Skip to content

How to normalize audio

One channel

Input/Output data: See how to pass data to KFR 

univector<float> audio;

Code 

audio /= absmaxof(audio);

  1. absmaxof function calculates the absolute maximum of the given array (max(abs(x[i])...))
  2. operator/= performs inplace division for array

Equivalent pure C++ code 

float absmax = 0;
for (size_t i = 0; i < audio.size(); ++i)
{
    absmax = std::max(absmax, std::abs(audio[i]));
}
for (size_t i = 0; i < audio.size(); ++i)
{
    audio[i] = audio[i] / absmax;
}

Stereo audio

Input/Output data: 

univector<univector<float>, 2> stereo;

Code 

const float peak = absmaxof(concatenate(stereo[0], stereo[1]));
unpack(stereo[0], stereo[1]) = pack(stereo[0], stereo[1]) / pack(peak, peak);

  1. concatenate function concatenates two arrays as if they were written sequentially in memory (but without any actual memory copies)
  2. absmaxof function calculates the absolute maximum of the concatenated arrays
  3. pack function combines two arrays as if it were a single array containing pairs of values
  4. operator/ divides pairs by the peak value
  5. unpack function unpacks the pairs to two target arrays