PDA

View Full Version : How to modulate tempo?


blurk
12-30-2008, 10:13 PM
Well, you did say no question is too silly...

I want to create a ramp up tempo change, i.e. start slow and build up to a faster tempo, then essentially maintain a constant tempo. I tried the naive approach (modulated the tempo parameter of the master clock), got into all sorts of weirdness, checked the documentation (in the Groove Clock section) and that says "You can also modulate the overall tempo by using parameter modulation on the tempo parameter of the master clock". So I am a bit flummoxed about the best way to do this.

The problem, of course, is that since I want a slow ramp, modulation sources like LFOs and the envelope are not appropriate (unless I'm missing something; always a possibility). Which leaves me the option of using the modulation sequencer. Unfortunately, the modulation sequencer (used in a naive way) is clocked by the master clock whose tempo it is modulating, so I end up with a completely uncontrolled rise in tempo. To be honest, I'm not entirely sure what it is doing, but given its behaviour it's clearly the wrong way to try and do this.

I've had some limited success stretching my basic knowledge of Numerology by adding another groove clock to control tempo for the bulk of the stack's step sequencers, with another modulation seq clocked to the master clock modulating that second groove clock's tempo. It kind of works, but since I want a ramp in, that modulation seq needs to be a one-shot and, as I've commented elsewhere, getting a reliable one-shot sequencer that doesn't kill the master transport is a bit beyond me at the moment.

So, has anyone else attempted this kind of tempo modulation? What's the best way to do this?

jim
01-01-2009, 11:54 PM
I'm working on a couple of examples. Since the envelope runs in 'real time' vs. beat time, and you just need a ramp, it is a good option -- however the details can be a bit tricky and I want to make sure I have a clean solution.

Another option I'm working on is a ModSeq with its own GrooveClock, so that tempo changes on the master clock don't affect it. The trick there is going to be to use a long step time in conjunction with a long sequence, so that you can have automation for the entire track.

Cheers,
Jim

blurk
01-02-2009, 04:58 AM
I'm working on a couple of examples. Since the envelope runs in 'real time' vs. beat time, and you just need a ramp, it is a good option -- however the details can be a bit tricky and I want to make sure I have a clean solution.
I think the reason I excluded the envelope was that I thought the time-scale was far too short.

Another option I'm working on is a ModSeq with its own GrooveClock, so that tempo changes on the master clock don't affect it. The trick there is going to be to use a long step time in conjunction with a long sequence, so that you can have automation for the entire track.

That sounds great! Though without seeing it yet, I wonder if I need to caution about the possibility of very long tracks. I've been running a variation of Jue's chaotic rhythm generator example as an ambient background for me all day. N2 is currently on bar #13924 as I type!

But seriously, anything that helps with entire-track length time scales will help. From the forum I gather there's a few of us for various reasons that are attempting to program up entire tracks without using the timeline. For me, it comes from working on things with a more programmatic background, and the timeline just takes me back into the strictly linear DAW paradigm that I'm kind of trying to escape. You'll have to ask the others for their reasons...

jasonwolf
01-02-2009, 11:09 AM
I think the reason I excluded the envelope was that I thought the time-scale was far too short.

When I initially read blurk's inquiry about modulated tempo, I at first thought about an envelope generator with a suitable attack and infinite sustain (to maintain bpm)... but that would take a lot of trial and error to get the ramp slope just right. (i.e to ramp from 85bpm to 135bpm)... the more I think about this, the more intrigued I am about this and looking forward to Jim's examples...

For me, it comes from working on things with a more programmatic background, and the timeline just takes me back into the strictly linear DAW paradigm that I'm kind of trying to escape. You'll have to ask the others for their reasons...

This is also very intriguing to me. I wrote script for MaxMSP that could control a collection of sequences and samples either randomly or controlled. In other words, I had a collection of about 2 dozen samples and small sequences that would randomly play... Blurk's wanted departure of a master timeline reminded me of this...

It would be really cool to be able to specify in N2, what stack to play when (at 2:20, stack 7 plays for 12 bars, etc.) or have an element of randomness that selects specific stacks to play at different times... and of course having a tempo demultiplexer would be fantastic (think that your master clock is 120 bpm, and the demultiplexer would control tempo derivatives for specific stacks... playing them at 1/3 the tempo, or 1/2 the tempo) mmmmm yummy!