Audio Mixer
Layer several audio tracks together β voice over music, multiple instruments, or stacked recordings β with independent gain, pan and start offset per track, mute and solo, and a real-time preview that plays the actual mix before you export it.
Quick Tips for a Clean Mix
Features
Layering, Not Joining
It's worth being precise about what "mixing" means here, because it's easy to confuse with the Audio Merger already in this category, and they do genuinely opposite jobs. The Merger takes several clips and plays them one after another β clip one ends, clip two begins β which is exactly right for stitching together separately-recorded segments of one continuous podcast episode. This Mixer instead plays every track at the same time, layered on top of each other, which is what's actually needed any time the goal is a voiceover sitting over background music, several instrument parts combined into one recording, or two people's separately-recorded audio that needs to sound like they were in the same room.
Getting this distinction right up front saves a wasted attempt β running tracks meant to play together through a sequencing tool produces one long clip where they never overlap at all, and running sequential clips through a layering tool produces a garbled mess of everything talking over everything else at once.
The two tools also pair well together in the right order: several separately-recorded podcast segments get joined into one continuous file with the Merger first, and only afterward does that finished episode get layered here under a single music bed that runs the whole way through β mixing before merging would mean re-doing the music timing every time a segment's length changed.
A Real Mix You Can Actually Hear Before Exporting
Most free online mixers make you render a file just to find out whether the balance is right, then reload and re-render again after every small adjustment. This one plays the actual live mix through a real Web Audio signal graph β every track's gain and pan node responds the instant a slider moves, even mid-playback, so you're hearing genuine feedback rather than guessing and re-exporting repeatedly. The panning itself uses equal-power panning rather than simple linear panning, which keeps a track sounding like a consistent, stable volume as it moves across the stereo field instead of noticeably dipping in the middle, a detail many simpler tools skip.
Export uses that exact same signal graph, just rendered offline instead of played live, which is why what you hear during preview and what actually gets exported are the same mix β not an approximation of it. The one place they can diverge is Start Offset and the Master Limiter, both of which are set at the moment playback begins and can't be changed retroactively on an already-running track β adjust those, then press Play again to hear the update before exporting.
Building the Balance, One Track at a Time
A mix rarely comes together by setting every slider once and being done β it's an iterative process of soloing a track to check it in isolation, pulling something back that's fighting for attention with the lead vocal or narration, and nudging pan slightly to give two similar-sounding tracks their own space instead of collapsing into the same spot in the stereo field. Mute and Solo exist specifically to make that back-and-forth fast β soloing isolates exactly what you're currently listening for, and switching solo off returns instantly to hearing everything together again.
Gain staging matters more than it might seem: a track that needs to be boosted more than a few dB to be heard is usually a track that was recorded too quietly to begin with, and pushing it that hard in the mix tends to bring its background noise up audibly along with it. Running a quiet track through the Volume Booster & Normalizer first, before adding it here, usually gives a cleaner result than trying to fix the same problem entirely with this mixer's gain slider.
Who Actually Needs to Layer Tracks Together
Podcasters lay a narration track over a bed of background music, pulling the music down under the voice rather than fighting it for space. Musicians combine separately recorded guitar, vocal and drum-machine takes into a single rough mix without needing a full DAW installed. Video creators layer a dialogue track with ambient sound effects and a music cue. Teachers record a lesson's narration separately from its background music and combine them here rather than trying to play both live during one take.
A podcaster with a clean narration file and a royalty-free music track pulls the music down to around -15dB, sets Start Offset on the music to begin two seconds after the voice, and solos each in turn to confirm nothing masks the other before exporting. A musician layering three instrument recordings pans each one slightly differently β one left, one right, one center β so all three stay audible instead of collapsing into a single crowded mono-sounding blend.
What This Mixer Isn't
It's a genuine mixing tool, not a full digital audio workstation β there's no EQ, compression per track, automation over time, or fade curves beyond what the offset and gain controls already provide. For a straightforward layered mix of a handful of tracks, that's enough; for detailed sound design or professional music production, a dedicated DAW is still the right tool for that depth of control.
It also can't fix problems already baked into a source track β background noise, clipping from an over-hot recording, or muddy audio all carry straight through into the mix regardless of how carefully the gain and pan are balanced around them. Cleaning a track up first, whether that's trimming out unwanted silence or cutting it down to just the useful section with the Audio Trimmer & Cutter, tends to produce a noticeably better mix than trying to compensate for a messy source afterward.
And it's a mix, not a mastered final product β the limiter guards against the most obvious clipping when several tracks add up, but it isn't a substitute for genuinely careful gain staging beforehand. A mix that still sounds harsh or unbalanced with the limiter engaged usually needs individual tracks pulled back further, not a heavier limiter setting fighting against levels that were too hot to begin with.