This piece began with an impromptu vocal and keyboard performance, without a specific beat or drum. To find the tempo, I used the keyboard’s MIDI recording. Then, I created “bounce” copies, which muted the original tracks. Next, I transcribed the voice from audio to MIDI. This is quite a detailed process. After that, I soft quantized the keyboard at the 1/8 note level to smooth out the beat. I also created a chord track to assist with the final score editing. Finally, I rendered the MIDI files to audio for the final mixdown.      

What really stood out to me was how intricate the vocal part was. Converting the vocal audio into a MIDI file seemed like a fairly good fit for the performance. The voice flows smoothly from one note to the next, but the solo trombone doesn’t quite match that fluidity. 

It’s definitely simpler to tweak the final score when the original recording is set to a beat, like a metronome. However, this can stifle the creativity and natural pauses that happen during a live performance. Plus, live performances aren’t always perfectly timed to the beat!