Beyond Audio Gating: Using MIDI Triggers and Key Spikes for Tighter Drum Control
The Limits of Audio-Based Gating
Audio-based gating breaks down when the signal stops behaving predictably. It’s one of those drum mixing issues that expose the limits of level-based decisions, especially once ghost notes, bleed, and real-world dynamics enter the picture.
Traditional gates open and close based on the level of incoming audio. When the signal crosses the threshold, the gate opens. When it drops below, the gate closes.
This works fine when every hit is the same volume and there’s no bleed. It falls apart when hits vary in dynamics or when spill approaches the same level as the drum you’re trying to isolate.
Soft ghost notes don’t reach the threshold, so they vanish. Loud cymbal bleed exceeds the threshold, so the gate opens when it shouldn’t. You end up chasing settings: lower the threshold to catch ghost notes, and bleed gets through; raise it to reject bleed, and quiet hits disappear.
The underlying problem is that you’re asking the gate to make decisions based on level alone. Level doesn’t reliably distinguish “drum hit” from “not drum hit” when dynamics vary and bleed exists.
External triggering solves this by telling the gate exactly when to open, regardless of what the audio level is doing.
What Are Key Spikes?
A key spike is a short, consistent audio pulse that triggers a gate’s sidechain. Instead of the gate listening to the drum audio and deciding when to open, it listens to a separate track of spikes and opens when a spike arrives.
A key spike is a short, consistent audio pulse that triggers a gate’s sidechain.
The spikes are generated from MIDI notes that correspond to each drum hit. Because they’re MIDI-derived, they can be edited, quantised, and made perfectly consistent in level. Every spike triggers the gate identically, so every drum hit gets the same treatment regardless of whether it was a ghost note or an accent.
The key spike track doesn’t go to the mix. It exists only to control processors.
This approach gives you several advantages: - Ghost notes trigger the gate even though they’re quiet - Bleed doesn’t trigger the gate because there’s no spike for it - Gate timing becomes predictable and editable - You can quantise or humanise the gating independently of the audio
Generating MIDI from Drum Audio
Every major DAW has tools for converting audio transients to MIDI notes. The names vary, but the process is similar:
- Select your drum track. Start with the snare or kick, whichever has the clearest transients and least bleed. These are usually the most reliable sources for MIDI extraction.
- Run transient detection. The DAW analyses the audio and places markers at each detected hit. You’ll have sensitivity controls: set them so ghost notes get detected but false triggers from bleed don’t.
- Convert markers to MIDI. Each marker becomes a MIDI note. Some DAWs put them on a new MIDI track automatically; others require you to export and re-import.
- Clean up the MIDI. This is the critical step. Auto-detection is never perfect. Zoom in and check that every real hit has a note and no bleed or noise created false notes. Delete extras. Add any missed hits manually.
- Decide on quantisation. For tight genres (metal, EDM, modern pop), you might snap the MIDI to a grid. For looser styles, leave it unquantised so the gating follows the natural timing of the performance.
DAW-specific notes on MIDI extraction:
Pro Tools: Beat Detective can extract groove and create MIDI. There’s also the “MIDI Learn” approach with some third-party plugins that convert audio to MIDI in real time.
Logic: “Replace or Double Drum Track” in the Edit menu. Works well but sometimes merges nearby hits. Review and correct.
Cubase: Hitpoints detection, then “Create MIDI Notes” from the Audio menu. Good detection accuracy.
Ableton: Convert drums to MIDI via the right-click menu on audio clips. Creates a MIDI clip from detected transients.
Reaper: Dynamic Split can create MIDI from transients. Requires some setup but very flexible.
Studio One: Audio to MIDI via the Audio Bend panel. Detection quality depends on source clarity.
Third-party tools also exist for audio-to-MIDI conversion, some with more sophisticated detection algorithms that consider frequency content (useful for distinguishing kick from snare in a stereo drum bus).
Creating the Spike Audio
Once you have clean MIDI, you need to turn it into audio spikes that can feed a sidechain.
- Create a sampler instrument track. Use your DAW’s built-in sampler or a simple plugin.
- Load a short, sharp sample. A click, blip, or even a synthesised transient. It doesn’t matter what it sounds like; it only needs to be short (under 10ms) and consistent.
- Turn off velocity sensitivity. Every MIDI note should trigger the same spike at the same level. You want consistency, not dynamics.
- Route the MIDI to the sampler. Play through the session and verify that every drum hit produces a spike.
- Print the sampler output to audio. Render or bounce the spike track so it exists as audio. This ensures timing accuracy and reduces CPU load.
- Mute the spike track’s output. It shouldn’t go to your mix bus. It only exists to feed sidechains.
Label the track clearly (“Snare Spike” or “Kick Trigger”). Colour-code it differently from your audio tracks so you don’t accidentally unmute it.
Routing Spikes to Gates
With spike tracks ready, routing is straightforward:
- Insert a gate on your drum track. Any gate with external sidechain input works.
- Set the sidechain input to the spike track. In most DAWs, this is a dropdown menu in the gate’s sidechain section.
- Lower the threshold. Since the spikes are consistent, you don’t need a precise threshold. Set it low enough that every spike opens the gate reliably.
- Focus on attack, hold, and release. These now shape the envelope purely based on your musical intent, not on trying to avoid false triggers.
Because the spikes define exactly when the gate opens, you can set aggressive timing without worrying about ghost notes getting cut or bleed sneaking through. The gate is no longer making decisions; it’s following instructions.
Because the spikes define exactly when the gate opens, you can set aggressive timing without worrying about ghost notes getting cut or bleed sneaking through.
Routing Spikes to Compressors
The same principle works for compression. Route the spike track to a compressor’s sidechain, and the compressor responds identically to every hit.
This is useful when you want:
- Consistent gain reduction across varying dynamics
- Predictable attack shaping (the compressor always “sees” the same level)
- Rhythmic pumping effects that follow the drums precisely
Set the compressor’s threshold relative to the spike level, not the drum audio. Then adjust attack and release to shape the sound.
For parallel compression, you can use spikes to trigger the compressed signal while leaving the dry signal unprocessed. This gives you control over exactly which hits get compressed treatment.
Routing Spikes to Transient Shapers
Some transient processors accept external triggers. When they do, the spike tells the plugin when to apply attack boost or sustain cuts.
This is helpful when the drum’s natural transient isn’t clear enough to trigger the shaper reliably, or when bleed causes inconsistent shaping.
Not all transient shapers support external input. Check your plugin’s documentation.
When Key Spikes Are Worth the Setup
Key spikes add a layer of complexity. You’re creating extra tracks, managing MIDI, and routing sidechains. That’s not always necessary.
Use key spikes when:
- Dynamics vary significantly and ghost notes keep disappearing
- Bleed level is close to drum level, causing false triggers
- You need machine-consistent gating (metal, EDM, sample-precise genres)
- You’re layering samples and need triggers to fire samples in sync
- You want to quantise the gating without quantising the audio
Skip key spikes when: -
- Dynamics are consistent and bleed is manageable
- Traditional gating is working fine
- The extra setup time isn’t justified by the improvement
- You’re working in a genre where slightly imperfect gating is acceptable or even desirable
Combining Audio and MIDI Triggering
A hybrid approach often works well:
- Use audio-based gating as the first pass. Clean up obvious bleed with threshold-based gating.
- Add key spike triggering for refinement. The spikes handle ghost notes and inconsistent hits that the audio gate missed or cut.
You can run both in series (audio gate first, then spike-triggered gate) or use them on different elements (audio gating for kick, spike triggering for snare).
Another hybrid: use the spike track to trigger sample replacement or reinforcement while using audio gating on the live drums. The spike ensures perfect sample alignment; the audio gate handles the acoustic cleanup.
Troubleshooting
Gate misses soft hits even with spikes
Check that the spike track has notes for all hits, including ghost notes. If detection missed them, add MIDI notes manually. Also verify that the threshold is set low enough relative to the spike level.
Gate opens too early or too late
Nudge the spike track. If the gate opens early, delay the spike track slightly. If late, advance it. Remember that gate attack time adds latency: if your gate has a 1ms attack, the spike should arrive 1ms before the audio transient (or you can use plugin delay compensation if available).
Rolls and flams create problems
MIDI detection often merges rapid hits or misses some notes entirely. Manually add spikes for flam grace notes. For rolls, you may need to increase detection sensitivity or draw in the MIDI by hand.
Alternatively, automate the gate to bypass during rolls and let the audio gating handle those sections, using spikes only for single hits.
Hardware triggers double-fire
If you’re using physical drum triggers (not MIDI from audio), adjust the trigger sensitivity and retrigger time on the module. Double-fires usually happen when the head vibration retrigs before the module’s “ignore window” expires.
Spikes audible in the mix
The spike track’s output should be muted or routed to “No Output.” Check your routing. The spike track feeds sidechains only; it should never reach the mix bus.
Live and Hybrid Contexts
External triggering is common in live sound for exactly the reasons that make it useful in the studio: unpredictable dynamics and severe bleed.
Hardware drum triggers (piezo or contact sensors mounted on shells) generate voltage spikes when the drum is struck. These signals feed hardware gates, sample modules, or IEM systems.
In hybrid setups where electronic pads and acoustic drums coexist, MIDI from the electronic kit can trigger gates on the acoustic mics. The drummer hits a pad; the MIDI note opens the gate on a parallel acoustic mic. This gives you the consistency of electronic triggering with the tone of acoustic drums.
In hybrid setups where electronic pads and acoustic drums coexist, MIDI from the electronic kit can trigger gates on the acoustic mics.
Latency matters in live contexts. Triggers need to precede or coincide with the audio; if they arrive late, the gate opens after the transient has passed. Most trigger systems are designed for minimal latency, but verify your setup and adjust if needed.
Beyond Gating: Creative Uses for Triggers
Once you have reliable triggers, you can use them for more than cleanup:
Sample layering. Trigger samples to reinforce weak hits or add consistent tone under the acoustic drums. Align samples to the spikes for perfect timing.
Reverb/effect triggering. Instead of putting reverb on a drum and gating the reverb, trigger the reverb with the spike track. This gives you reverb on every hit, even quiet ones, with consistent tail lengths.
Sidechain pumping. Use drum spikes to trigger sidechain compression on synths, pads, or bass. More precise than audio sidechaining because the spike doesn’t have variable dynamics.
Rhythmic gating on other elements. Use the drum spike track to gate or duck other instruments in time with the drums. Creates tight, interlocked grooves.
Sample Triggering: Phase and Philosophy
If you’re using triggers to fire samples (whether for reinforcement or replacement), two things will determine whether the result sounds natural or uncanny.
Phase alignment, not just time alignment. A triggered sample needs to be in phase with the original drum, not just lined up in time. If the sample arrives at the right moment but its waveform is inverted relative to the acoustic drum, the two will cancel rather than reinforce. The result is a thinner sound, not a bigger one.
If the sample arrives at the right moment but its waveform is inverted relative to the acoustic drum, the two will cancel rather than reinforce.
Check this the same way you’d check any phase relationship: flip the polarity on the sample track. If the combined sound gets bigger, leave it flipped. If it gets smaller, flip it back. Then nudge timing until the attack is tight. Phase first, then time.
Blend vs replace: know which you’re doing. Adding 10-20% of a consistent sample under the acoustic kick to fill out the low end is very different from replacing the kick entirely.
Blending is subtle reinforcement. The sample provides body, consistency, or frequency content that the acoustic sound lacks. Done well, nobody notices it’s there. The acoustic drum is still the sound; the sample just supports it.
Replacing is substitution. The sample becomes the primary sound; the acoustic drum is reduced or muted. This gives you total control but risks the “uncanny valley” where something feels wrong but listeners can’t identify why. Replaced drums often sound too perfect, too consistent, or disconnected from the performance.
Most engineers working on real drum recordings are blending, not replacing. The goal is to enhance, not substitute. Keep the sample level low enough that muting it makes the drum sound weaker, not different. If muting the sample changes the character entirely, you’ve crossed from reinforcement into replacement.
A Practical Key Spike Workflow
For a typical multitrack drum session:
- Extract MIDI from kick and snare tracks. These are your primary triggers. Clean up the MIDI carefully.
- Create spike tracks for kick and snare. Use a simple click sample with velocity sensitivity off.
- Set up tom triggers if needed. Toms benefit from spike triggering when bleed is severe or when you’re reinforcing with samples.
- Route spikes to gates on each drum track. Set low thresholds, then dial in attack/hold/release based on tempo and feel.
- Test with the full kit. Play through the session and verify that gates open correctly for every hit. Adjust spike timing or add/remove MIDI notes as needed.
- Use automation to bypass spike gating during sections where it causes problems (very fast rolls, extremely dynamic passages).
- Blend with audio-based processing where appropriate. Spikes don’t have to do everything; they’re another tool in the workflow.
Where This Fits
MIDI triggering and key spikes represent the most controlled end of the drum cleanup spectrum. At the other end is minimal processing: polarity flips, gentle EQ, and accepting the bleed as part of the sound.
Most sessions fall somewhere in between. Phase alignment gets the mics working together. Bleed control (gating, filtering, manual editing) cleans up the worst spill. Decay shaping balances resonance against noise. And triggers provide surgical precision when needed.
The decision about which tools to use depends on the genre, the quality of the tracking, and how much time you have. Key spikes take effort to set up, but on the right session, they solve problems that no amount of threshold tweaking can fix.