chaosmoon

I'm working on a midi-clock syncable sequencer and up to now am working in the following way:

incoming ticks are counted with [counter 0 0 95]

each value is looked up in a table for gate value (note-on)

if 1 then lookup note value, swing(hence the 96 values) velocity, length etc.

The sequencer is only 16 steps so now my question is:

Would it be more efficient to have 16 [sel] objects with values defined by the swing offset, listening to the counter and define the rest if they are triggered?

So thats 1 table being polled 96 times a bar or 16 [sel] objects being polled 96 times a bar. After writing this i suspect my current setup might be better.

incoming ticks are counted with [counter 0 0 95]
each value is looked up in a table for gate value (note-on)
if 0 then nothing
if 1 then lookup note value, swing(hence the 96 values) velocity, length etc.
make-note

So thats 1 table being polled 96 times a bar or 16 [sel] objects being polled 96 times a bar. After writing this i suspect my current setup might be better.


efficiency-question-when-looking-up-values-table-vs-sel

you dont have to do 16 different select objects, just use 1:

you dont have to do 16 different select objects, just use 1:
select 0 1 2 3 4 5 6 7 8 9 etc...


the problem is that the values arent fixed, they change based on "swing" value. So its 0-5, 6-11 etc. I can change the value for a single [sel] object but not for multiple values in a single [sel] object AFAIK.

My gut instinct is that one table object will be more efficient than multiple sel objects for what you describe.

My gut instinct is also that it won't be that big a difference. But the only way to be sure is to try both and see.

In general, message passing does eat some CPU, so if there is a clear way to do something with one object I will prefer that to a solution that needs multiple objects. But if I need multiple objects to do something I don't lose any sleep over it.

efficiency question when looking up values [table] vs [sel]