void analyzeLoudness(const std::vector<char> &data, int sampleRate, ChannelConfig chans,
SampleType type, std::vector<float> &out, float valuesPerSecond)
{
int samplesPerSegment = static_cast<int>(sampleRate / valuesPerSecond);
int numSamples = bytesToFrames(data.size(), chans, type);
int advance = framesToBytes(1, chans, type);
out.reserve(numSamples/samplesPerSegment);
int segment=0;
int sample=0;
while (segment < numSamples/samplesPerSegment)
{
float sum=0;
int samplesAdded = 0;
while (sample < numSamples && sample < (segment+1)*samplesPerSegment)
{
// get sample on a scale from -1 to 1
float value = 0;
if (type == SampleType_UInt8)
value = ((char)(data[sample*advance]^0x80))/128.f;
else if (type == SampleType_Int16)
{
value = *reinterpret_cast<const int16_t*>(&data[sample*advance]);
value /= float(std::numeric_limits<int16_t>::max());
}
else if (type == SampleType_Float32)
{
value = *reinterpret_cast<const float*>(&data[sample*advance]);
value = std::max(-1.f, std::min(1.f, value)); // Float samples *should* be scaled to [-1,1] already.
}
sum += value*value;
++samplesAdded;
++sample;
}
float rms = 0; // root mean square
if (samplesAdded > 0)
rms = std::sqrt(sum / samplesAdded);
out.push_back(rms);
++segment;
}
}
size_t bytesToFrames(size_t bytes, ChannelConfig config, SampleType type)
{
return bytes / framesToBytes(1, config, type);
}
