FFT Settings

When performing Fast Fourier Transform (FFT) processing, the sound is divided into small blocks and processed one block at a time. The FFT size value controls the size of the these blocks. The number of samples to process is calculated by taking the value as a power of 2. A value of 10 gives 2 to the 10th power, or 1024 samples. By increasing the number of samples, frequencies are processed at a higher resolution, which helps to eliminate chirping and other mechanical sound distortions, but it tends to add more echo. Usually values from 11 to 12 give the best trade-off between distortions and echo.

To smooth out transitions from one block to the next, it is necessary to overlap blocks. The Overlap value controls how much of the FFT analysis of one block overlaps the next. A high value makes the transition between each block smoother. It also requires more processing time since overlapping samples are recalculated several times. A low value may result in rougher transitions, but processes faster. For complex audio and tempo or pitch modifying effects, higher values may give better quality.