Title: Call-and-response filter motion: without third-party plugins (Advanced)

Alright, let’s build one of those “the bass is talking” moments in drum and bass, but we’re doing it the clean way: stock Ableton only, and with automation that feels like it belongs to the groove instead of fighting it.

The goal is call-and-response, not in melody, but in movement. Two filter “voices” that alternate in a predictable DnB phrase, and each voice has its own personality. Tight, resonant, speech-like motion for the call… and a wider, more open answer for the response.

Before we touch any devices, set the context. Tempo: 172 to 175 BPM. I’ll imagine 174. Put down a simple two-step: kick on one, snare on two and four, hats doing steady sixteenths or a slightly swung sixteenth pattern if you want. The reason we do this first is simple: filter motion lands way harder when the drums are locked and predictable. The drums are your grid. The filter is your actor.

Now choose a sound source. Bass is the clearest example, because the tonal shift is obvious and the movement can feel like speech. If you want the quickest stock setup, use Operator. Oscillator A on saw or square, keep it mono and stable. In Operator’s filter section, set a low-pass 24, somewhere around 700 to 1.5k as a starting point, with moderate resonance. Then add a Saturator after Operator, Analog Clip mode, drive maybe 3 to 8 dB, Soft Clip on. That gives the filter something to bite into.

If you want a heavier mid character, Wavetable is also stock and great. Basic shapes, slight detune if you want, but keep an eye on mono if you go wide. For DnB, I usually like keeping the very bottom solid and letting the mids do the talking, and we’ll address that later.

Now we build the actual call-and-response system. Drop an Audio Effect Rack after your synth and saturation. Inside that rack, create two chains. Name one CALL and the other RESPONSE. Think of them as two different characters. Same notes, same rhythm… different mouth shapes.

Put an Auto Filter on each chain. We’re going to automate manually, because we want phrasing-level intention, not a random LFO doing its own thing.

For the CALL chain, set Auto Filter to band-pass, 24 dB slope. Put the frequency somewhere like 550 Hz to start, and make it resonant. Not subtle. Think 50 to 75 percent resonance. Add a little drive, maybe 2 to 6 dB. And turn the LFO off.

For the RESPONSE chain, set Auto Filter to low-pass, but make it more open: 12 dB slope works well. Frequency maybe around 2 kHz to start, resonance lower, like 15 to 35 percent, and less drive, around 0 to 3 dB. This chain should feel like release, like the bass opening its jaw instead of talking through its teeth.

That contrast is the whole trick. CALL is mid-focused and resonant, RESPONSE is broader and more airy. You’re not just changing cutoff, you’re changing character.

Next, we need a way to alternate between them like a conversation. The cleanest method for arrangement control is Chain Selector automation.

Open the chain list in the rack, find Chain Selector, and set the chain zones so CALL is active at value 0 and RESPONSE is active at value 127. Then map Chain Selector to Macro 1 and name that macro “Call/Response.”

Now you can automate that macro in Arrangement. Start simple: bar 1 is CALL at 0, bar 2 is RESPONSE at 127, bar 3 back to CALL, bar 4 back to RESPONSE. That’s your basic four-bar statement.

And for that classic DnB “little answer before the snare,” try switching every half-bar in a fill. For example: beat 1 through 3 is CALL, and then beat 3 to 4 is RESPONSE, so you get that small reply right before the snare. It feels like punctuation.

One important note here: use instant jumps, not ramps, when you’re switching the chain selector. Ramps create morphing, and that can be cool, but it’s a different vibe. We’ll talk about smooth crossfades later as an advanced variation.

Now, the main thing people mess up: they build the two chains, they switch them, and then they don’t actually write motion inside each voice. So let’s write motion.

On the CALL chain, automate Auto Filter frequency with short, rhythmic shapes. Think in eighth notes and sixteenth notes. Quick ramps, quick dips, and little jabs right before snares. Here’s a mental one-bar example you can copy as a starting template: at beat one, sweep up fast from 600 to 900 Hz. At beat two, drop from 900 down to 450. At beat three, do a small wiggle, like 500 to 700 and back. And right before the snare on beat four, do a hard dip down toward 350, then reset right after. That dip is your “question mark.”

When you draw this, use a little finesse. In Live, if you hold Alt or Option while dragging automation points, you get finer control. That matters because the difference between “talking” and “tone vanished” can be a tiny move.

On the RESPONSE chain, do the opposite kind of motion: longer arcs. Half-bar sweeps, one-bar curves. For example, beat one to three: open from 1.8k to 3.5k. Then beat three to four: close back to 1.5k going into the snare. It should feel like breathing behind the drums rather than chattering on top of them.

And here’s a teacher move: don’t let frequency do all the work, or your automation starts feeling one-dimensional. Give each voice a second parameter so the characters really separate. For the CALL, keep frequency moves smaller and add tiny resonance nudges. For the RESPONSE, do broader frequency arcs and add small drive nudges. Same synth, same notes, but now the “mouth” changes in multiple ways.

Now let’s go advanced: make the filter react to the groove, but without losing your phrase control. That’s where Envelope Follower comes in.

Put an Envelope Follower on the bass track, and set its sidechain input to your drum group, or better: just hats or ghost percussion. Hats driving movement often feels more “rolling” and less jumpy than a kick. Start with attack around 5 to 20 milliseconds, release around 80 to 180 milliseconds, moderate smoothing.

Then map the Envelope Follower to the Auto Filter frequency in the CALL chain with a tiny range. Tiny. Think 5 to 15 percent of a useful window. This is not meant to rewrite your automation, it’s meant to add micro-timing life on top of it. Manual automation is your macro phrasing. Envelope follower is your micro groove.

You can even split the duties: hats drive the CALL for chatter, and kick drives the RESPONSE for a bigger push. Just keep the mapping range conservative so it feels intentional, not chaotic.

Now, let’s lock this into an arrangement so it actually sounds like a drop.

Bars one through four: establish the pattern. Bar one CALL, bar two RESPONSE, bar three CALL with a slight variation in the automation, bar four RESPONSE slightly more open, and maybe a tiny rise into bar five.

Bars five through eight: one rule change. Switch every half-bar for a couple bars. Then bar seven, lean mostly CALL to create tension. Bar eight, open the RESPONSE, and in the last quarter of bar eight, do a hard low-pass choke so it resets the loop with impact.

That’s a DJ-friendly structure: simple first, variation second, and a clear loop point.

Now we need to keep it mix-ready, because DnB basses get messy fast when you automate everything.

If this is your main bass, split sub and mids. Create another Audio Effect Rack for the whole bass chain, make two chains: SUB and MID. On the SUB chain, put EQ Eight and low-pass it somewhere like 80 to 120 Hz. Keep the sub steady. Minimal automation down there. On the MID chain is where your call-and-response rack lives, and that’s where aggressive movement belongs. This is how you get translation on club systems without the bottom end wobbling in volume.

Also, resonance can spike levels. Auto Filter resonance is basically a boost when you hit that sweet spot. So put a safety device right after your call-and-response rack: a Limiter with the ceiling at about minus 0.8 dB and minimal gain, or a Glue Compressor doing only one to two dB of gain reduction. Not for loudness. For control. That way you can push resonance without random peaks ruining your headroom.

Another practical issue: sometimes chain switching clicks, especially with resonant band-pass. If you hear ticks, you’ve got two stock fixes.

One fix is to put a Utility after the rack and automate a tiny fade down and up around switch points. Five to twenty milliseconds is enough. You won’t perceive it as a volume move, you’ll perceive it as “smooth switching.”

The other fix is to avoid chain selector jumps entirely and do a Utility duel. Put a Utility at the end of each chain, and automate the gains in opposite directions: CALL at 0 dB while RESPONSE is down at negative infinity, then swap. And instead of hard mutes, draw micro-fades. This method also keeps reverb or delay tails more consistent if you’re using sends, because the audio doesn’t hard cut in a single sample.

Now a few pro sound design upgrades, still stock.

If you want darker, heavier vocal snarls, put a Saturator inside the CALL chain, not just globally. Push it harder than the RESPONSE chain, and maybe tame harshness with EQ Eight around 2.5k to 4.5k if it gets barky.

If you want formant-style talking without any special plugin, add EQ Eight after the Auto Filter in the CALL chain. Make two narrow boosts, Q around 3 to 6, one around 500 to 800 Hz and another around 1.2k to 2.2k. Then automate those boost frequencies slightly in opposite directions. That’s the stock “vowel” trick. It’s subtle, but once you hear it, you’ll start using it constantly.

If you want rhythmic syllables, put Auto Pan before the filter with phase at 0 degrees so it acts like tremolo. Set it to one-eighth or one-sixteenth, amount 10 to 30 percent. Now the filter is shaping a pulsing input, which reads like articulation.

And if you want a little controlled chaos, Beat Repeat can work, but keep it dry-only and subtle: place it before the rack, interval one bar, grid one-sixteenth or one-thirty-second, chance 10 to 25 percent. The repeats become consonants that your filter motion “speaks.”

Last advanced timing trick: you can add swing to automation without touching grooves at all. In Arrangement, nudge a few automation breakpoints late by five to fifteen milliseconds on offbeats, especially pre-snare moves. You’ll get that lazy pull, but the MIDI stays locked.

Now let’s wrap with a practice challenge you can actually finish in one session.

Make an eight-bar MIDI clip with a simple bass pattern, even just one or two notes. Build the two-chain call-and-response rack with Auto Filter on each chain. Automate the chain switching: bars one to four, switch every bar; bars five to eight, switch every half-bar. Duplicate it to make sixteen bars, and invert the logic in bars nine to sixteen: wherever it was CALL, make it RESPONSE, and vice versa. Then resample your best four to eight bars, and do two edits: reverse a single RESPONSE hit, and add a one-eighth-bar low-pass choke at the end of bar sixteen.

If it still grooves with the drums after the inversion, you didn’t just draw automation. You built an actual conversation.

Quick recap so it sticks. You created two filter voices using stock Auto Filter inside an Audio Effect Rack. You alternated them using chain selector automation or Utility gating. You wrote distinct motion inside each voice, combining short “talking” shapes for the call and longer “breathing” arcs for the response. You optionally added Envelope Follower for groove-locked micro movement. And you kept it professional by stabilizing the sub, controlling resonance peaks, and smoothing switches when needed.

If you tell me what your bass source is—Operator or Wavetable—and whether you’re going for Reese, neuro mid, foghorn, or jungle stab, I can give you tight, safe macro ranges: specific min and max frequencies, resonance targets, and drive levels that stay heavy without blowing up your mix.