v0.1.1107 Aug 2026·Media

Web Audio 3D spatial panning (PannerNode) - synthesis surface complete

The last common Web Audio node lands: PannerNode, which places a sound in 3D space relative to a listener. This completes an arc that took Web Audio from a silent stub to a comprehensive synthesis engine over three releases - oscillators, gain, constant and buffer sources, biquad filters, delay, waveshaper, a real analyser FFT, stereo routing (panner / splitter / merger), convolution reverb, and now 3D positional audio. The panner uses the equal-power model: it reduces the source-and-listener geometry to a horizontal azimuth, pans with the same equal-power law as StereoPanner, and attenuates by distance using the inverse, linear or exponential distance model. It stays deterministic by construction like the rest of the audio engine, verified against exact geometry: a source to the right is hard-panned right, a source in front is centred at 1/sqrt(2) per channel, and a centred source at twice the reference distance is attenuated to exactly one half by the inverse model. Media moves 38 to 39; overall holds at 51.

Parity with the reference browser

74overall, as of v0.1.343

Web APIs / DOM surface42

Security architecture66

GPU / graphics acceleration68

DevTools / extensions72

Networking85

Media85

Accessibility85

Real-world site compatibility99

HTML parsing100

CSS parsing + selectors100

Layout: block / flex / grid / tables / writing-modes100

Paint + visual100

JavaScript language100

JavaScript performance (JIT)100

HTML parsing100

What changed 3 changes

Media	PannerNode (equalpower model): the source position and the context AudioListener position/orientation are reduced to a horizontal azimuth, panned with the equal-power law, then attenuated by the distance model (inverse / linear / exponential, with refDistance / maxDistance / rolloffFactor). positionX/Y/Z params plus the legacy setPosition/setOrientation, and a real context.listener with its own position and orientation. Verified against exact geometry: right -> hard right, front -> centred at 1/sqrt(2), distance 2 under the inverse model -> exactly half gain.
Media	AudioListener is now a real object on the context: positionX/Y/Z and forward/up orientation vectors (defaulting to the standard front = -z, up = +y), plus setPosition and setOrientation. The panner reads it, so moving the listener moves every spatialized source.
Media	This completes the Web Audio synthesis surface: over v0.1.108-0.1.110 the engine went from stereo-less mono to stereo routing (StereoPanner / ChannelSplitter / ChannelMerger), convolution reverb (ConvolverNode) and 3D positional audio (PannerNode), each deterministic by construction. 4 new property tests (21 Web Audio tests total), conformance gate 6375/6375 with the native-JIT and no-nanbox builds green. Elevation and HRTF are the honest remaining panner gaps.