Analogue-sequencer

> i've got a few questions & requests, mostly about accumulators..
> 
> - does the Reset on Pattern Switch on/off Acc.Config option affect 
> the pattern in which it is set only? for example, with the RoPS set 
> to off in pattern 2 but to on in patterns 1 & 3, would the acc. 
> continue accumulating on a switch from pattern 1 to pattern 2, and 
> also on a switch from pattern 2 to 3? 

When a different pattern is selected for a track, from a playlist step or
part change, the RoPS flag is checked, and if it is set, the accumulators
are reset.
In your example, with RoPS on for patterns 1 & 3, and off on 2, the
accumulators wouldn't reset from 1 to 2, but would from 2 to 3.

> then the requests (in order of importance to me..): 
> 
> - would it be possible to add an aux event to offset the 
> accumulators? so you could shift the "0"-point up or down, with the 
> range remaining the same. 

You mean to select an alternative 'zero' point to apply when the
accumulators are reset on a new pattern or by RTZ ?
That shouldn't be too hard, though there is no space to store the values per
pattern. It would have to be an event that temporarily set the zero-point.

> - would it be possible to have a separate accumulator for delay & 
> length, besides the ones already implemented for note & velocity? i 
> can reach this effect by using 2 aux tracks, but a separate acc. for 
> these would be a lot easier..

Each of the accumulators adds a processing overhead. If I add too many of
them, latency would increase. Only the initial latency for a clock pulse is
a problem as the rest of the processing time is masked by the midi
transmission latency, but the accumulators and aux events have the biggest
impact in the initial stage (the processing before the first midi message is
generated).
It may be possible to make the aux D accumulator more flexible - so it could
be assigned to delay or length.

> - would it be possible to add some sort of 'divider'-aux event for 
> the random and acc. events, so you could have a delay-, note- or 
> length-value randomly jump within steps defined by the divider. i.e., 
> set the randomize delay to max with a divider of 3 would give steps 
> of 3/6/9..i think it'd be useful.

You mean like a quantise event ? That could be done quite easily - you would
need to randomise the value with one aux, then 'quantise to the nearest
multiple of n' with another. Does that sound like it'd do what you want ?

> - could the 'offset [XX] rel' events have a (user-configurable?) 
> range of +/- 127 instead of the +/- 64 it is at the moment? i realise 
> you've now divided the 127 available steps evenly between the + and - 
> sides, but personally i could live with steps of 2 if i could get the 
> full range. preferably user-configurable of course..

That would need to be a different event. I'll have a think about it.

It could take a while for any of this to appear of course - we're on a 3
hour feeding cycle here, so sleep is a luxury, never mind spare time.
Hopefully we'll get into a routine soon.

Cheers,
Colin f