MuCodec: Ultra Low-Bitrate Music Codec
Not on the map. Why?
Abstract (arXiv)
Music codecs are a vital aspect of audio codec research, and ultra low-bitrate compression holds significant importance for music transmission and generation. Due to the complexity of music backgrounds and the richness of vocals, solely relying on modeling semantic or acoustic information cannot effectively reconstruct music with both vocals and backgrounds. To address this issue, we propose MuCodec, specifically targeting music compression and reconstruction tasks at ultra low bitrates. MuCodec employs MuEncoder to extract both acoustic and semantic features, discretizes them with RVQ, and obtains Mel-VAE features via flow-matching. The music is then reconstructed using a pre-trained MEL-VAE decoder and HiFi-GAN. MuCodec can reconstruct high-fidelity music at ultra low (0.35kbps) or high bitrates (1.35kbps), achieving the best results to date in both subjective and objective metrics. Code and Demo: https://xuyaoxun.github.io/MuCodec_demo/.
Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). First posted on arXiv 20 Sep 2024.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2024: 1; 2025: 4; 2026: 4.
9 papers.
- 2024WavChat: A Survey of Spoken Dialogue Modelscited by 90
- 2025LeVo: High-Quality Song Generation with Multi-Preference Alignmentcited by 25
- 2025SongBloom: Coherent Song Generation via Interleaved Autoregressive Sketching and Diffusion Refinementcited by 23
- 2026Muse: Towards Reproducible Long-Form Song Generation with Fine-Grained Style Controlcited by 8
- 2026LeVo 2: Stable and Melodious Song Generation via Hierarchical Representation Modeling and Progressive Post-Trainingcited by 6
- 2025On the Design of Diffusion-based Neural Speech Codecscited by 1
- 2025DuoTok: Source-Aware Dual-Track Music Tokenization for Vocal-Accompaniment Generationcited by 0
- 2026S-PRESSO: Ultra Low Bitrate Sound Effect Compression With Diffusion Autoencoders And Offline Quantizationcited by 0
- 2026SqueezeComposer: Temporal Speed-up is A Simple Trick for Long-form Music Composingcited by 0
- 2026LeVo 2: Stable and Melodious Song Generation via Hierarchical Representation Modeling and Progressive Post-Training29 Jun 2026
- 2026SqueezeComposer: Temporal Speed-up is A Simple Trick for Long-form Music Composing22 Mar 2026
- 2026S-PRESSO: Ultra Low Bitrate Sound Effect Compression With Diffusion Autoencoders And Offline Quantization16 Feb 2026
- 2026Muse: Towards Reproducible Long-Form Song Generation with Fine-Grained Style Control7 Jan 2026
- 2025DuoTok: Source-Aware Dual-Track Music Tokenization for Vocal-Accompaniment Generation25 Nov 2025
- 2025SongBloom: Coherent Song Generation via Interleaved Autoregressive Sketching and Diffusion Refinement9 Jun 2025
- 2025LeVo: High-Quality Song Generation with Multi-Preference Alignment9 Jun 2025
- 2025On the Design of Diffusion-based Neural Speech Codecs11 Apr 2025
- 2024WavChat: A Survey of Spoken Dialogue Models15 Nov 2024
Similar, not cited
Close in wording, no citation either way
- 2026UniSRCodec: Unified and Low-Bitrate Single Codebook Codec with Sub-Band Reconstruction
- 2024SemantiCodec: An Ultra Low Bitrate Semantic Audio Codec for General Sound
- 2025FlowDec: A flow-based full-band general audio codec with high perceptual quality
- 2024Gull: A Generative Multifunctional Audio Codec
- 2021SoundStream: An End-to-End Neural Audio Codec
- 2025QINCODEC: Neural Audio Compression with Implicit Neural Codebooks
- 2022High Fidelity Neural Audio Compression
- 2026Ultra-Low-Bitrate Mel-Spectrogram-based Neural Speech Coding with Flow-Matching-based Refinement and Vocoding-driven Reconstruction ↗