How Audio Flows

Updated Oct 26, 2026 4 min read

ON THIS PAGE

Everything in Aura Mixer makes sense once you can picture the journey a sound takes through it. This page builds that picture. No knobs to turn yet — just the map.

The big picture

 SOURCES              INPUT CHANNELS            OUTPUT BUSES           DEVICES
─────────            ───────────────           ──────────────         ─────────
 🎙 Microphone  ──→  [ channel DSP ] ──┐   ┌─→ [ bus DSP ] ──→ 🎧 Headphones
 🎯 Spotify     ──→  [ channel DSP ] ──┼──→┼─→ [ bus DSP ] ──→ 🔊 Speakers
 🔊 Device      ──→  [ channel DSP ] ──┘   └─→ [ bus DSP ] ──→ 🔷 OBS (virtual)
                                       ROUTING

Four stages, always in this order:

  1. Sources — anything that produces audio: a mic, a single app, a whole playback device, a virtual cable.
  2. Input channels — each source lands on a channel strip and passes through that channel's processing chain.
  3. Routing — the matrix decides which buses each channel feeds. One channel can feed several buses at once.
  4. Output buses — each bus sums everything routed to it, applies its own processing, and delivers the result to a real device (headphones, speakers) or a virtual one (OBS, Discord).

The key insight: a channel is processed once, but can be heard everywhere — and every bus is its own independent mix. Pulling the music fader down in your stream mix doesn't touch what you hear in your headphones.

Inside an input channel

By default, audio passes through the channel strip's processors in the order a studio engineer would patch them:

IN → GAIN → GATE → EQ → COMP → FX inserts → FADER → SPATIAL → PAN → out to routing
Stage What it does Why this position
GAIN Sets a healthy input level (−20…+20 dB), with optional PAD (−20 dB) and phase invert Everything downstream works best with a sane level coming in
GATE Cuts low-level noise (fan hum, keyboard) when you're not speaking Gating before EQ/compression keeps noise from being amplified
EQ Shapes tone — up to 12 parametric bands Tone first, then dynamics react to the corrected sound
COMP Evens out loud and quiet moments Works on the EQ'd signal
FX inserts Your custom rack: de-esser, reverb, pitch… After the core chain, before the fader
FADER The strip's main level — your mixing control Mixing happens after processing, so the sound stays consistent at any fader position
SPATIAL / PAN Stereo width, Haas delay, then left-right placement — or, in Binaural HRTF mode, point-source HRTF positioning instead of pan Last, on the final stereo image — but only for stereo/mono sources (see the surround note below)

The middle of the chain — Gain, Gate, EQ, Compressor and any FX inserts — is re-orderable: on the FX tab you can drag cables between the nodes to change which runs first. Before it, a channel can invert its phase; after the compressor sits the optional clip protection limiter (SETUP → Calibration), then the fader. The HOME tab's signal-flow strip shows the default order and lights up each block that's active; clicking a block jumps to its tab.

💡 Tip

Processors you switch off aren't just bypassed — they're removed from the processing graph entirely and cost zero CPU.

⚠️ Surround sources fold first

When a 5.1 / 7.1 source is set to Binaural HRTF, the surround→headphone fold does not happen at the SPATIAL stage — it runs at the very start of the chain, before Gain, Gate, EQ and Comp. Those processors are stereo-only, so the multichannel signal is first decoded to a binaural stereo image, and the rest of the chain then shapes that result (your channel EQ/Comp act on the already-folded binaural signal). The "SPATIAL last" rule above applies only to stereo/mono positioning — width, panning, and point-source HRTF. An imported Custom HRIR profile also gets a Treble Softening control to tame any residual high-frequency edge. See Spatial — Binaural 3D.

Inside an output bus

Buses have their own, shorter chain, applied to the sum of everything routed in:

sum of channels → EQ → LIMITER → FADER → DELAY → device
  • EQ — correct the room or the headphones, not the individual sources.
  • LIMITER — a brick-wall safety ceiling that stops the summed mix from clipping.
  • FADER — the bus's overall level.
  • DELAY — shifts the whole bus in time, e.g. to sync audio with video (lip-sync).

Metering checkpoints

Every stage has meters, and they answer different questions:

  • Channel HOME tab, INPUT meter — is the source arriving at all, and at a good level?
  • Channel strip VU — what is this channel contributing after its processing?
  • Bus strip VU — what is the final mix on this destination doing?

When something sounds wrong, walking these meters left to right tells you where the problem starts — which is exactly how the No sound? checklist works.

Still stuck? Ask the community on Discord.
An unhandled error has occurred. Reload 🗙

Rejoining the server...

Rejoin failed... trying again in seconds.

Failed to rejoin.
Please retry or reload the page.

The session has been paused by the server.

Failed to resume the session.
Please retry or reload the page.