Welcome back. Today we’re dialing in dub delay timing at 170 BPM, specifically for jungle rollers in Ableton Live. This is an intermediate session, so I’m going to assume you already know how to set up sends and returns, and you’ve got a break-driven loop running.

Here’s the big idea: in jungle and drum and bass, dub delay isn’t just “an echo effect.” It’s part of the groove. It’s propulsion. It’s the thing that fills the tiny gaps between snare, hats, and ghost notes, without ever stealing the punch. And at 170 BPM, the difference between “rolling and alive” and “smeared and messy” is usually just timing, filtering, and ducking.

So let’s build a return that’s designed for this tempo, then we’ll talk about how to throw it like a DJ, not wash the whole mix in echoes.

First, quick session context so the timing makes sense. Set your project to 170 BPM. Ideally you’ve got a break track, maybe a clean snare or clap layer on top, a bass or reese, and a stab or vocal chop you can use for throws. Dub delay works best when it’s selective. Think moments, not constant.

Now create a dedicated return track. Name it Return A, “Dub Delay.” Set the return fader at 0 dB. That’s important because you’ll control how much delay you hear from the send knobs and automation, not by riding the return all the time.

On this return, we’re going to build a simple, reliable chain: Echo first, then EQ Eight, then Saturator, then a Compressor with sidechain, and then Utility at the end. Timing first, cleanup second, character third, groove control fourth, and final width or level management last.

Alright, drop Echo on the return.

Set Echo to Sync mode so it locks to the tempo. Set Dry/Wet to 100%, because it’s a return. Now, for the key move: set Time to dotted eighth, 1/8D.

This is the classic jungle swagger setting. Dotted eighth has that skip. It leans forward. It creates a syncopated answer that feels like it’s pushing the track along, especially when you throw it off a snare or a vocal tail.

Now set Feedback somewhere in the 35 to 55 percent range. Start at about 42. At 170, you really want to respect feedback. If you drift into 70 percent territory, you’re basically asking for a constant cloud that steps on everything.

For Ping Pong, you can try it on, but don’t assume wide is better. Wide can be cool, but it can also blur in a club. We’ll manage that later with Utility.

For character, keep it fairly clean. If you want a little softness, add a small amount of Diffuse, but keep it subtle. And add a touch of modulation for tape-ish movement: set the modulation rate around 0.15 to 0.35 Hz, and the amount around 10 to 25 percent. You’re going for “alive,” not “chorus soup.”

Now let’s do a quick timing reality check, because this helps you predict clutter instead of guessing.

At 170 BPM, one beat, a quarter note, is about 353 milliseconds. An eighth note is about 176 milliseconds. A sixteenth is about 88 milliseconds. And dotted eighth, 1/8D, is about 265 milliseconds.

If your snare is landing on 2 and 4, the space between snares is two beats, about 706 milliseconds. So a dotted eighth repeat chain lands around 265 milliseconds, then 529, then 794. That third repeat often lands just after the next snare. That can be perfect, like a little tail that blooms after the hit, or it can smear your snare crack. So we’re going to build the rest of the chain to make it intentional and controlled.

Next device: EQ Eight after Echo. This is where you stop the delay from wrecking your low end.

Turn on a high-pass filter, 24 dB slope, somewhere around 180 to 300 Hz. Start at 220. This is non-negotiable for rollers. Low frequencies feeding back turn into mud fast, and they fight your bass and kick.

Then add a low-pass filter up top, around 8 to 12 kHz. Start at 10 kHz. This gives you that classic darker dub tail, and it stops the delay from turning into harsh hat splash.

If your delay is hissing over the snare crack, make a small dip around 2 to 4 kHz, maybe 2 to 5 dB, and listen. The point is not to dull the whole mix, it’s to carve the delay so the snare stays the main transient.

And here’s a teacher trick: automate the low-pass cutoff sometimes. On a turnaround, you can sweep it down from, say, 12 kHz to 4 or 5 kHz. That “getting eaten by tape” vibe reads instantly as dub, and it sounds like movement without adding more notes.

Next: Saturator after EQ. Set it to Analog Clip or Soft Sine. Drive around 2 to 6 dB, start at 3.5. Turn Soft Clip on, and then trim the output so the return isn’t clipping.

This does two things. It thickens the repeats so you don’t have to push them louder to feel them. And it rounds the transient edge, which helps the delay read as groove instead of sounding like a sharp copy pasted echo.

Now the secret sauce for rollers: sidechain ducking.

Drop a Compressor after Saturator. Turn on Sidechain. For the sidechain input, I recommend your clean snare layer if you have it. Ducking to the snare means the delay tucks out of the way right on the hit, then blooms in the gap after. That’s the “breathing” feeling that keeps things punchy and rolling.

Set ratio to about 4 to 1. Attack around 2 to 10 milliseconds, start at 5. Release around 80 to 180 milliseconds, start at 120. Then lower the threshold until you see about 2 to 6 dB of gain reduction when the snare hits.

Listen for the pocket. If the delay feels like it’s arriving too late, shorten the release. If it feels too chattery and nervous, lengthen the release so it glues a bit more.

If your snare pattern is inconsistent because the break is doing weird ghost stuff, you can create a dedicated ghost trigger track: a MIDI track with a short click or snare on 2 and 4, routed only to the sidechain, muted in the mix. That gives you consistent ducking every time, which is huge for tight rollers.

Now add Utility at the end.

If you’re using Ping Pong and it’s getting too wide, pull the Width down to somewhere between 70 and 100 percent. Or, better: keep the low end mono. Turn on Bass Mono and set it around 200 to 300 Hz so your delay doesn’t mess with club translation.

At this point, your return is built. Now we need to use it correctly, because this is where most people go wrong.

You’re not going to send everything to this delay all the time. Think of it as a weapon. Throws, not a bath.

Let’s do a classic move: snare throws.

On your snare layer track, automate Send A so it only spikes on specific hits. Great places are the last snare of bar 4, bar 8, bar 16, those phrase edges. In Arrangement View, draw a quick spike from minus infinity up to maybe minus 6 dB for just that hit, then back down.

What happens is the snare hits clean, the delay ducks under it because of the sidechain, and then the dotted eighth repeats create a mini fill that drags you into the next bar. That’s jungle storytelling right there.

Next, vocal or stab throws for atmosphere.

Take a one-shot “hey” or a reggae-style stab. Place it right before a drop or in the space of a breakdown. Send it into the dub delay, and automate the feedback for one bar: maybe from 35 percent up to 65 percent, then snap back down to 40. Combine that with a low-pass sweep down to 5 kHz on that same bar. You get this darker spiral moment, but it doesn’t take over the whole mix because you bring it back immediately.

Now a micro technique for breaks, because this one saves mixes.

Don’t send your entire break loop to the delay. Instead, duplicate your break track, chop out just one hat or ghost note, and send only that chopped piece into the delay return. That gives you rhythmic sparkle and movement without turning your whole drum bus into mush.

Let’s also talk about timing choices quickly, because 1/8D isn’t the only option.

1/16 is tight and fast, great for hats or metallic textures, but keep feedback low and high-pass higher, like 400 to 800 Hz, so it doesn’t clutter. Straight 1/8 is a safe default for a classic rhythmic echo that sits more predictably. 1/4 is huge and spacious, but at 170 it can swallow your groove if you overdo it. And if you have 3/16 available, it can get that “break science” off-kilter funk, especially layered quietly with high-pass.

A really good workflow is to duplicate the return, match the level, EQ, saturation, and ducking exactly, and only change the delay time. One return is 1/8D, one is 1/8. Then you A/B them while the loop plays. Don’t stare at the numbers. Listen for which one locks to your break swing.

Now, three arrangement moves that make this feel like real drum and bass.

First: 16-bar turnarounds. In bars 15 and 16, increase the snare send and nudge feedback up slightly. Then right before the downbeat of bar 17, cut feedback hard, or pull the send down to zero, so the drop lands clean. The contrast makes the drop hit heavier.

Second: call and response with bass. Let the bass do a one-bar phrase, then answer with a stab or vocal throw that delays in the next half bar. It’s a conversation, not clutter.

Third: drop impact control. In pre-drop, let the delay feel bigger and wider. On the drop, don’t just turn it down. Often it’s better to switch the timing shorter, like from 1/8D to 1/16 with lower feedback. The ear reads tightness as weight. You can keep the vibe without losing punch.

Common mistakes to avoid while you’re dialing this in.

Number one: leaving lows in the delay. Anything under about 200 Hz feeding back is instant mud. Number two: too much feedback at 170. It stacks fast. Number three: delay on everything. Rollers need clarity. Number four: no ducking. Without sidechain, your delay will step on the snare and kill your punch. Number five: over-wide ping pong. Headphones can lie. Clubs will tell the truth.

Now let’s do a quick 15-minute practice loop so you can lock this into muscle memory.

Load an Amen-style break and a clean snare layer. Build Return A exactly like we did: Echo, EQ Eight, Saturator, sidechain Compressor, Utility. Set Echo to 1/8D, feedback 42. Sidechain the compressor from the snare layer.

Loop 16 bars. Automate Send A on the snare so it only throws on bar 8 and bar 16, on the last hit of the bar. On bar 16 only, automate feedback up to about 60 percent for that one bar, and automate the EQ low-pass down to about 5 kHz. Then snap both back at bar 17.

After that, resample or freeze-and-flatten the return audio for 8 bars, and chop one or two delay tails into fills. Reverse one, pitch one up, fade one into the next phrase. This is how you turn “delay as an effect” into “delay as an arrangement asset,” which is extremely jungle.

Before we wrap, one extra advanced thought: if your ducking feels heavy-handed, you can make it more frequency-selective. One simple way is to EQ before the compressor so the compressor mostly reacts to the mids where the snare lives, leaving low-mid warmth more intact. Or go deeper with multiband ducking. The goal is the same: snare stays the loudest transient, delay lives in the gaps.

Alright, recap.

Build a dedicated return so you can throw delay rhythmically. At 170 BPM, start with dotted eighth for jungle swagger, then audition straight eighth and sixteenth depending on your break swing. Filter the delay hard: high-pass around 200 to 300 Hz, low-pass around 8 to 12 kHz. Add saturation for tape-ish bite. Sidechain duck it to the snare so it breathes with the groove. And automate sends, feedback, and filtering at phrase edges so it feels intentional.

If you tell me whether your roller is more two-step or full break, and what your snare pattern is, I can recommend the most musical time setting, like 1/8D versus 3/16 versus 1/8, and give you a throw map for a full 64-bar arrangement.