Skip to main content

My steps for creating choir learning tracks

· 4 min read
Jeroen Janssen
Jeroen Janssen

Introduction​

I have been creating choir learning tracks since 2010 when I joined a mixed choir. Since I was a new singer I had a lot of "catching up" to do to learn the existing repertoire and this was when I started to combine my passion for music and computers. In 15 years technology has improved a lot to speedup and improve the process.

Here are the steps I take and the software I am using.

Steps​

Scan the sheetmusic​

I am using PlayScore 2: a sheetmusic scanning app

It is no fun to manually type in all the notes, so I am using PlayScore 2 to help speedup the process. Most of the time I need to correct only some small things, but it really is an improvement from "long ago".

I export musicxml from PlayScore 2 for the next step.

Correct the notation​

I am using Dorico: a music notation program

I use Dorico to perform any corrections after the sheetmusic is scanned. I have used other programs in the past, but Dorico's insert mode is really powerful to automatically shuffle music left and right as you insert or delete notes from a voice, or change the durations of notes.

During this process I also regularly listen & read the sheetmusic to detect issues, e.g. missing note ties between pages, missing accidentals, etc.

  • I am using Pianoteq, a physically modeling based virtual piano.

    This virtual instrument can be configured in great detail and sounds amazing.

Split the voice parts​

I also use Dorico to split the voice parts (e.g. when you have a reduced choir score with split female/male voices). Since I need to be able to playback each vocal part separately I use the "Paste Special, Explode into Staves" functionality.

I export midi from Dorico for the next step.

Generate part mixes​

I am using Logic Pro: a DAW

While you can export audio from Dorico, I have choosen to do this from Logic because I am using my own mixing template that allows me to easily export different learning track mixes in one go.

This mixing template is something that I want to explain in a future post, but in short it works this way:

  • open mixing template (containing input midi tracks for S1,S2,A1,A2,T1,T2,B1,B2, piano guide track)
  • import the midi as exported from Dorico
  • record enable the different mix tracks (per voice)
    • part track: single part is predominant, the other parts have a softer volume
    • part stereo track: same as part track, but the predominant part is mixed in stereo to one side and the other parts are on the other side
    • part only track: single part, other parts are not present
    • part with guide track: single part but including a piano (accompaniment)
  • play the whole song through once
  • bulk export the different mix tracks (to wav)

Convert to MP3​

I am using Audacity: a free audio editor

This is the last step and is mainly needed because I cannot bulk export to MP3 from Logic. I use the default "MP3 Conversion" Macro from Audacity to convert all the mixes (could be up to 32 wav files) in one go.

After this I shortly check all audio files and then make them available to the choir members.

(optional step) vocal tracks​

Make AI vocal tracks​

I am using Synthesizer V Studio: a vocal synthesizer

At the end of 2023 I started experimenting with vocal synthesis to improve the learning tracks with actual vocals. This is really promising, but still a bit time consuming to do from a lyric and finetuning point of view (which is also a great topic for a next post)

Here is an example of a SynthV Choir cover: