The spectral oscillator analyzes the spectrum of the loaded sample, that is, the progression of the frequencies, amplitudes, and phases from the sample start to the end of the sample. The analyzed spectrum is used to generate a waveform for playback.
During this re-synthesis, the sample is split into its individual frequency components and overtones. These can then be filtered, amplified, or adjusted in relation to each other. The spectrum becomes independent from the original temporal progression of the sample, which means that the sound is not played faster or more slowly for higher or lower notes, but instead keeps its timing. This allows you to change the playback speed without affecting the pitch.
You can also slow down the playback speed up to the point where only a static waveform is played back. When playing pitches other than the original sample pitch, it is possible to preserve the natural formants of the sound or to tweak them independently to create more artificial variations.