posted 27 November 2010 20:21            

I am getting over my irrational fear of the sequencer engines. I am editing the values, it is not as unfriendly as I initially suspected. In-fact I can see it opening up yet another world of possibility for me to explore.

As ever it has led me to some questions. I am changing the rate of sequencer by multiplying the BPM for triplets etc. Unless I change the events on the timeline, I can't get it to keep strictly intime with my arrangement. I would like to delay the message so it happens on the next scheduled moment of my choosing.

I have had a rummage yet I am obviously not searching the right terms.

Many thanks

posted 28 November 2010 13:52            

no mention of 'alignWith' contained in the KymaXrev, I did discover a few posts mentioning it, I kinda thought I had cracked it with:

(!BPM / 60.0) * (!RateMadness alignWith: (1 bpm:!BPM))

It sometimes plays the first beat, yet does not carry on. This sound is on a timeline.
I am pretty sure I must have the brackets in the wrong places.

I can see how you use this message to reset the gate on a sequencer using Cristians expression, however I don't see how to implement this for a VCS change!
:-)

Many thanks

C

posted 28 November 2010 16:11            

((!RateMadness hasChangedInLast: (!BPM bpm s)) alignWith: (1 bpm:!BPM))) sampleAndHold: (!BPM/60 * !RateMadness)

posted 18 December 2012 12:00            

Ta

posted 19 December 2012 11:19            

...
A new CapyTalk expression, alignWith:, for aligning one trigger with another. For example, say you wanted to press a Gate button but not have it take effect until the next beat. You could use the following expression in a Trigger or Gate parameter field:

posted 22 December 2012 15:40            

You've uncovered an anomaly here that I've been trying to get round for a long time.

Your expression

!Gate alignWith: (1 bpm: !BPM)

works fine if the thing that you are gating is has nothing to do with the BPM, but if you are gating a change of BPM itself, then the BPM you are alining with has to be changed in the same way.

It can't do this and you end up nesting alignWith: inside alignWith: add infinitum.

You need a form of Capytalk feed back so that the BPM resulting from the alignWith statement is the same BPM that the alignWith change is resulting from.

In other words the answer has to be calculated using the the answer as one of it's variables.

I came across this when I tried to make a midi recorder that would store notes and matching time stamps into ram. The problem was that to play it back the time had to be controlled by the time data but the reading of the time data had to be controlled by the time. It's the sort of thing that DSP programming could do with ease, but not so easy in Cappytalk.

Maybe it would work with a one tick (1ms) delay which was feeding a sound to global controller. That would allow the feed back to happen, and could allow a delayed nested version of itself, to sit inside the expression.

Maybe people have a neater way of doing this?

Charlie, I'm wondering if this anomaly is the problem you are having?

Thanks

posted 22 December 2012 21:20            

Perhaps you could use the CapyTalk expression triggerTimeToNext: to accomplish this? For example if you had:

!Gate triggerTimeToNext: <eventExpression>

it generates a trigger at the time interval(s) generated by <eventExpression>. !Gate turns the generation of triggers on or off.

For example:

code:

---

   1 timeToNext: !Duration s  
   1 triggerTimeToNext: !Random abs s
   1 triggerTimeToNext: (2 ms randExp s)

---

posted 23 December 2012 15:40            

I think this may help with the recored midi play back but I don't think it will work with a Bar defined BPM change.

If we have a slider with desired BPM . Nothing random in this scenario.

We want the BPM to change to the desired BPM but only at the beginning of the bar (every fourth beet). If we move the slider at any other time it must not have any effect on the BPM until the fourth beet has happened.

So we use a sample and hold on the BPM slider but the trigger of the sample and hold has to come from a count 4 of the pulse that is being generated by the BPM that is being controlled by the result of the sample and holds output.

Would the above expression help help in this case.

Thanks

posted 23 December 2012 16:41            

((1 bpm: !curBPM) triggerEvery: 4) sampleAndHold: !BPM

posted 26 December 2012 15:50            

That looks like it should do it, and it doesn't need the one tick delay as that is inherent in the sound to global controller.

Thanks