xref: /aosp_15_r20/external/aac/libSBRdec/src/psdec.cpp (revision e54365361535b070c2db7374cec45c159c7d0e7a)

/* -----------------------------------------------------------------------------
Software License for The Fraunhofer FDK AAC Codec Library for Android

© Copyright  1995 - 2018 Fraunhofer-Gesellschaft zur Förderung der angewandten
Forschung e.V. All rights reserved.

 1.    INTRODUCTION
The Fraunhofer FDK AAC Codec Library for Android ("FDK AAC Codec") is software
that implements the MPEG Advanced Audio Coding ("AAC") encoding and decoding
scheme for digital audio. This FDK AAC Codec software is intended to be used on
a wide variety of Android devices.

AAC's HE-AAC and HE-AAC v2 versions are regarded as today's most efficient
general perceptual audio codecs. AAC-ELD is considered the best-performing
full-bandwidth communications codec by independent studies and is widely
deployed. AAC has been standardized by ISO and IEC as part of the MPEG
specifications.

Patent licenses for necessary patent claims for the FDK AAC Codec (including
those of Fraunhofer) may be obtained through Via Licensing
(www.vialicensing.com) or through the respective patent owners individually for
the purpose of encoding or decoding bit streams in products that are compliant
with the ISO/IEC MPEG audio standards. Please note that most manufacturers of
Android devices already license these patent claims through Via Licensing or
directly from the patent owners, and therefore FDK AAC Codec software may
already be covered under those patent licenses when it is used for those
licensed purposes only.

Commercially-licensed AAC software libraries, including floating-point versions
with enhanced sound quality, are also available from Fraunhofer. Users are
encouraged to check the Fraunhofer website for additional applications
information and documentation.

2.    COPYRIGHT LICENSE

Redistribution and use in source and binary forms, with or without modification,
are permitted without payment of copyright license fees provided that you
satisfy the following conditions:

You must retain the complete text of this software license in redistributions of
the FDK AAC Codec or your modifications thereto in source code form.

You must retain the complete text of this software license in the documentation
and/or other materials provided with redistributions of the FDK AAC Codec or
your modifications thereto in binary form. You must make available free of
charge copies of the complete source code of the FDK AAC Codec and your
modifications thereto to recipients of copies in binary form.

The name of Fraunhofer may not be used to endorse or promote products derived
from this library without prior written permission.

You may not charge copyright license fees for anyone to use, copy or distribute
the FDK AAC Codec software or your modifications thereto.

Your modified versions of the FDK AAC Codec must carry prominent notices stating
that you changed the software and the date of any change. For modified versions
of the FDK AAC Codec, the term "Fraunhofer FDK AAC Codec Library for Android"
must be replaced by the term "Third-Party Modified Version of the Fraunhofer FDK
AAC Codec Library for Android."

3.    NO PATENT LICENSE

NO EXPRESS OR IMPLIED LICENSES TO ANY PATENT CLAIMS, including without
limitation the patents of Fraunhofer, ARE GRANTED BY THIS SOFTWARE LICENSE.
Fraunhofer provides no warranty of patent non-infringement with respect to this
software.

You may use this FDK AAC Codec software or modifications thereto only for
purposes that are authorized by appropriate patent licenses.

4.    DISCLAIMER

This FDK AAC Codec software is provided by Fraunhofer on behalf of the copyright
holders and contributors "AS IS" and WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES,
including but not limited to the implied warranties of merchantability and
fitness for a particular purpose. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
CONTRIBUTORS BE LIABLE for any direct, indirect, incidental, special, exemplary,
or consequential damages, including but not limited to procurement of substitute
goods or services; loss of use, data, or profits, or business interruption,
however caused and on any theory of liability, whether in contract, strict
liability, or tort (including negligence), arising in any way out of the use of
this software, even if advised of the possibility of such damage.

5.    CONTACT INFORMATION

Fraunhofer Institute for Integrated Circuits IIS
Attention: Audio and Multimedia Departments - FDK AAC LL
Am Wolfsmantel 33
91058 Erlangen, Germany

www.iis.fraunhofer.de/amm
[email protected]
----------------------------------------------------------------------------- */

/**************************** SBR decoder library ******************************

   Author(s):

   Description:

*******************************************************************************/

/*!
  \file
  \brief  parametric stereo decoder
*/

#include "psdec.h"

#include "FDK_bitbuffer.h"

#include "sbr_rom.h"
#include "sbr_ram.h"

#include "FDK_tools_rom.h"

#include "genericStds.h"

#include "FDK_trigFcts.h"

/********************************************************************/
/*                       MLQUAL DEFINES                             */
/********************************************************************/

#define FRACT_ZERO FRACT_BITS - 1
/********************************************************************/

SBR_ERROR ResetPsDec(HANDLE_PS_DEC h_ps_d);

/***** HELPERS *****/

/***************************************************************************/
/*!
  \brief  Creates one instance of the PS_DEC struct

  \return Error info

****************************************************************************/
int CreatePsDec(HANDLE_PS_DEC *h_PS_DEC, /*!< pointer to the module state */
                int aacSamplesPerFrame) {
  SBR_ERROR errorInfo = SBRDEC_OK;
  HANDLE_PS_DEC h_ps_d;
  int i;

  if (*h_PS_DEC == NULL) {
    /* Get ps dec ram */
    h_ps_d = GetRam_ps_dec();
    if (h_ps_d == NULL) {
      goto bail;
    }
  } else {
    /* Reset an open instance */
    h_ps_d = *h_PS_DEC;
  }

  /*
   * Create Analysis Hybrid filterbank.
   */
  FDKhybridAnalysisOpen(&h_ps_d->specificTo.mpeg.hybridAnalysis,
                        h_ps_d->specificTo.mpeg.pHybridAnaStatesLFdmx,
                        sizeof(h_ps_d->specificTo.mpeg.pHybridAnaStatesLFdmx),
                        NULL, 0);

  /* initialisation */
  switch (aacSamplesPerFrame) {
    case 960:
      h_ps_d->noSubSamples = 30; /* col */
      break;
    case 1024:
      h_ps_d->noSubSamples = 32; /* col */
      break;
    default:
      h_ps_d->noSubSamples = -1;
      break;
  }

  if (h_ps_d->noSubSamples > MAX_NUM_COL || h_ps_d->noSubSamples <= 0) {
    goto bail;
  }
  h_ps_d->noChannels = NO_QMF_CHANNELS; /* row */

  h_ps_d->psDecodedPrv = 0;
  h_ps_d->procFrameBased = -1;
  for (i = 0; i < (1) + 1; i++) {
    h_ps_d->bPsDataAvail[i] = ppt_none;
  }
  {
    int error;
    error = FDKdecorrelateOpen(&(h_ps_d->specificTo.mpeg.apDecor),
                               h_ps_d->specificTo.mpeg.decorrBufferCplx,
                               (2 * ((825) + (373))));
    if (error) goto bail;
  }

  for (i = 0; i < (1) + 1; i++) {
    FDKmemclear(&h_ps_d->bsData[i].mpeg, sizeof(MPEG_PS_BS_DATA));
  }

  errorInfo = ResetPsDec(h_ps_d);

  if (errorInfo != SBRDEC_OK) goto bail;

  *h_PS_DEC = h_ps_d;

  return 0;

bail:
  if (h_ps_d != NULL) {
    DeletePsDec(&h_ps_d);
  }

  return -1;
} /*END CreatePsDec */

/***************************************************************************/
/*!
  \brief  Delete one instance of the PS_DEC struct

  \return Error info

****************************************************************************/
int DeletePsDec(HANDLE_PS_DEC *h_PS_DEC) /*!< pointer to the module state */
{
  if (*h_PS_DEC == NULL) {
    return -1;
  }

  {
    HANDLE_PS_DEC h_ps_d = *h_PS_DEC;
    FDKdecorrelateClose(&(h_ps_d->specificTo.mpeg.apDecor));
  }

  FreeRam_ps_dec(h_PS_DEC);

  return 0;
} /*END DeletePsDec */

/***************************************************************************/
/*!
  \brief resets some values of the PS handle to default states

  \return

****************************************************************************/
SBR_ERROR ResetPsDec(HANDLE_PS_DEC h_ps_d) /*!< pointer to the module state */
{
  SBR_ERROR errorInfo = SBRDEC_OK;
  INT i;

  /* explicitly init state variables to safe values (until first ps header
   * arrives) */

  h_ps_d->specificTo.mpeg.lastUsb = 0;

  /*
   * Initialize Analysis Hybrid filterbank.
   */
  FDKhybridAnalysisInit(&h_ps_d->specificTo.mpeg.hybridAnalysis, THREE_TO_TEN,
                        NO_QMF_BANDS_HYBRID20, NO_QMF_BANDS_HYBRID20, 1);

  /*
   * Initialize Synthesis Hybrid filterbank.
   */
  for (i = 0; i < 2; i++) {
    FDKhybridSynthesisInit(&h_ps_d->specificTo.mpeg.hybridSynthesis[i],
                           THREE_TO_TEN, NO_QMF_CHANNELS, NO_QMF_CHANNELS);
  }
  {
    INT error;
    error = FDKdecorrelateInit(&h_ps_d->specificTo.mpeg.apDecor, 71, DECORR_PS,
                               DUCKER_AUTOMATIC, 0, 0, 0, 0, 1, /* isLegacyPS */
                               1);
    if (error) return SBRDEC_NOT_INITIALIZED;
  }

  for (i = 0; i < NO_IID_GROUPS; i++) {
    h_ps_d->specificTo.mpeg.h11rPrev[i] = FL2FXCONST_DBL(0.5f);
    h_ps_d->specificTo.mpeg.h12rPrev[i] = FL2FXCONST_DBL(0.5f);
  }

  FDKmemclear(h_ps_d->specificTo.mpeg.h21rPrev,
              sizeof(h_ps_d->specificTo.mpeg.h21rPrev));
  FDKmemclear(h_ps_d->specificTo.mpeg.h22rPrev,
              sizeof(h_ps_d->specificTo.mpeg.h22rPrev));

  return errorInfo;
}

/***************************************************************************/
/*!
  \brief  Feed delaylines when parametric stereo is switched on.
  \return
****************************************************************************/
void PreparePsProcessing(HANDLE_PS_DEC h_ps_d,
                         const FIXP_DBL *const *const rIntBufferLeft,
                         const FIXP_DBL *const *const iIntBufferLeft,
                         const int scaleFactorLowBand) {
  if (h_ps_d->procFrameBased ==
      1) /* If we have switched from frame to slot based processing  */
  {      /* fill hybrid delay buffer.                                */
    int i, j;

    for (i = 0; i < HYBRID_FILTER_DELAY; i++) {
      FIXP_DBL qmfInputData[2][NO_QMF_BANDS_HYBRID20];
      FIXP_DBL hybridOutputData[2][NO_SUB_QMF_CHANNELS];

      for (j = 0; j < NO_QMF_BANDS_HYBRID20; j++) {
        qmfInputData[0][j] =
            scaleValue(rIntBufferLeft[i][j], scaleFactorLowBand);
        qmfInputData[1][j] =
            scaleValue(iIntBufferLeft[i][j], scaleFactorLowBand);
      }

      FDKhybridAnalysisApply(&h_ps_d->specificTo.mpeg.hybridAnalysis,
                             qmfInputData[0], qmfInputData[1],
                             hybridOutputData[0], hybridOutputData[1]);
    }
    h_ps_d->procFrameBased = 0; /* switch to slot based processing. */

  } /* procFrameBased==1 */
}

void initSlotBasedRotation(
    HANDLE_PS_DEC h_ps_d, /*!< pointer to the module state */
    int env, int usb) {
  INT group = 0;
  INT bin = 0;
  INT noIidSteps;

  FIXP_SGL invL;
  FIXP_DBL ScaleL, ScaleR;
  FIXP_DBL Alpha, Beta;
  FIXP_DBL h11r, h12r, h21r, h22r;

  const FIXP_DBL *PScaleFactors;

  if (h_ps_d->bsData[h_ps_d->processSlot].mpeg.bFineIidQ) {
    PScaleFactors = ScaleFactorsFine; /* values are shiftet right by one */
    noIidSteps = NO_IID_STEPS_FINE;
  } else {
    PScaleFactors = ScaleFactors; /* values are shiftet right by one */
    noIidSteps = NO_IID_STEPS;
  }

  /* dequantize and decode */
  for (group = 0; group < NO_IID_GROUPS; group++) {
    bin = bins2groupMap20[group];

    /*!
    <h3> type 'A' rotation </h3>
    mixing procedure R_a, used in baseline version<br>

     Scale-factor vectors c1 and c2 are precalculated in initPsTables () and
    stored in scaleFactors[] and scaleFactorsFine[] = pScaleFactors []. From the
    linearized IID parameters (intensity differences), two scale factors are
     calculated. They are used to obtain the coefficients h11... h22.
    */

    /* ScaleR and ScaleL are scaled by 1 shift right */

    ScaleR = PScaleFactors[noIidSteps + h_ps_d->specificTo.mpeg.pCoef
                                            ->aaIidIndexMapped[env][bin]];
    ScaleL = PScaleFactors[noIidSteps - h_ps_d->specificTo.mpeg.pCoef
                                            ->aaIidIndexMapped[env][bin]];

    Beta = fMult(
        fMult(Alphas[h_ps_d->specificTo.mpeg.pCoef->aaIccIndexMapped[env][bin]],
              (ScaleR - ScaleL)),
        FIXP_SQRT05);
    Alpha =
        Alphas[h_ps_d->specificTo.mpeg.pCoef->aaIccIndexMapped[env][bin]] >> 1;

    /* Alpha and Beta are now both scaled by 2 shifts right */

    /* calculate the coefficients h11... h22 from scale-factors and ICC
     * parameters */

    /* h values are scaled by 1 shift right */
    {
      FIXP_DBL trigData[4];

      inline_fixp_cos_sin(Beta + Alpha, Beta - Alpha, 2, trigData);
      h11r = fMult(ScaleL, trigData[0]);
      h12r = fMult(ScaleR, trigData[2]);
      h21r = fMult(ScaleL, trigData[1]);
      h22r = fMult(ScaleR, trigData[3]);
    }
    /*****************************************************************************************/
    /* Interpolation of the matrices H11... H22: */
    /*                                                                                       */
    /* H11(k,n) = H11(k,n[e]) + (n-n[e]) * (H11(k,n[e+1] - H11(k,n[e])) /
     * (n[e+1] - n[e])    */
    /* ... */
    /*****************************************************************************************/

    /* invL = 1/(length of envelope) */
    invL = FX_DBL2FX_SGL(GetInvInt(
        h_ps_d->bsData[h_ps_d->processSlot].mpeg.aEnvStartStop[env + 1] -
        h_ps_d->bsData[h_ps_d->processSlot].mpeg.aEnvStartStop[env]));

    h_ps_d->specificTo.mpeg.pCoef->H11r[group] =
        h_ps_d->specificTo.mpeg.h11rPrev[group];
    h_ps_d->specificTo.mpeg.pCoef->H12r[group] =
        h_ps_d->specificTo.mpeg.h12rPrev[group];
    h_ps_d->specificTo.mpeg.pCoef->H21r[group] =
        h_ps_d->specificTo.mpeg.h21rPrev[group];
    h_ps_d->specificTo.mpeg.pCoef->H22r[group] =
        h_ps_d->specificTo.mpeg.h22rPrev[group];

    h_ps_d->specificTo.mpeg.pCoef->DeltaH11r[group] =
        fMult(h11r - h_ps_d->specificTo.mpeg.pCoef->H11r[group], invL);
    h_ps_d->specificTo.mpeg.pCoef->DeltaH12r[group] =
        fMult(h12r - h_ps_d->specificTo.mpeg.pCoef->H12r[group], invL);
    h_ps_d->specificTo.mpeg.pCoef->DeltaH21r[group] =
        fMult(h21r - h_ps_d->specificTo.mpeg.pCoef->H21r[group], invL);
    h_ps_d->specificTo.mpeg.pCoef->DeltaH22r[group] =
        fMult(h22r - h_ps_d->specificTo.mpeg.pCoef->H22r[group], invL);

    /* update prev coefficients for interpolation in next envelope */

    h_ps_d->specificTo.mpeg.h11rPrev[group] = h11r;
    h_ps_d->specificTo.mpeg.h12rPrev[group] = h12r;
    h_ps_d->specificTo.mpeg.h21rPrev[group] = h21r;
    h_ps_d->specificTo.mpeg.h22rPrev[group] = h22r;

  } /* group loop */
}

static const UCHAR groupTable[NO_IID_GROUPS + 1] = {
    0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10, 11,
    12, 13, 14, 15, 16, 18, 21, 25, 30, 42, 71};

static void applySlotBasedRotation(
    HANDLE_PS_DEC h_ps_d, /*!< pointer to the module state */

    FIXP_DBL *mHybridRealLeft, /*!< hybrid values real left  */
    FIXP_DBL *mHybridImagLeft, /*!< hybrid values imag left  */

    FIXP_DBL *mHybridRealRight, /*!< hybrid values real right  */
    FIXP_DBL *mHybridImagRight  /*!< hybrid values imag right  */
) {
  INT group;
  INT subband;

  /**********************************************************************************************/
  /*!
  <h2> Mapping </h2>

  The number of stereo bands that is actually used depends on the number of
  availble parameters for IID and ICC: <pre> nr. of IID para.| nr. of ICC para.
  | nr. of Stereo bands
   ----------------|------------------|-------------------
     10,20         |     10,20        |        20
     10,20         |     34           |        34
     34            |     10,20        |        34
     34            |     34           |        34
  </pre>
  In the case the number of parameters for IIS and ICC differs from the number
  of stereo bands, a mapping from the lower number to the higher number of
  parameters is applied. Index mapping of IID and ICC parameters is already done
  in psbitdec.cpp. Further mapping is not needed here in baseline version.
  **********************************************************************************************/

  /************************************************************************************************/
  /*!
  <h2> Mixing </h2>

  To generate the QMF subband signals for the subband samples n = n[e]+1 ,,,
  n_[e+1] the parameters at position n[e] and n[e+1] are required as well as the
  subband domain signals s_k(n) and d_k(n) for n = n[e]+1... n_[e+1]. n[e]
  represents the start position for envelope e. The border positions n[e] are
  handled in DecodePS().

  The stereo sub subband signals are constructed as:
  <pre>
  l_k(n) = H11(k,n) s_k(n) + H21(k,n) d_k(n)
  r_k(n) = H21(k,n) s_k(n) + H22(k,n) d_k(n)
  </pre>
  In order to obtain the matrices H11(k,n)... H22 (k,n), the vectors h11(b)...
  h22(b) need to be calculated first (b: parameter index). Depending on ICC mode
  either mixing procedure R_a or R_b is used for that. For both procedures, the
  parameters for parameter position n[e+1] is used.
  ************************************************************************************************/

  /************************************************************************************************/
  /*!
  <h2>Phase parameters </h2>
  With disabled phase parameters (which is the case in baseline version), the
  H-matrices are just calculated by:

  <pre>
  H11(k,n[e+1] = h11(b(k))
  (...)
  b(k): parameter index according to mapping table
  </pre>

  <h2>Processing of the samples in the sub subbands </h2>
  this loop includes the interpolation of the coefficients Hxx
  ************************************************************************************************/

  /******************************************************/
  /* construct stereo sub subband signals according to: */
  /*                                                    */
  /* l_k(n) = H11(k,n) s_k(n) + H21(k,n) d_k(n)         */
  /* r_k(n) = H12(k,n) s_k(n) + H22(k,n) d_k(n)         */
  /******************************************************/
  PS_DEC_COEFFICIENTS *pCoef = h_ps_d->specificTo.mpeg.pCoef;

  for (group = 0; group < NO_IID_GROUPS; group++) {
    pCoef->H11r[group] += pCoef->DeltaH11r[group];
    pCoef->H12r[group] += pCoef->DeltaH12r[group];
    pCoef->H21r[group] += pCoef->DeltaH21r[group];
    pCoef->H22r[group] += pCoef->DeltaH22r[group];

    const int start = groupTable[group];
    const int stop = groupTable[group + 1];
    for (subband = start; subband < stop; subband++) {
      FIXP_DBL tmpLeft =
          fMultAdd(fMultDiv2(pCoef->H11r[group], mHybridRealLeft[subband]),
                   pCoef->H21r[group], mHybridRealRight[subband]);
      FIXP_DBL tmpRight =
          fMultAdd(fMultDiv2(pCoef->H12r[group], mHybridRealLeft[subband]),
                   pCoef->H22r[group], mHybridRealRight[subband]);
      mHybridRealLeft[subband] = tmpLeft;
      mHybridRealRight[subband] = tmpRight;

      tmpLeft =
          fMultAdd(fMultDiv2(pCoef->H11r[group], mHybridImagLeft[subband]),
                   pCoef->H21r[group], mHybridImagRight[subband]);
      tmpRight =
          fMultAdd(fMultDiv2(pCoef->H12r[group], mHybridImagLeft[subband]),
                   pCoef->H22r[group], mHybridImagRight[subband]);
      mHybridImagLeft[subband] = tmpLeft;
      mHybridImagRight[subband] = tmpRight;
    } /* subband */
  }
}

/***************************************************************************/
/*!
  \brief  Applies IID, ICC, IPD and OPD parameters to the current frame.

  \return none

****************************************************************************/
void ApplyPsSlot(
    HANDLE_PS_DEC h_ps_d,      /*!< handle PS_DEC*/
    FIXP_DBL **rIntBufferLeft, /*!< real bands left qmf channel (38x64)  */
    FIXP_DBL **iIntBufferLeft, /*!< imag bands left qmf channel (38x64)  */
    FIXP_DBL *rIntBufferRight, /*!< real bands right qmf channel (38x64) */
    FIXP_DBL *iIntBufferRight, /*!< imag bands right qmf channel (38x64) */
    const int scaleFactorLowBand_no_ov, const int scaleFactorLowBand,
    const int scaleFactorHighBand, const int lsb, const int usb) {
/*!
The 64-band QMF representation of the monaural signal generated by the SBR tool
is used as input of the PS tool. After the PS processing, the outputs of the
left and right hybrid synthesis filterbanks are used to generate the stereo
output signal.

<pre>

           -------------            ----------            -------------
          | Hybrid      | M_n[k,m] |          | L_n[k,m] | Hybrid      | l[n]
 m[n] --->| analysis    |--------->|          |--------->| synthesis   |----->
           -------------           | Stereo   |           -------------
                 |                 | recon-   |
                 |                 | stuction |
                \|/                |          |
           -------------           |          |
          | De-         | D_n[k,m] |          |
          | correlation |--------->|          |
           -------------           |          |           -------------
                                   |          | R_n[k,m] | Hybrid      | r[n]
                                   |          |--------->| synthesis   |----->
 IID, ICC ------------------------>|          |          | filter bank |
(IPD, OPD)                          ----------            -------------

m[n]:      QMF represantation of the mono input
M_n[k,m]:  (sub-)sub-band domain signals of the mono input
D_n[k,m]:  decorrelated (sub-)sub-band domain signals
L_n[k,m]:  (sub-)sub-band domain signals of the left output
R_n[k,m]:  (sub-)sub-band domain signals of the right output
l[n],r[n]: left/right output signals

</pre>
*/
#define NO_HYBRID_DATA_BANDS (71)

  int i;
  FIXP_DBL qmfInputData[2][NO_QMF_BANDS_HYBRID20];
  FIXP_DBL *hybridData[2][2];
  C_ALLOC_SCRATCH_START(pHybridData, FIXP_DBL, 4 * NO_HYBRID_DATA_BANDS);

  hybridData[0][0] =
      pHybridData + 0 * NO_HYBRID_DATA_BANDS; /* left real hybrid data */
  hybridData[0][1] =
      pHybridData + 1 * NO_HYBRID_DATA_BANDS; /* left imag hybrid data */
  hybridData[1][0] =
      pHybridData + 2 * NO_HYBRID_DATA_BANDS; /* right real hybrid data */
  hybridData[1][1] =
      pHybridData + 3 * NO_HYBRID_DATA_BANDS; /* right imag hybrid data */

  /*!
  Hybrid analysis filterbank:
  The lower 3 (5) of the 64 QMF subbands are further split to provide better
  frequency resolution. for PS processing. For the 10 and 20 stereo bands
  configuration, the QMF band H_0(w) is split up into 8 (sub-) sub-bands and the
  QMF bands H_1(w) and H_2(w) are spit into 2 (sub-) 4th. (See figures 8.20
  and 8.22 of ISO/IEC 14496-3:2001/FDAM 2:2004(E) )
  */

  /*
   * Hybrid analysis.
   */

  /* Get qmf input data and apply descaling */
  for (i = 0; i < NO_QMF_BANDS_HYBRID20; i++) {
    qmfInputData[0][i] = scaleValue(rIntBufferLeft[HYBRID_FILTER_DELAY][i],
                                    scaleFactorLowBand_no_ov);
    qmfInputData[1][i] = scaleValue(iIntBufferLeft[HYBRID_FILTER_DELAY][i],
                                    scaleFactorLowBand_no_ov);
  }

  /* LF - part */
  FDKhybridAnalysisApply(&h_ps_d->specificTo.mpeg.hybridAnalysis,
                         qmfInputData[0], qmfInputData[1], hybridData[0][0],
                         hybridData[0][1]);

  /* HF - part */
  /* bands up to lsb */
  scaleValues(&hybridData[0][0][NO_SUB_QMF_CHANNELS - 2],
              &rIntBufferLeft[0][NO_QMF_BANDS_HYBRID20],
              lsb - NO_QMF_BANDS_HYBRID20, scaleFactorLowBand);
  scaleValues(&hybridData[0][1][NO_SUB_QMF_CHANNELS - 2],
              &iIntBufferLeft[0][NO_QMF_BANDS_HYBRID20],
              lsb - NO_QMF_BANDS_HYBRID20, scaleFactorLowBand);

  /* bands from lsb to usb */
  scaleValues(&hybridData[0][0][lsb + (NO_SUB_QMF_CHANNELS - 2 -
                                       NO_QMF_BANDS_HYBRID20)],
              &rIntBufferLeft[0][lsb], usb - lsb, scaleFactorHighBand);
  scaleValues(&hybridData[0][1][lsb + (NO_SUB_QMF_CHANNELS - 2 -
                                       NO_QMF_BANDS_HYBRID20)],
              &iIntBufferLeft[0][lsb], usb - lsb, scaleFactorHighBand);

  /* bands from usb to NO_SUB_QMF_CHANNELS which should be zero for non-overlap
     slots but can be non-zero for overlap slots */
  FDKmemcpy(
      &hybridData[0][0]
                 [usb + (NO_SUB_QMF_CHANNELS - 2 - NO_QMF_BANDS_HYBRID20)],
      &rIntBufferLeft[0][usb], sizeof(FIXP_DBL) * (NO_QMF_CHANNELS - usb));
  FDKmemcpy(
      &hybridData[0][1]
                 [usb + (NO_SUB_QMF_CHANNELS - 2 - NO_QMF_BANDS_HYBRID20)],
      &iIntBufferLeft[0][usb], sizeof(FIXP_DBL) * (NO_QMF_CHANNELS - usb));

  /*!
  Decorrelation:
  By means of all-pass filtering and delaying, the (sub-)sub-band samples s_k(n)
  are converted into de-correlated (sub-)sub-band samples d_k(n).
  - k: frequency in hybrid spectrum
  - n: time index
  */

  FDKdecorrelateApply(&h_ps_d->specificTo.mpeg.apDecor,
                      &hybridData[0][0][0], /* left real hybrid data */
                      &hybridData[0][1][0], /* left imag hybrid data */
                      &hybridData[1][0][0], /* right real hybrid data */
                      &hybridData[1][1][0], /* right imag hybrid data */
                      0                     /* startHybBand */
  );

  /*!
  Stereo Processing:
  The sets of (sub-)sub-band samples s_k(n) and d_k(n) are processed according
  to the stereo cues which are defined per stereo band.
  */

  applySlotBasedRotation(h_ps_d,
                         &hybridData[0][0][0], /* left real hybrid data */
                         &hybridData[0][1][0], /* left imag hybrid data */
                         &hybridData[1][0][0], /* right real hybrid data */
                         &hybridData[1][1][0]  /* right imag hybrid data */
  );

  /*!
  Hybrid synthesis filterbank:
  The stereo processed hybrid subband signals l_k(n) and r_k(n) are fed into the
  hybrid synthesis filterbanks which are identical to the 64 complex synthesis
  filterbank of the SBR tool. The input to the filterbank are slots of 64 QMF
  samples. For each slot the filterbank outputs one block of 64 samples of one
  reconstructed stereo channel. The hybrid synthesis filterbank is computed
  seperatly for the left and right channel.
  */

  /*
   * Hybrid synthesis.
   */
  for (i = 0; i < 2; i++) {
    FDKhybridSynthesisApply(
        &h_ps_d->specificTo.mpeg.hybridSynthesis[i],
        hybridData[i][0], /* real hybrid data */
        hybridData[i][1], /* imag hybrid data */
        (i == 0) ? rIntBufferLeft[0]
                 : rIntBufferRight, /* output real qmf buffer */
        (i == 0) ? iIntBufferLeft[0]
                 : iIntBufferRight /* output imag qmf buffer */
    );
  }

  /* free temporary hybrid qmf values of one timeslot */
  C_ALLOC_SCRATCH_END(pHybridData, FIXP_DBL, 4 * NO_HYBRID_DATA_BANDS);

} /* END ApplyPsSlot */