First it produces some kind of spectra-like data, this is called further the "WAVE->Pitch" conversion. Now there are two different algorithms implementing this step, each of them has its own advantages and disadvantages.
Then it multiplies that spectral data by coefficients of "Equalizer", making some pitches louder and some pitches more quiet (This is only for Widi for Windows version).
The third step is an actual retrieval of notes from the spectra, it is called "Pitch->MIDI" process and there are also two algorithms making that.
The final step is producing the MIDI file structure, and you can choose some attributes of the output MIDI file.
Nowadays investigations of more complex schemes are taking place, it seems that they will allow to differentiate musical instruments and make fewer errors.

1. gets as input both compressed and uncompressed WAVE, MPEG Layer 1, 2 and 3 and CD tracks.
2. produces multi-track volume differentiated MIDI files..
3. uses simple harmonic model to eliminate wrong notes.
4. allows automatic fine tuning to the key of source file.
5. does not differentiate instruments in source file, but allows you to use any instrument in the output MIDI file.
6. can produce intermediate GS files for quicker operation (See GS files for details).
7. pitches to transcribe are selectable (See Equalizer).
8. subject to tuning:
9. algorithm selection.
10. minimal duration of notes (wrong short notes can be cut off).
11. number of voices.
12. time resolution (usually higher than minimal note duration).
13. bass note shifting.
14. length of input file is not limited.

wave to midi,wave midi converter,mp3 to mid converter,wave, midi, mp3, cd to mid,cda to midi, mid, convertor, wave to midi convert, wave to mid, mp3 to midi, wave to mid, midi guitar, wave2mid, guitar midi, wave2midi, mp3 2 mid, audio to midi, wave2mid