%5d : Y%f7.3:U%f7.3:V%f7.3:F%f7.3:S%7.3f\n",
                    cpi->common.current_video_frame, y2, u2, v2,
                    frame_psnr2, frame_ssim2);
            fclose(f);
          }
#endif
        }
      }

      if (cpi->b_calculate_ssimg) {
        double y, u, v, frame_all;
        frame_all =  vp9_calc_ssimg(cpi->Source, cm->frame_to_show,
                                    &y, &u, &v);
        cpi->total_ssimg_y += y;
        cpi->total_ssimg_u += u;
        cpi->total_ssimg_v += v;
        cpi->total_ssimg_all += frame_all;
      }

    }
  }

#endif

#if HAVE_ARMV7
#if CONFIG_RUNTIME_CPU_DETECT
  if (cm->rtcd.flags & HAS_NEON)
#endif
  {
    vp9_pop_neon(store_reg);
  }
#endif

  return 0;
}

int vp9_get_preview_raw_frame(VP9_PTR comp, YV12_BUFFER_CONFIG *dest,
                              vp9_ppflags_t *flags) {
  VP9_COMP *cpi = (VP9_COMP *) comp;

  if (cpi->common.refresh_alt_ref_frame)
    return -1;
  else {
    int ret;
#if CONFIG_POSTPROC
    ret = vp9_post_proc_frame(&cpi->common, dest, flags);
#else

    if (cpi->common.frame_to_show) {
      *dest = *cpi->common.frame_to_show;
      dest->y_width = cpi->common.Width;
      dest->y_height = cpi->common.Height;
      dest->uv_height = cpi->common.Height / 2;
      ret = 0;
    } else {
      ret = -1;
    }

#endif // !CONFIG_POSTPROC
    vp9_clear_system_state();
    return ret;
  }
}

int vp9_set_roimap(VP9_PTR comp, unsigned char *map, unsigned int rows,
                   unsigned int cols, int delta_q[4], int delta_lf[4],
                   unsigned int threshold[4]) {
  VP9_COMP *cpi = (VP9_COMP *) comp;
  signed char feature_data[SEG_LVL_MAX][MAX_MB_SEGMENTS];
  MACROBLOCKD *xd = &cpi->mb.e_mbd;
  int i;

  if (cpi->common.mb_rows != rows || cpi->common.mb_cols != cols)
    return -1;

  if (!map) {
    vp9_disable_segmentation((VP9_PTR)cpi);
    return 0;
  }

  // Set the segmentation Map
  vp9_set_segmentation_map((VP9_PTR)cpi, map);

  // Activate segmentation.
  vp9_enable_segmentation((VP9_PTR)cpi);

  // Set up the quant segment data
  feature_data[SEG_LVL_ALT_Q][0] = delta_q[0];
  feature_data[SEG_LVL_ALT_Q][1] = delta_q[1];
  feature_data[SEG_LVL_ALT_Q][2] = delta_q[2];
  feature_data[SEG_LVL_ALT_Q][3] = delta_q[3];

  // Set up the loop segment data s
  feature_data[SEG_LVL_ALT_LF][0] = delta_lf[0];
  feature_data[SEG_LVL_ALT_LF][1] = delta_lf[1];
  feature_data[SEG_LVL_ALT_LF][2] = delta_lf[2];
  feature_data[SEG_LVL_ALT_LF][3] = delta_lf[3];

  cpi->segment_encode_breakout[0] = threshold[0];
  cpi->segment_encode_breakout[1] = threshold[1];
  cpi->segment_encode_breakout[2] = threshold[2];
  cpi->segment_encode_breakout[3] = threshold[3];

  // Enable the loop and quant changes in the feature mask
  for (i = 0; i < 4; i++) {
    if (delta_q[i])
      vp9_enable_segfeature(xd, i, SEG_LVL_ALT_Q);
    else
      vp9_disable_segfeature(xd, i, SEG_LVL_ALT_Q);

    if (delta_lf[i])
      vp9_enable_segfeature(xd, i, SEG_LVL_ALT_LF);
    else
      vp9_disable_segfeature(xd, i, SEG_LVL_ALT_LF);
  }

  // Initialise the feature data structure
  // SEGMENT_DELTADATA    0, SEGMENT_ABSDATA      1
  vp9_set_segment_data((VP9_PTR)cpi, &feature_data[0][0], SEGMENT_DELTADATA);

  return 0;
}

int vp9_set_active_map(VP9_PTR comp, unsigned char *map,
                       unsigned int rows, unsigned int cols) {
  VP9_COMP *cpi = (VP9_COMP *) comp;

  if (rows == cpi->common.mb_rows && cols == cpi->common.mb_cols) {
    if (map) {
      vpx_memcpy(cpi->active_map, map, rows * cols);
      cpi->active_map_enabled = 1;
    } else
      cpi->active_map_enabled = 0;

    return 0;
  } else {
    // cpi->active_map_enabled = 0;
    return -1;
  }
}

int vp9_set_internal_size(VP9_PTR comp,
                          VPX_SCALING horiz_mode, VPX_SCALING vert_mode) {
  VP9_COMP *cpi = (VP9_COMP *) comp;

  if (horiz_mode <= ONETWO)
    cpi->common.horiz_scale = horiz_mode;
  else
    return -1;

  if (vert_mode <= ONETWO)
    cpi->common.vert_scale  = vert_mode;
  else
    return -1;

  return 0;
}


int vp9_calc_ss_err(YV12_BUFFER_CONFIG *source, YV12_BUFFER_CONFIG *dest) {
  int i, j;
  int Total = 0;

  unsigned char *src = source->y_buffer;
  unsigned char *dst = dest->y_buffer;

  // Loop through the Y plane raw and reconstruction data summing (square differences)
  for (i = 0; i < source->y_height; i += 16) {
    for (j = 0; j < source->y_width; j += 16) {
      unsigned int sse;
      Total += vp9_mse16x16(src + j, source->y_stride, dst + j, dest->y_stride,
                            &sse);
    }

    src += 16 * source->y_stride;
    dst += 16 * dest->y_stride;
  }

  return Total;
}


int vp9_get_quantizer(VP9_PTR c) {
  VP9_COMP   *cpi = (VP9_COMP *) c;
  return cpi->common.base_qindex;
}