Demo: Musical Sound Separation Based on Binary Time-Frequency Masking
The signals are generated from MIDI data of quartet composed by J. S. Bach using instrument samples from RWC database. The tenor and the alto line are mixed with equal energy. The first five-second signals are used for testing. For more details about the separation system, see: Y. Li and D. L. Wang, Musical Sound Separation Based on Binary Time-Frequency Masking, EURASIP Journal on Audio, Speech, and Music Processing, vol. 2009, Article ID 130567.
กก
| Instruments | Mixture (0 dB) | Original | Separated | Comments |
| Flute |
|
|
|
good separation with little artifacts |
| Clarinet |
|
|
||
| Clarinet |
|
|
|
good separation with some artifacts |
| Trumpet |
|
|
||
| Clarinet |
|
|
|
poor separation result |
| Trumpet |
|
|