# Decoupling Magnitude and Phase Estimation with Deep ResUNet for Music Source Separation

arXiv: https://arxiv.org/abs/2109.05418 · HTML: https://mlsanity.com/paper/2109.05418/

## Key facts

- Cited by 48 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Sep 2021.

## Built on by (most cited first, 30 of 48)

- [AudioLDM: Text-to-Audio Generation with Latent Diffusion Models](https://mlsanity.com/paper/2301.12503/) (2023)
- [AudioLDM 2: Learning Holistic Audio Generation with Self-supervised Pretraining](https://mlsanity.com/paper/2308.05734/) (2023)
- [Text-to-Audio Generation using Instruction-Tuned LLM and Latent Diffusion Model](https://mlsanity.com/paper/2304.13731/) (2023)
- [Tango 2: Aligning Diffusion-based Text-to-Audio Generations through Direct Preference Optimization](https://mlsanity.com/paper/2404.09956/) (2024)
- [Hybrid Spectrogram and Waveform Source Separation](https://mlsanity.com/paper/2111.03600/) (2021)
- [Music Source Separation with Band-split RNN](https://mlsanity.com/paper/2209.15174/) (2022)
- [HiFi++: a Unified Framework for Bandwidth Extension and Speech Enhancement](https://mlsanity.com/paper/2203.13086/) (2022)
- [Separate What You Describe: Language-Queried Audio Source Separation](https://mlsanity.com/paper/2203.15147/) (2022)
- [SemantiCodec: An Ultra Low Bitrate Semantic Audio Codec for General Sound](https://mlsanity.com/paper/2405.00233/) (2024)
- [Separate Anything You Describe](https://mlsanity.com/paper/2308.05037/) (2023)
- [Moisesdb: A dataset for source separation beyond 4-stems](https://mlsanity.com/paper/2307.15913/) (2023)
- [Neural Vocoder is All You Need for Speech Super-resolution](https://mlsanity.com/paper/2203.14941/) (2022)
- [ResGrad: Residual Denoising Diffusion Probabilistic Models for Text to Speech](https://mlsanity.com/paper/2212.14518/) (2022)
- [FlowSep: Language-Queried Sound Separation with Rectified Flow Matching](https://mlsanity.com/paper/2409.07614/) (2024)
- [Audio Generation with Multiple Conditional Diffusion Model](https://mlsanity.com/paper/2308.11940/) (2023)
- [Multitrack Music Transcription with a Time-Frequency Perceiver](https://mlsanity.com/paper/2306.10785/) (2023)
- [The Chamber Ensemble Generator: Limitless High-Quality MIR Data via Generative Modeling](https://mlsanity.com/paper/2209.14458/) (2022)
- [CWS-PResUNet: Music Source Separation with Channel-wise Subband Phase-aware ResUNet](https://mlsanity.com/paper/2112.04685/) (2021)
- [FFC-SE: Fast Fourier Convolution for Speech Enhancement](https://mlsanity.com/paper/2204.03042/) (2022)
- [A Stem-Agnostic Single-Decoder System for Music Source Separation Beyond Four Stems](https://mlsanity.com/paper/2406.18747/) (2024)
- [Benchmarks and leaderboards for sound demixing tasks](https://mlsanity.com/paper/2305.07489/) (2023)
- [Learning Source Disentanglement in Neural Audio Codec](https://mlsanity.com/paper/2409.11228/) (2024)
- [MECAT: A Multi-Experts Constructed Benchmark for Fine-Grained Audio Understanding Tasks](https://mlsanity.com/paper/2507.23511/) (2025)
- [Semantic Grouping Network for Audio Source Separation](https://mlsanity.com/paper/2407.03736/) (2024)
- [Learning Temporal Resolution in Spectrogram for Audio Classification](https://mlsanity.com/paper/2210.01719/) (2022)
- [BrainECHO: Semantic Brain Signal Decoding through Vector-Quantized Spectrogram Reconstruction for Whisper-Enhanced Text Generation](https://mlsanity.com/paper/2410.14971/) (2024)
- [Segment-level Metric Learning for Few-shot Bioacoustic Event Detection](https://mlsanity.com/paper/2207.07773/) (2022)
- [MedleyVox: An Evaluation Dataset for Multiple Singing Voices Separation](https://mlsanity.com/paper/2211.07302/) (2022)
- [DIFFRENT: A Diffusion Model for Recording Environment Transfer of Speech](https://mlsanity.com/paper/2401.08102/) (2024)
- [Real-time Low-latency Music Source Separation using Hybrid Spectrogram-TasNet](https://mlsanity.com/paper/2402.17701/) (2024)

## Abstract (arXiv)

Deep neural network based methods have been successfully applied to music source separation. They typically learn a mapping from a mixture spectrogram to a set of source spectrograms, all with magnitudes only. This approach has several limitations: 1) its incorrect phase reconstruction degrades the performance, 2) it limits the magnitude of masks between 0 and 1 while we observe that 22% of time-frequency bins have ideal ratio mask values of over~1 in a popular dataset, MUSDB18, 3) its potential on very deep architectures is under-explored. Our proposed system is designed to overcome these. First, we propose to estimate phases by estimating complex ideal ratio masks (cIRMs) where we decouple the estimation of cIRMs into magnitude and phase estimations. Second, we extend the separation method to effectively allow the magnitude of the mask to be larger than 1. Finally, we propose a residual UNet architecture with up to 143 layers. Our proposed system achieves a state-of-the-art MSS result on the MUSDB18 dataset, especially, a SDR of 8.98~dB on vocals, outperforming the previous best performance of 7.24~dB. The source code is available at: https://github.com/bytedance/music_source_separation