Back to blog
August 19, 2021

Starting with a blank canvas in Ableton




The third installment of Start to Finish: Ill Factor - Episode 3 - Introducing Jared Evan finds ill Factor and Jared working on ideas for the melody of the song they’re writing. In the excerpt, ill Factor shows a cool trick for experimenting with different melodic ideas.

Up and Down in Ableton

At this point in the production, Jared has recorded a number of melodic possibilities for the lead vocal and autotune. Ill Factor wants to try out some ideas he has for tweaking some of them, and he’ll be using Waves Tune Real-Time’s MIDI input feature to do so. He points out that you could also use other pitch-correction plug-ins that recognize incoming MIDI—for example, Antares Autotune Pro.

He’s working in Ableton Live, but this technique should work in any DAW, although the setup may differ, depending on the program. The idea is to route MIDI input into the pitch-correction plug-in, which you’ve inserted on the vocal track to enhance your autotune.

In Ableton Live, ill Factor creates a MIDI track and sets its output to the track called Ruff Vox, and once he does, the inputs for Waves Tune Real-Time show up right underneath, and he selects the Channel 1 input.

Ill Factor sends the MIDI (routing highlighted) into an instance of Waves Tune Real-Time that’s inserted on the vocal track.


He inserts Waves Tune Real-Time on the vocal track and clicks on the Target Pitch button, which sets it to receive incoming MIDI notes. Then, as he plays back the track with Jason’s vocals, he changes the melody by pressing keys on his MIDI controller. Waves Tune instantly transposes whatever note is coming into it to match the note he’s playing.

Using this technique in Ableton, he can show Jason his melodic ideas for the song using the singer’s voice. You’ll notice that ill Factor turned on the formant correction on Waves Tune Real-Time. That will help keep Jason’s voice from sounding too “sampler like”—that is, changing timbre when changing the pitch— when Waves Tune transposes notes. The formant correction can’t wholly hide timbre changes, particularly on large transpositions. Still, because this is only a writing session, and Jason will eventually sing a complete lead vocal, it’s not a critical issue.

Like Waves Tune Real-Time, Auto-Tune Pro offers a MIDI input mode (Target Notes).


Mellow Dining in Ableton Live

If ill Factor had wanted to, he could have recorded the MIDI data created when he was playing his controller to change the vocal. He then could have edited it after the fact to experiment with even more melodic options.

If you use Melodyne as your tuning plug-in or the built-in tuning features in DAWs like Logic Pro X, Digital Performer and Cubase, you can edit the notes, and thus change the vocal parts, after the fact. In Melodyne, you’ll need first to transfer the vocals into your DAW in real-time, and then you can edit them.

Many DAWs, such as Logic Pro X, have built-in pitch correction, which you can use to move pitches around to experiment with melodic changes.


The disadvantage of doing it this way is that it’s not as immediate. You won’t get the instant gratification of changing notes of the vocal from your MIDI keyboard during playback. However, if you’re not a good keyboard player, using Melodyne or another “non real-time” pitch corrector means that the software will take care of inputting the notes of the vocal part, and you won’t have to.

After transferring audio into Melodyne, you can move pitches around at will.


The Mix in Ableton

We’ve seen how a tuning plug-in with MIDI input can help with songwriting and arranging, but it can also be useful during the production process. You can use tuning plug-ins or built-in DAW tuning features to alter melodies or create harmonies of existing vocal or even instrument tracks.

However, there is a major caveat: With formant correction turned on, the larger the transposition, the more the vocal timbre will change. If you transpose up too much, you’ll get the “Mickey Mouse”effect. If you transpose down, it can create the “monster” effect.

If you listen to tracks transposed by pitch-correction software, they’ll frequently sound unnatural. But blended into the mix, they can work well, particularly in an electronic genre. That said, if you’re going for an organic vibe, it’s much better to have a real singer (or singers) do the harmonies.

A tip for making the MIDI track reproduce the vocals more realistically is to make sure that the notes in a phrase have little or no space between them. That will smooth out the transitions from note to note and make the final result sound more legato and authentic.

Keeping the note spaces inside phrases small or non-existent can help smooth out a vocal part created by MIDI through a tuning plug-in.


Here’s an example of harmonies generated by Autotune Pro in Target Notes (MIDI input) mode. You’ll hear a four-bar vocal line repeat twice. The first time there are no harmonies. The second time, you’ll hear a doubled harmony part that starts a third above the melody.

Here’s how the harmonies were created:


  • • An empty mono audio track was created for each harmony (well call them “Harmony 1” and “Harmony 2”) and Autotune Pro was inserted on each one.
    • A MIDI track (which we’ll call “MIDI 1”) was recorded to match the lead vocal.

    • The MIDI track was edited so that it would create a harmony when sent through Autotune Pro.

    • MIDI 1 was then routed to the MIDI input of Autotune Pro (the same way Ill Factor did it) on the Harmony 1 track.

    • MIDI 1 was duplicated to create MIDI 2.

    • The notes in MIDI 2 were very slightly randomized—by a few ticks here or there—so that the parts created wouldn’t be identical.

    • MIDI 2’s output was routed into an instance of Autotune Pro on the Harmony 2 track.

    • Harmony 1 and Harmony 2 were recorded as audio using Live’s Arrangement Record feature.

Thanks to Autotune’s ability to transpose incoming MIDI, it was relatively easy to create the harmony vocal tracks. Here’s what is sounded like. The first time through is just the lead vocal and the instruments. When it repeats, you’ll hear the doubled harmony vocals, panned about 2 o’clock and 10 o’clock.