TS-AudioToMIDI Web Online Help - Performing recognition of pre-recorded audio

TS-AudioToMIDI allows recognizing music recorded as digital audio file (plain or compressed wave will do as well as audio CD track. Read Supported formats for details). You can select sound file to recognize by pressing "Open Wave File" button at upper left corner of TS-AudioToMIDI window. After wave file is chosen, "Play/Stop Wave File" button (next to "Open Wave File") becomes active, as well as "Convert Wave File" button. You can listen to the wave file you have opened and convert it with the help of these two buttons. Most controls in the main window serve to set up recognition settings and MIDI options. If you are new to TS-AudioToMIDI, first convert file chosen without changing default settings by simply pressing "Convert Wave File" button. After this transcription starts. It may take a while depending on the length of the recognized file and your computer configuration (see System Requirements). During transcription process most controls will be disabled and the upper time counter at the right of MIDI sequencer window will indicate the progress of transcribing. You can stop the process by pressing "Convert Wave File" once more. After transcription process finishes, MIDI file created will be automatically loaded to MIDI sequencer placed at the bottom of the window. You can listen to the result of recognition, but saving it as MIDI file is available only in registered version.

For most musical records default settings don't not provide the best recognition accuracy possible. Adjusting certain parameters will improve it. A very powerful function allowing to detect these parameters is Synchronous playback function. When both digital audio and MIDI file are loaded, "Sync Play WAVE and MIDI" button becomes active. Press it to hear initial sound and transcription result simultaneously. This function helps to conveniently compare both records and see, whether significant notes were sipped or, in opposite, waste ones were added. Each spectral peak can be interpreted as a note or rejected. TS-AudioToMIDI allows for two criteria when making such a decision. The first is note duration. By setting up Minimal Note duration value at the lower right corner of main window you indicate that all notes shorter than this value will be rejected by the program. Second criterion is note's amplitude. This parameter is controlled by Sharpness and Selectivity together with Threshold and Noise Gate sliders.

Several Recognition Methods are provided by TS-AudioToMIDI. Each has it's advantages and disadvantages. Monophonic methods provide recognition of one (the loudest) voice in melody but with great accuracy. Polyphonic method transcribes several voices. Minimal amplitude to be detected as note is controlled by Threshold and Noise Gate controls. Selecting appropriate recognition algorithm for each particular musical piece will improve transcribing accuracy greatly.

Another two powerful instruments for increasing recognition accuracy are Graphic Equalizer and Filter. First allows smooth control of sound level for each frequency. Second is intended for forbidding particular notes to appear in output MIDI file (the same effect as tuning corresponding Equalizer slider to zero). Filter window is positioned under the fingerboard. Pressing the keys of the fingerboard with left mouse button you will hear corresponding notes. This helps to detect straight away, which frequencies need to be weaken, and which require amplification.

Related topics:
TS-AudioToMIDI Main Window