The task of separating a mix of several inter-weaving melodies from a mono recording into multiple tracks is attempted by filtering in the spectral domain. The transcribed score is provided in MIDI format a priori. In each time frame a filter is constructed for each instrument in the mix, whose effect is to filter out all harmonics of that instrument from the DFT spectrum. The complication of overlapping harmonics arising from separate notes is discussed and two filter shapes that were found to be fairly successful at separating overlapping harmonics are presented. In comparing the separated audio tracks to the original instrumental parts, signalto-residual ratios (SRR’s) in excess of 20 dB have been achieved. Audio demonstrations are on the internet [1].