I need to convert one dimensional array of size N to two dimensional array of size A*B > N. Let us take such case:
int oneDimensionalArray[6] = {7, 8, 10, 11, 12, 15};
//then the second array would be
int twoDimensionalArray[2][4] = {{7, 8, 10, 11},
{10, 11, 12, 15}};
This is used in so called overlap-add method used in digital sound processing. I have tried this approach which gives improper results:
for(unsigned long i = 0; i < amountOfWindows; i++)
{
for(unsigned long j = hopSize; j < windowLength; j++)
{
//buffer without the overlapping
if( (i * amountOfWindows + j) >= bufferLength)
break;
windowedBuffer[i][j] = unwindowedBuffer[i * amountOfWindows + j];
}
}
for(unsigned long i = 1; i < amountOfWindows; i++ )
{
for(unsigned long j = 0; j < hopSize; j++)
{
// Filling the overlapping region
windowedBuffer[i][j] = windowedBuffer[i-1][windowLength - hopSize + i];
}
}
I've also tried finding the relation using the modulo operation but I can't find the right one. This is the one that I've tried:
windowedBuffer[m][n % (windowLength - hopSize)] = unwindowedBuffer[n];
Since you already know hopSize
(from your comment), what you want is simply:
for (size_t i = 0; i < amountOfWindows; ++i) {
for (size_t j = 0; j < windowLength; ++j) {
windowedBuffer[i][j] = unwindowedBuffer[i * hopSize + j];
}
}
Where amountOfWindows
, windowLength
and hopSize
are you parameters (resp. 2, 4 and 2 in your example).