Welcome in. Today we’re doing feedback delay tricks in Ableton Live for drum and bass… but with one important safety rule: we are only doing this with resampling.

That means we’re not going to leave some crazy feedback loop running live and hope it behaves. We’re going to print the chaos to audio, then edit it like it’s a sample. That’s how you get those dubby throws, rolling spirals, and nasty transition tails without the classic “why is my project suddenly screaming” moment.

By the end, you’ll have a safe feedback-delay return you can use on basically any DnB project, plus three effects you can drop into your arrangements right away: a dub snare throw, a rolling hat spiral, and a neuro-ish scream tail.

Alright, let’s build the foundation first.

Step one: set up a resampling print track.

Create a new audio track. Name it RESAMPLE PRINT. On that track, set Audio From to Resampling. Set Monitor to Off. And don’t arm it yet. This track is basically your recorder for anything you hear in the project: delays, distortion, reverb, the whole chain. When you arm it and hit record, you’re printing exactly what’s coming out of the master.

Teacher tip here: resampling works best when you do it in short takes. Two to four bars at a time. Beginners often record one giant long pass, and then you end up scrolling for 20 minutes trying to find the good two seconds. We’re not doing that.

Now step two: build the actual safe feedback throw on a return track.

Create a return track, and name it A – FB THROW.

And we’re going to build a little stock-device chain that sounds properly DnB, but also has built-in guardrails.

First, add Utility. Set the gain to minus 6 dB to start. That’s just headroom. Optionally, push the width to about 120% if you want a bit more stereo vibe. Keep it reasonable for now.

Next, add Echo. If you don’t have Echo, you can use the older Delay, but Echo is perfect for this.

Set Echo to Sync mode. Start with a delay time of one eighth note. That’s the classic DnB throw timing. For a junglier swing, you can try three sixteenths later.

Set Dry/Wet to 100%, because it’s a return. And set Feedback somewhere around 70 to 88% as a starting range. We will push it higher, but we’ll do that when we’re ready to record.

Now inside Echo, use its filter. High-pass somewhere around 200 to 500 Hz to keep low end out of the feedback. And low-pass around 6 to 10 kHz to darken the tail. DnB delays are often darker than you think. Darker equals cleaner mix.

After Echo, add Saturator. Set Drive around 3 to 8 dB and turn Soft Clip on. Saturation helps the feedback feel more present, and soft clip helps prevent sudden spikes.

After that, add Auto Filter. Set it to low-pass mode, set the frequency around 8 kHz, and a little resonance, like 10 to 25%. If you want movement, turn on the LFO and keep the amount small and the rate slow. We’re going for “alive,” not “spinning out.”

And finally, add a Limiter at the end of the return. Always. Set the ceiling around minus 0.3 dB. This limiter is your seatbelt. Feedback is unpredictable by nature. The limiter makes it fun instead of scary.

Quick gain staging coach note: while you’re experimenting, try to keep that return channel peaking around minus 12 to minus 6 dB. You can always turn it up later. If it’s blasting while you design the sound, you’ll make timid choices and everything will still end up harsh.

Cool. Now we’ve got a return we can throw sounds into.

Let’s do Trick number one: the Dub Snare Throw.

The goal here is simple: on a chosen snare hit, we throw it into the feedback delay for a moment, print the tail, and then we place that audio tail like an arrangement element.

Go to your drum break, drum bus, or whatever track has your snare. Find the snares on 2 and 4.

Now automate Send A on that snare track. Keep it all the way down, basically off, most of the time. Then, on one snare hit, do a quick spike. Think of it like a momentary performance gesture, not a long fade. One sixteenth note to one eighth note long is a great starting point.

Level-wise, that spike might land anywhere from about minus 6 dB up to 0 dB on the send, depending on how hot your drums are. If you do it and it instantly overwhelms, pull it back. If it feels too subtle, push it up.

Arrangement idea: do this on the last snare before a fill, or the snare right before the drop. That’s a classic tension builder that doesn’t require any new sounds.

Now, on the return, push Echo feedback. Try 82 to 90%. This is where it starts to bloom into that rolling dub tail. If it feels too bright and it’s fighting your cymbals, lower the low-pass to around 6 to 8 kHz.

Now we print.

Arm the RESAMPLE PRINT track. Hit record. Record one to four bars covering the snare hit and the tail. Then stop.

Now edit that printed audio clip. Crop it so you’re keeping the good tail. Add short fades at the beginning and end so it doesn’t click. Then place that tail right after the snare, or let it lead into the next bar.

Optional shaping: if the tail feels too long and messy, add a Gate on the printed audio, not on the return. That way you’re shaping the sample, not changing the feedback behavior. And if it’s poking your ears, grab EQ Eight and gently reduce harshness in that 2 to 5 kHz area.

Before we move on, do the reset protocol. This saves your future self.
Turn the send back down. Bring Echo feedback back to something safe, like 50 to 60%. And if you moved the return fader, put it back where you like it. This is how you avoid “why is this project broken” later.

Alright. Trick number two: the Rolling Hat Spiral.

This is that hypnotic, rhythmic stereo texture you hear in jungle and modern rollers, where hats start to feel like they’re folding into themselves. We’re going to create it from something super simple.

Pick a closed hat loop or shaker pattern. One bar is enough.

On the hat track, send a steady amount to Return A. Start subtle, like minus 18 to minus 10 dB on the send. We’re not trying to replace the hats. We’re adding motion behind them.

Now go to the return and tweak Echo.
Set the time to one sixteenth for fast nervous energy, or three sixteenths for a more rolling swing.
Set feedback around 75 to 85%.

If Echo has modulation controls available, add a tiny bit of modulation. Mod amount very small, like 2 to 10%, and a slow rate. This is the difference between “static delay” and “liquid, moving texture.”

Now resample four to eight bars. Arm RESAMPLE PRINT, record, and stop.

Then we slice it.
Take the printed clip, and use Slice to New MIDI Track. You can slice by transients, or do one eighth notes for more consistent chunks.

Now you can play those slices like little rhythmic FX hits, or create that “barcode” stutter before a drop by placing a few one-sixteenth chunks rapidly in the final beat. Keep it quiet and dark. The trick is urgency without stealing the spotlight.

Mix tip: if your mix is already wide and messy, consider putting a Utility after Echo on the return and reducing width to like 60 to 80%. Or do the opposite if you want it to bloom: 140% can sound huge. Just remember, the wider it is, the more careful you have to be with brightness.

Trick number three: the Neuro-ish Scream Tail.

This one is for transitions. It’s aggressive, it’s dark, and it sounds like the track is ripping open for a second.

Choose a one-shot source. A bass stab, a reese hit, a foghorn, even a single snare impact can work.

Put it on an audio track. Then do a hard send to Return A for a brief moment. Something like minus 3 dB up to 0 dB on the send. It’s a throw, not a constant wash.

Now on the return, we go heavier.
Set Echo time to one eighth or one quarter.
Push feedback into the 85 to 92% zone. But only when you’re ready to print. This is “print-only territory.”

Set Echo’s filter darker: high-pass around 300 to 800 Hz, low-pass around 4 to 7 kHz. Dark and mean.

On Saturator, push the drive harder, like 8 to 12 dB, soft clip on.

Optional: add Redux before the limiter for some digital teeth. Keep it subtle. A little downsample goes a long way.

Now resample two to four bars while it evolves.

Once it’s printed, do the fun part: reverse it.
Go into the clip view and hit Reverse. Now you’ve got that vacuum-suck into the drop.

You can add a dark reverb after reversing if you want extra space. Or warp it, switch warp mode to Texture, and stretch it longer for more tension without sounding like a cheesy synth riser. Because it’s literally made from your own material, it automatically matches the vibe of your track.

Arrangement move that always works: place that reversed tail in the last bar before the drop, then cut to dry drums and dry bass on the downbeat. That contrast makes the drop hit harder than adding more layers.

Now the bonus concept: “freeze the feedback” without freezing anything.

Because we’re resampling, the trick is to push feedback until it’s almost self-oscillating, record a short pass, and then pick the sweet spot.

Crank Echo feedback until it’s right on the edge where it wants to run away. Record eight to sixteen seconds into RESAMPLE PRINT. Then audition and grab the best one to two seconds. Loop it, fade it, or slice it into fills.

This is controlled chaos. You’re capturing that moment where it sounds alive, then you’re turning it into a reliable piece of audio you can place anywhere.

Let’s quickly cover the most common mistakes so you can avoid the pain.

Mistake one: no limiter on the return. Feedback spikes happen fast. Always limiter last.

Mistake two: letting low end feed the delay. If bass and sub get into feedback, your mix turns to soup. High-pass early, either in Echo’s filter or with Auto Filter.

Mistake three: leaving feedback high while you keep working. You’ll forget, hit play later, and it’ll scream. Print it, then reset feedback to a safe value.

Mistake four: over-bright tails fighting cymbals and vocals. Roll off highs, and if needed, use EQ Eight to notch that harsh 3 to 5 kHz zone.

And mistake five: not committing. The whole point of this approach is committing to audio so you can treat FX like arrangement pieces: crop, fade, reverse, slice, place.

Now let’s lock it in with a mini practice exercise.

Build a basic 16-bar DnB phrase: kick and snare, hats, simple bass.

Then create three printed feedback moments.
In bar 4, do a snare throw at one eighth timing.
In bar 8, do a hat spiral at one sixteenth timing.
In bar 16, do a heavy scream tail at one quarter timing with distortion.

Resample each one into RESAMPLE PRINT, in short takes. Create an FX audio track and drag only the best parts there. Name them and color-code them right away. Future you will thank you.

Then arrange:
Put the snare throw tail across bar 4 into bar 5.
Layer the hat spiral quietly through bars 9 to 12 as motion.
Reverse the scream tail into bar 16, then hard cut to the drop.

And here’s a final pro-level mixing move that’s still beginner-friendly: after resampling, put EQ Eight on the printed FX track, switch it to M/S mode, and gently roll off some highs on the sides only. That keeps width without turning into hiss. You can also high-pass the sides a bit higher than the mid so the low-mid stays centered and clean.

Recap.

We used a return-track feedback chain, Echo into Saturator into filtering, with a limiter at the end. We used momentary sends like performance gestures. We resampled every time, edited the best moments, and used those prints as arrangement elements.

If you tell me your Ableton version and whether you’re using Echo or the older Delay device, I can describe a simple macro setup so you can control Throw, Feedback, Tone, Width, and Dirt from one rack and keep this workflow even faster.