Generic vs. Song Style
When creating styles you have to decide if the style is a
• generic style (e.g. a ballad style)
or a
• song style (e.g. a style for the song "Yesterday").

The generic style is intended for use in many songs, therefore the accompaniment must be flexible. This is most often done by increasing the accompaniment complexity AND number of instruments used from the "Main A" part to the "Main D". And for the Intro, the Fill and the Ending parts as well.

Example: The Main A part may have only a drum, a bass and a chord channel, all playing rather simple patterns. Main B may add another chord channel. Main C may add a pad channel. And Main D may add a phrase channel.

If we consider a song style we will have to create style parts for what traditionally are called the "A" and the "B" part. And for "the bridge" as well. This means that complexity and number of instruments used are dictated by the original song.

In song styles the Intro and the Ending parts will have to sound like the original intro and ending. Quite often these parts include both the chord progression (which in other parts is played with the left hand) and the melody line (which in other parts is played with the right hand).

If the style is created with these "melodic" Intro's and Ending's, all the performer has to do is to hit the song's root chord. Then the style will play a melodic intro, and when this is finished the performer takes over by playing one of the Main parts.

Creating the channel parts
We are now ready to add notes to the channels. There are several ways to do this:
• Record the notes using your keyboard
• Pick the notes you need from an existing style
• Program the notes in the sequencer software
or combinations of these methods.

When recording from the keyboard all the small variations (and the errors) are recorded. The small variations (e.g. timing, note loudness, note length etc.) will give the style a more natural sound - keep them. But of course the errors (too large variation) have to be corrected. The best approach is to correct the erratic notes individually instead of quantisizing, which will remove the natural sound.

When creating the style in sequencer software, the performance often becomes too perfect. The best approach is to add some (random) variation to timing, note loudness, note length etc. Quite often the sequencer software has a "Humanize" feature for this purpose.

A pitfall to be avoided is to copy and paste too much. There is no idea in creating a 4 measure long part if all the four measures are identical. And - as stated before - it is a good idea to increase the pattern complexity in generic styles going from the A variation to the D variation.

We will discuss the above topic further in the coming parts.

All style parts should be recorded using notes in CMaj7. You are able to change chords within the parts, but the overall performance has to be in this key. The following rules are recommended:
• In Bass and the two Phrase channels: Use only C, D, E, G, A and B notes
• In Pad and the two Chord channels: Use only C, E, G and B notes

Intro and Ending Parts
Intro's and Endings can be created in three different methods:
• Without chord progression and melody.
In this situation the performer will have to play the chord progression with his left hand and a melodic phrase in his right hand.
• With chord progression, but without melody line.
In this situation the performer will have to start the accompaniment by hitting the root chord with his left hand and then play a melodic phrase in his right hand.
• With chord progression and melody line.
In this situation the performer will only have to start the accompaniment by hitting the root chord with his left hand. The Intro and ending parts are in reality complete MIDI files. This method is often used in Song Styles.