In this guide
The 30-second answer
Advanced audio is mostly about balance: voice on top, music underneath, effects as accents, and nothing fighting anything else. Duck the music under speech, clean up noise before you mix, and check the whole mix on phone speakers — the speakers your viewers actually use.
Viewers forgive shaky video. They do not forgive bad audio. A video with clean, balanced sound feels professional even when the footage is rough, and a video with muddy, uneven sound feels amateur even when the footage is beautiful. Audio is where a little extra effort pays off most.
This guide covers the advanced tools — ducking, cleanup, equalization, and layered mixing — and a workflow for using them without getting lost.
The advanced audio toolkit
Ducking automatically lowers the music whenever someone speaks, then brings it back up in the gaps. It is the single most valuable advanced audio tool, because it solves the most common problem — music drowning out voice — without manual volume automation. If you use one advanced tool, use this one.
Noise cleanup removes steady background sounds: hum, hiss, fan noise, room tone. It works best on constant noise and worst on unpredictable sounds like traffic or chatter. Apply it gently — heavy cleanup makes voices sound robotic and underwater, which is worse than a little noise.
Equalization (EQ) adjusts the balance of bass, midrange, and treble in a sound. Cut low rumble from voice recordings, tame harshness in the upper mids, and give music and voice different frequency homes so they stop competing. EQ is the tool that makes a mix sound clean rather than just loud.
Compression and normalization even out volume. Compression narrows the gap between quiet and loud moments so speech stays audible; normalization sets the overall level so the video is neither too quiet nor clipping. Together they are why professional videos sound consistent from start to finish.
Fade curves and audio keyframes give you manual control when automation is not enough. Fade music out under a key line, bring a sound effect up exactly on a visual beat, or dip the audio for a half-second pause. The keyframes guide covers the same idea for motion; audio keyframes do the same for sound.
Build a layered mix step by step
A mix has layers, and the order you build them matters. Voice first, music second, effects last — the same priority the viewer hears.
- Start with the voice track alone. Clean it up first — light noise cleanup, a low-frequency cut for rumble — and set its level so speech is clear and consistent throughout.
- Add the music track underneath, at a level where it is clearly audible in the gaps but never competes with speech. As a rule of thumb, music under speech should sit noticeably lower than the voice.
- Turn on ducking so the music dips automatically whenever the voice is present. Check the dips: they should feel smooth, not pumping. If the music pumps audibly, reduce the ducking depth.
- Add sound effects one at a time, each placed on a visual moment — a whoosh on a transition, a pop on a reveal. Effects are accents; if two effects overlap, one of them is unnecessary.
- Use EQ to separate the layers. Give the voice the midrange, roll the music’s mids back slightly, and keep effects short and bright so they cut through without volume.
- Listen to the whole mix on phone speakers, then on headphones. Phone speakers reveal balance problems headphones hide — especially muddy bass and buried speech. Fix what the phone reveals.
Mix at a moderate volume. Mixing loudly makes everything sound good and hides balance problems; mixing quietly exposes them. If a layer disappears at low volume, it was too quiet to matter. If a layer dominates at low volume, it will dominate everywhere.
Clean up noisy audio
Cleanup starts before the mix. Record in the quietest place you can, and capture a few seconds of room tone — silence in the same room — before you start speaking. That room tone is what cleanup tools use as a reference, and it makes the difference between a clean result and a robotic one. The voiceover guide covers recording technique in detail.
Apply cleanup in small doses. Run it once at a light setting and listen: is the noise reduced without the voice changing character? If the voice sounds thin or metallic, you have gone too far — back off and accept a little remaining noise. A natural voice with faint background noise beats a pristine-sounding robot every time.
Some noise cannot be cleaned. Wind, traffic, and overlapping conversations are unpredictable, and cleanup tools guess badly at them. For those, the fix is not more cleanup — it is re-recording the line, replacing the audio with a music bed, or cutting the noisy section entirely. Knowing when to abandon cleanup is part of the skill.
A repeatable mixing workflow
Mixing feels overwhelming because people do it all at once. Break it into passes, and each pass has one job. First pass: levels only. Set voice, music, and effects to roughly the right volumes, and nothing else. Second pass: cleanup and EQ. Fix noise and frequency clashes, still without touching automation. Third pass: ducking and fades. Automate the music under speech and smooth the transitions between sections. Fourth pass: the phone-speaker check. Listen on the smallest speaker you have, fix what it reveals, and stop.
Take breaks between passes. Ears adapt fast — after twenty minutes of mixing, your judgment of balance drifts, and loud things start sounding normal. A ten-minute break resets your ears more reliably than any tool. The best mixers work in short sessions for exactly this reason.
Compare against a reference. Pick a video whose sound you admire, play ten seconds of it, then ten seconds of yours, back to back. The comparison exposes problems your ears had normalized: too much bass, buried voice, harsh effects. Reference checks are the fastest way to close the gap between amateur and professional sound.
Fix common audio problems
Music drowns out the voice. Lower the music under speech, turn on ducking, or both. If the music still competes, the problem is frequency, not volume — use EQ to pull the music’s midrange back so the voice owns it.
Hiss or hum in the background. Light noise cleanup on the voice track, applied before mixing. If the hiss survives cleanup, the recording itself is the problem — re-record in a quieter room rather than fighting it in the mix.
The mix sounds muddy. Too many layers share the same frequencies. Cut low frequencies from everything except the layer that needs them, shorten overlapping effects, and give each layer its own frequency home with EQ.
Levels jump between clips. Normalize each clip’s level before mixing, so no clip arrives dramatically louder or quieter than its neighbors. Level jumps are the most common sign of an unmixed video, and the easiest to fix.
Pops and clicks at cuts. Add tiny fades — even a few frames — at every audio cut. A hard cut in audio is audible in a way a hard cut in video is not, and a micro-fade removes it without anyone noticing the fade itself.
When to use advanced audio — and when not to
Not every video needs advanced audio. A quick clip with a trending track and no speech needs almost none of this: pick the track, trim it, export. Advanced tools earn their time when a video has voice and music together, multiple audio layers, or a quality problem you can hear and name. If you cannot name the problem, you do not need the tool yet.
The trap is over-processing. Every tool you add changes the sound, and changes stack: cleanup plus EQ plus compression plus ducking can turn a natural voice into a flat, lifeless one. Apply one tool, listen, and stop when the problem is solved. The goal is not a heavily processed mix; it is a mix where nobody notices the processing at all.
Advanced audio also rewards patience more than gear. A phone microphone in a quiet room, mixed carefully, beats an expensive microphone mixed carelessly. The tools in this guide are free inside the app; the skill is in the listening, and that improves with every video you mix. Start with ducking on your next voice-and-music video, add cleanup when you need it, and let the rest follow naturally.
Frequently asked questions
Do I need external audio tools for advanced mixing? No. Ducking, cleanup, EQ, and level control inside Instagram Edits cover most videos. External tools help for specialized work — podcast-style editing or heavy restoration — but they are not a requirement for clean, professional-sounding reels.
Can I duck music automatically under speech? Yes — that is what ducking does. It lowers the music whenever voice is present and restores it in the gaps. Check the dips after applying it: they should feel smooth, and the music should still be audible in the gaps, not vanish entirely.
What is the best loudness for a reel? Aim for consistent, comfortable speech level with no clipping — the loudest peaks should never distort. Viewers adjust their own volume; your job is consistency, so nobody reaches for the volume button mid-video.
Does noise cleanup hurt voice quality? It can, if applied heavily. Light cleanup removes steady noise with little effect on the voice; aggressive cleanup makes voices sound thin and robotic. Use the lightest setting that solves the problem, and re-record rather than over-cleaning.
Can I mix on headphones? You can, but always do a final check on phone speakers. Headphones reveal detail and hide balance problems; phone speakers reveal balance problems and hide detail. Both checks together give you the full picture.
Ready to mix? Open Instagram Edits, add a voice track and a music track to a finished edit, turn on ducking, and listen on phone speakers. One pass with ducking alone transforms most videos.
Get Edits on Google Play Get Edits on the App Store
Next: All Tutorials →
Bottom line: a great mix is a balanced mix. Voice on top, music underneath, effects as accents, and a phone-speaker check before you export. The music and audio guide and the voiceover guide cover the building blocks; this page is the finishing layer. When the sound is clean, the whole video feels finished — and when it is not, no amount of visual polish will save it.