ddw_music

I wanted to manipulate MIDI note data in such a way that I could modify the time between note onsets and note durations independently.

I could be missing something, but it seems [coll] might be the best choice to hold the note data -- use one or two [timer]s to get the note durations and inter-note deltas, then pack those up with the note number, velocity and an integer index from a [counter].

The issue with [coll] is that I can't find a simple way to stop reading out the note data at the end of the sequence. (next) or (bang) --> [coll] wraps around at the end. Currently I'm using [t i i] to compare the current index against the previous index -- if the current index is greater, keep going; otherwise, it's looped back around to the beginning and I can send a signal to the [metro] to stop banging.

Is there an easier way? I rechecked the Tutorials, but they only discuss [seq] and this doesn't expose the timing information as measurements to mess around with algorithmically.

NB about "simple," I'm comparing against SuperCollider, where, given note data in the format [notenum, on_velocity, delta, sustain], I can simply do this and it stops at the end with no fuss.

    #[midinote, velocity, delta, sustain], Pseq(notes, 1),

    amp, Pkey(velocity).linlin(0, 127, 0.3, 1.0)

By contrast, I find it highly amusing (love the Cycling 74 developers' sense of humor) that [coll] can output a bang at the end of a (dump), but not when stepping through.

Pbind(
	instrument, defname,
	#[midinote, velocity, delta, sustain], Pseq(notes, 1),
	amp, Pkey(velocity).linlin(0, 127, 0.3, 1.0)
).play;

detecting-the-end-of-a-coll-when-stepping-through

hi in a similar situation i created a second instance of that same coll and queried its length with the "length" message and then set a counter to control the number of bangs before stepping thru the data.

i don´t have the patch around at the moment, but maybe you get the idea.

hi in a similar situation i created a second instance of that same coll and queried its length with the "length" message and then set a counter to control the number of bangs before stepping thru the data.
i don´t have the patch around at the moment, but maybe you get the idea.

Can you put an "end" entry into your coll... at index 99999 or somthing suitably high?

Assuming a coll with ordered indices 1..N, you can step through it with a counter, banging the counter after receiving each output. If the counter becomes greater N, coll doesn't find an entry and produces no output. Thus the loop will stop "automatically".

Some nice suggestions, thanks. I'm not sure they are necessarily simpler than what I have now, but reasonable.

One issue with broc's suggestion is that I would want to use a [toggle] to start and stop playback, and also show the playback status. So, at the end of playback, I send a (0) message back into the toggle -- and this depends on some message being generated somewhere when we hit the end. The absence of output wouldn't help here.

Now to try to explain all this to max/msp noobs *lol*

I also used a 'length' message with a loadbang in a similar situation.

here is one way to trigger on absence of a bang

Yes, I understand that my approach is not ideal in every context. But actually I'm using it in a typical "looper" configuration, ie. playing a coll of notes within a loop. In this case start of the coll player is triggered by the outer loop and no explicit end message required.

Detecting the end of a [coll] when stepping through