Analogue-sequencer

Folks,

Paul said...

> I was thinking that a way of making Playlist Edit more intuitive 
to a
> newcomer would be to have the first 8 keys represent playlist 
position
> and the second 8 act as selectors for patterns.

It occurs to me that a big part of the problem is the overlap of the 
steps used to indicate playlist position and current pattern, and 
the indirectness of choosing the patterns on steps other than the 
current one.

Modifying Paul's idea slightly, what if the first 8 keys are for 
selecting patterns, and keys 9 - 16 represent playlist steps.
The labelling for PART could become PART/PLAYLIST STEP to make this 
clear.
One LED above the playlist step keys would be green to indicate the 
current playlist position, the rest would be red to indicate the 
length of the playlist if it were longer than one step.
The playlist will always start on step 1 (key 9), and a bar of LEDs 
will always show the length. The PLED display can lose the length 
indication and be less cluttered.

The pattern on the current playlist step would be indicated by red 
on one of the LEDs 1 to 8, and pressing one of the step keys 1 to 8 
alone would select a new pattern for the current step.
But if you hold one of the playlist step keys (9 - 16) the LEDs 
above 1 - 8 will show the pattern on that playlist step, and the 
keys will allow it to be edited.
To alter the length, you would hold the current softkey and press 
the playlist step key for the last step.
Transpose and repeat control would move to the knobs above the new 
playlist step keys, but otherwise operate in the same way, as would 
chg and hold. No need for adv when stopped though - all playlist 
steps would be directly accessible.

This would leave the PAGE key free to somehow enable selection of 
patterns 9 to 16 in 3 bank mode. Holding PAGE plus a playlist step 
plus a pattern key seems a bit much, so maybe PAGE could toggle an 
indication in the display showing whether 'upper' or 'lower' 
patterns would be selected.

I don't think this version would be any more space consuming than 
the current implementation. When I get a chance, I'll implement it 
as a test build and see what people think.

Cheers,
Colin f