aiqbench analysis
This arXiv paper proposes MixtureTT, a diffusion-transformer system that changes the instruments in a polyphonic recording directly from the mix, instead of separating stems first and transferring each one. The authors report gains over separate-then-transfer pipelines on objective and subjective tests, plus competitiveness with single-instrument methods given clean stems. Readers should note these are preprint results on specific SATB benchmarks, so generalization to wider real-world audio is still unproven.
Remix the Timbre: Diffusion-Based Style Transfer Across Polyphonic Stems
arXiv:2605.09259v2 Announce Type: replace-cross Abstract: Timbre transfer aims to modify the timbral identity of a musical recording while preserving its original melody and rhythm. While single-instrument timbre transfer has made substantial progress, existing approaches to multi-instrument settings commonly rely on separate-then-transfer pipelines that propagate source-separation artifacts and process the resulting stems independently. This paper proposes MixtureTT, to the best of our knowledge the first system for flexible per-stem timbre transfer directly from a polyphonic mixture. Given a mixture and a separate timbre reference for each target voice, MixtureTT jointly transfers all stems to the specified instruments. Its Joint Stem Diffusion Transformer first models cross-stem harmonic context and then refines each output under its own content and timbre conditions. All stems are generated within a single joint diffusion process rather than through separate per-stem diffusion runs. This design avoids propagating waveform-level separation errors through a cascade and produces more coherent multi-stem outputs. Despite operating under a harder input condition, MixtureTT outperforms practical separate-then-transfer pipelines on objective and subjective evaluations, while remaining competitive with single-instrument methods given ground-truth isolated stems. Results on both synthesized and real-audio SATB benchmarks further demonstrate its generalization. Controlled ablations show that jointly modeling cross-stem dependencies improves over an equivalent independent-stem variant, while the structure and timing of this coordination are also critical for effective mixture-level timbre transfer.
Abstract:Timbre transfer aims to modify the timbral identity of a musical recording while preserving its original melody and rhythm. While single-instrument timbre transfer has made substantial progress, existing approaches to multi-instrument settings commonly rely on separate-then-transfer pipelines that propagate source-separation artifacts and process the resulting stems independently. This paper proposes MixtureTT, to the best of our knowledge the first system for flexible per-stem timbre transfer directly from a polyphonic mixture. Given a mixture and a separate timbre reference for each target voice, MixtureTT jointly transfers all stems to the specified instruments. Its Joint Stem Diffusion Transformer first models cross-stem harmonic context and then refines each output under its own content and timbre conditions. All stems are generated within a single joint diffusion process rather than through separate per-stem diffusion runs. This design avoids propagating waveform-level separation errors through a cascade and produces more coherent multi-stem outputs. Despite operating under a harder input condition, MixtureTT outperforms practical separate-then-transfer pipelines on objective and subjective evaluations, while remaining competitive with single-instrument methods given ground-truth isolated stems. Results on both synthesized and real-audio SATB benchmarks further demonstrate its generalization. Controlled ablations show that jointly modeling cross-stem dependencies improves over an equivalent independent-stem variant, while the structure and timing of this coordination are also critical for effective mixture-level timbre transfer.
Sound (cs.SD); Artificial Intelligence (cs.AI)
arXiv:2605.09259 [cs.SD]
(or arXiv:2605.09259v2 [cs.SD] for this version)
https://doi.org/10.48550/arXiv.2605.09259 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 309 words
Related coverage
Runway’s WorldPrompt and the Engineering of Real-Time Worlds
Latent Space · Sep 25, 2026
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · Sep 17, 2026
Research Note: Filtering Subversion-Relevant Information From Pretraining Data Is Feasible
LessWrong · Oct 5, 2026
MAST: Mask-Guided Attention Control for Training-Free Regional-Multi Style Transfer
arXiv · Sep 9, 2026