lol *SIX* sections. ...
ok here's my take on it:
you have Pattern mode, and Song mode.
64 user patterns, called Styles, each of which has 6 sections,
plus a factory library of 128 pre-programmed 'styles'.
and 20 user-programmable Songs,
which are 16 tracks/parts, plus an accompaniment 'track'/part, where you chain patterns/styles.
you can, if you want, make a 'Song', and only use the pattern chain part, inserting 'sections' of
patterns, and the chord transposition feature.
but you also have linear Song tracks, that play alongside the pattern part.
a pattern/style has 8 tracks, loops in playback, and the sounds and mix are assigned via the mixer.
these tracks can be assigned to parts 1-8 or 9-16 (mapped to midi channels 1-16) and correspond
to parts:tracks/midi channels 1-16 in the Song mode, where sounds and mix are also assigned via
a mixer page.
and this is the slightly problematic bit - though it needn't be: pattern and song share tracks. the stuff
playing on the 1-8/9-16-assigned Pattern tracks are the same mixer tracks assigned in the Song mode.
so say you have a pattern section, just looping all the way through in song mode, you could just use the
corresponding song tracks to record midiCC changes. or overdub etc.
and then use the other 8 Song tracks for whatever you want.
or you can 'unpack' a (pattern) onto song tracks, with 'Get/Put Phrase' Job functions. eg: this would allow you
to introduce variations - also takes up more memory, since it isn't just repeating the data from the pattern.
you can also do this the other way round, jamming something into a Song track/part, and then grabbing a bit of
it, and copying it over into a Pattern track. this can be a more natural way of getting your performance take.
note that: every pattern/Style track recorded becomes a 'Phrase', that can be called up in any pattern 'Section',
on any 'Track' - which you can also do with the factory Styles. eg: so if you can't be bothered to record a drum
part, you can re-use bits of a factory Style in your new pattern/Style. (hence the terms Get/Put 'Phrase')
generally, starting with Pattern mode is better for jamming ideas and getting the parts of your song together.
alternatively, you could ignore Pattern mode, and play your stuff directly into Song mode tracks. or combine
the two. at simplest level, pattern sections looping on their 'pattern chain' part, with song tracks providing the
linear parts you want to add over the patterns.
one thing worth noting is that a Style is its own microcosm with its own mix/effects settings; when you go into
Song mode, the Song is 'boss', and its mixer/effect settings override the Style (at least i think? ...can you change
Voice assignments here, for the pattern playing? don't think so ...)
for example, i believe this because if you do something obscure like setting a channel midi out to 'off', you have to
input a sysex message in the Song mode, and call that up before getting your Pattern mode to play in the same way,
eg: excluding the midi output for that part. ( switching midi out per track is quite painful; wish they'd included it as a
standard function, like on the RM1x )
you can probably confirm all this using an external 16 part sound module, ignoring the QY70/100 audio output.
Eric refers to this as a 24 'midi channel' setup - i think it's 24 'tracks' (8+16), applied to 16 midi channels. but you'll
always have one side (1-8/9-16) of the Song setup sharing the voice channels with the Pattern mode.
i've never used Phrases much, but i think they could be really useful if you work a method like that - eventually, maybe:
using pattern first, and then expanding into a song is probably the easiest/fastest way to go.
