English

RIPPLE: Generating Multi-Channel Phase, Not Recovering It

Machine Learning 2026-07-30 v1 Sound

Abstract

Generative models synthesize magnitude spectra with high fidelity, while phase is delegated to a recovery module---Griffin--Lim, a vocoder, or a latent decoder---applied independently to each channel. For multi-channel waveforms this delegation is costly: the physical content of spatial audio and three-component seismograms lives in the phase relationships between channels, precisely what channel-independent recovery cannot produce. The cost is also invisible, since the magnitude-based metrics common to both fields barely move when inter-channel phase coherence collapses---so a pipeline can discard the physical information in its output while still scoring well. We argue that phase should be generated, not recovered, and present RIPPLE (Rectified Inter-channel Phase with Prior-based LEarning), which reinterprets Griffin--Lim as a phase **prior** rather than a final estimator: initialized from the source phase, this prior carries the inter-channel structure to be preserved, and a rectified flow refines it toward the target under an explicit inter-channel phase loss. Tested on first-order ambisonics environment transfer and seismic cross-station translation---two physically unrelated domains---RIPPLE outperforms recovery-based pipelines on the coherence metrics that downstream analyses consume. The seismic case is decisive: across architecturally distinct generators, per-channel recovery leaves S-wave polarization error near the 57.357.3^\circ random expectation, whereas learned phase reduces it to 33.833.8^\circ.

Cite

@article{arxiv.2607.27775,
  title  = {RIPPLE: Generating Multi-Channel Phase, Not Recovering It},
  author = {Jaehyuk Lee and Yeajin Lee and Dayeon Shin and Donghun Lee},
  journal= {arXiv preprint arXiv:2607.27775},
  year   = {2026}
}