top of page
marula web logo222.webp
  • Instagram
  • Facebook
  • Threads
  • Youtube
  • Patreon
  • Bandcamp
  • Beatport
  • Spotify

Bitwig 6 Automation Layering: A Different Way to Build Movement


Most automation workflows follow the same logic: you draw a curve on the timeline, it does its thing once, and you either loop the whole section or draw another curve elsewhere. It's fine. It works. But it also means your automation is essentially just a shape that plays back at a fixed point in time, which puts a ceiling on how alive a sound can feel over a long section.

Bitwig 6 changes that, and honestly, the automation system is the thing I keep coming back to out of everything in this update. Not because it's flashy, but because it flips the whole mental model. Automation is now clip-based, which means you can treat it the same way you treat note clips, loop them independently, give them different lengths, stack them, swap them out. Once that clicks, a lot of ideas open up.

I want to walk through a specific technique I've been using: layering multiple automation sources on a single parameter, each running at its own loop length, blended together with a Mix device. The results are a lot more interesting than anything you'd get from an LFO or a single drawn curve.

Why Clip-Based Automation Is a Different Paradigm

Before getting into the technique, it's worth being clear about what actually changed. In previous versions of Bitwig, automation lived on the timeline, like it does in most DAWs. It was attached to time positions, not to clips. You could draw it, you could loop the arranger playback, but the automation itself wasn't a portable, independently looping object.

In Bitwig 6, automation is stored inside clips. That means a clip can contain both note data and automation, and the automation will loop whenever the clip loops, regardless of the timeline. You can give that clip a different length to your note clip and now you've got polymetric automation, something that's previously required LFOs or fairly involved modulation routing to achieve.

The LFO comparison is worth addressing directly. You can absolutely use an LFO to get a looping modulation pattern, and I've done a lot of that. But an LFO is always going to be a shape, a sine, a triangle, a custom drawn wave, repeating at a set rate. With clip-based automation, you can draw stepped patterns, you can have a section that holds a value for three beats and then drops suddenly, you can mix smooth curves with sharp edges, and you can change the pattern mid-arrangement just by swapping the clip. It's more like programming automation in a step sequencer than dialling in a modulation rate.

Setting Up the Bitwig 6 Automation Layering Technique

I'll walk through this using Bitwig's Polymer device as the target, but the approach works with anything. Start with a loop running, something with filter movement potential. Set up a polymeter on the clip if you want, something that isn't locking to the bar.

First pass: draw in some basic cutoff automation directly in the clip. Smooth filter sweep over eight bars or so. This gives you a foundation, the overall movement of the sound. This part's not unusual, just standard clip automation.

Now, here's where it gets interesting.

Introducing Macros as Modulation Layers

Rather than drawing more automation directly to the cutoff, add a Macro device. Name the first macro something like "Cutoff L1" (Layer 1) and assign it to the filter cutoff. On the surface that looks redundant, you could just draw more automation onto the cutoff directly. But the macro gives you something the direct approach doesn't: you can set it to bipolar.

A bipolar macro can push the cutoff up or down from whatever position it's currently at. This becomes useful because the audible effect of positive and negative values changes depending on where your base cutoff is sitting. If the filter is mostly closed, negative values won't do much, but positive values will open it up noticeably. If the filter is mostly open, it flips. That means the same looping automation pattern will behave differently at different points in your arrangement, just because the base position changes. You get variation for free.

Now create a new automation clip for this macro and draw something in, a short pattern, maybe just a bar or two, set it to loop. This is your first automation layer.

Adding a Second Layer

Add a second macro, "Cutoff L2". Before assigning it anywhere, bring in a Mix device. Assign Cutoff L1 and Cutoff L2 both to the Mix device (set both to 100%, both bipolar), then assign the Mix output to the filter cutoff. The crossfader on the Mix device is now your blend control between the two automation layers.

With the crossfader at the far left, you're hearing only L1. At the far right, only L2. Anywhere in between, a blend of both.

For L2, draw something structurally different from L1. I went for a stepped pattern rather than a smooth curve, using the spray can to create something more rhythmic and abrupt. Set it to a different loop length than L1, say a bar and a third instead of a bar, and now the two patterns are drifting in and out of phase with each other over time.

This is the polymetric part. Neither layer is synced to the other. They interact differently on every cycle. The sound develops without you having to draw anything new.

Automating the Crossfader

Here's the third layer: automate the crossfader itself. Now you have a parameter controlling how much of L1 versus L2 you're hearing, and that balance can shift over time or per section. At some points you're fully in the stepped pattern, at others you're in the smooth curve, and in between you get hybrids that neither layer could produce on its own.

The combined effect, two looping automation layers out of phase with each other, blended by a third automated parameter, creates movement that genuinely doesn't repeat in any predictable way. It's not random, it's structured, but the structure is complex enough that it doesn't feel mechanical.

Using Macros to Control Amount, Not Absolute Values

One other thing worth covering: the difference between automating a parameter directly versus automating a macro assigned to that parameter.

If you automate, say, envelope decay time directly, you're setting absolute values. The automation sets the decay to exactly 200ms at this point, exactly 400ms at that point, regardless of where you've manually set it. That makes it hard to tweak the sound without breaking the automation.

If you route through a macro instead, the macro controls how much the automation can push the decay, while the base decay time stays as you set it on the device. So you can have automation that's modulating the decay "amount", dial in how aggressively that modulation applies, and still adjust the underlying decay time as a separate creative decision. Macros as amount controllers rather than absolute position setters is something I use a lot in general, not just with Bitwig 6, but the new automation system makes it a lot more practical to set up.

Live Control Still Works

One more thing, because it comes up. When you've got layered automation running like this, you can still grab a parameter and move it with a MIDI controller or with your mouse. The automation continues to run underneath. You're effectively offsetting from whatever the automation is doing at that moment. So this system doesn't lock you out of hands-on control during performance or recording; it just adds a moving baseline underneath it.

What to Try Next

If you want to experiment with this: start with one parameter, the filter cutoff is obvious but resonance, verb mix, or oscillator detune all work well. Set up two macros, draw two different shaped patterns at two different loop lengths, and blend them with a Mix device. Once that's running, listen to how the sound changes over a few minutes, it'll surprise you.

The full walkthrough is in the video on the Marula Music YouTube channel if you want to see the exact routing. And if you're working with Bitwig in general, the Parseq parameter lock sequencer article is another approach to polymetric modulation worth pairing with this.


Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page