resampler.cc

// Copyright (c) 2010 The Chromium OS Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

#include <stdio.h>
#include <stdlib.h>

#include "log.h"
#include "resampler.h"
#include "libresample/libresample.h"

namespace speech_synthesis {

Resampler::Resampler(TtsDataReceiver *destination,
                     int source_rate,
                     int dest_rate,
                     int buffer_size)
    : destination_(destination),
      source_rate_(source_rate),
      dest_rate_(dest_rate),
      buffer_size_(buffer_size) {
  const int high_quality = 0;
  factor_ = dest_rate_ * 1.0 / source_rate_;
  resample_handle_ = resample_open(high_quality, factor_, factor_);
  in_floats_ = new float[buffer_size_];
  out_floats_ = new float[buffer_size_];
  out_int16s_ = new int16_t[buffer_size_];
}

Resampler::~Resampler() {
  resample_close(resample_handle_);
  delete[] in_floats_;
  delete[] out_floats_;
  delete[] out_int16s_;
}

tts_callback_status Resampler::Receive(int rate,
                                       int num_channels,
                                       const int16_t* data,
                                       int num_samples,
                                       tts_synth_status synth_status) {
  if (num_samples > buffer_size_) {
    LOG(ERROR) << "Got " << num_samples << " samples, buffer size is " <<
            buffer_size_ << "\n";
    exit(-1);
  }

  if (rate != source_rate_) {
    LOG(ERROR) << "Got input rate of " << rate << " expected " <<
        source_rate_ << "\n";
    exit(-1);
  }

  for (int i = 0; i < num_samples; i++) {
    in_floats_[i] = data[i];
  }

  int last_flag = (synth_status == TTS_SYNTH_DONE? 1 : 0);

  int input_index = 0;
  while (input_index < num_samples) {
    int in_buffer_used;
    int out_samples = resample_process(resample_handle_,
                                       factor_,
                                       &in_floats_[input_index],
                                       num_samples - input_index,
                                       last_flag,
                                       &in_buffer_used,
                                       out_floats_,
                                       buffer_size_);
    input_index += in_buffer_used;

    for (int i = 0; i < out_samples; i++) {
      int value = static_cast<int>(out_floats_[i]);
      if (value > 32767)
        value = 32767;
      if (value < -32768)
        value = -32768;
      out_int16s_[i] = static_cast<int16_t>(value);
    }

    tts_synth_status status;
    if (synth_status == TTS_SYNTH_DONE && input_index == num_samples) {
      status = TTS_SYNTH_DONE;
    } else {
      status = TTS_SYNTH_PENDING;
    }

    tts_callback_status callback_status = destination_->Receive(
        dest_rate_, 1, out_int16s_, out_samples, status);
    if (callback_status != TTS_CALLBACK_CONTINUE) {
      return callback_status;
    }
  }

  return TTS_CALLBACK_CONTINUE;
}

}  // namespace speech_synthesis