The Spheres Dataset: Multitrack Orchestral Recordings for Music Source Separation and Information Retrieval

2025-11-26

AI Summary

Classical music source separation and music information retrieval suffer from a lack of high-quality, professionally recorded multitrack datasets. Method: We introduce Orchestral-23, the first professional multitrack orchestral dataset—comprising complete movements, scales, and solo excerpts totaling over one hour—recorded using a 23-channel microphone array (including close, main, and ambient microphones) to simultaneously capture multichannel audio and room impulse responses, enabling controllable crosstalk modeling. The dataset provides high-fidelity isolated stems and spatial acoustic features. Contribution/Results: Leveraging Orchestral-23, we establish orchestral-family-level source separation and microphone crosstalk suppression baselines using the X-UMX model, demonstrating both feasibility and challenges of source separation in highly reverberant classical music scenarios. This work establishes a reproducible benchmark and introduces a new research paradigm for classical music signal processing.

Technology Category

• Machine Learning: Multimodal Learning
• Search and Optimization: Mixed Discrete/Continuous Search
• Cognitive Modeling & Cognitive Systems: Affective Computing

Application Category

• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Security and Privacy: Data transparency and provenance

Abstract

This paper introduces The Spheres dataset, multitrack orchestral recordings designed to advance machine learning research in music source separation and related MIR tasks within the classical music domain. The dataset is composed of over one hour recordings of musical pieces performed by the Colibrì Ensemble at The Spheres recording studio, capturing two canonical works - Tchaikovsky's Romeo and Juliet and Mozart's Symphony No. 40 - along with chromatic scales and solo excerpts for each instrument. The recording setup employed 23 microphones, including close spot, main, and ambient microphones, enabling the creation of realistic stereo mixes with controlled bleeding and providing isolated stems for supervised training of source separation models. In addition, room impulse responses were estimated for each instrument position, offering valuable acoustic characterization of the recording space. We present the dataset structure, acoustic analysis, and baseline evaluations using X-UMX based models for orchestral family separation and microphone debleeding. Results highlight both the potential and the challenges of source separation in complex orchestral scenarios, underscoring the dataset's value for benchmarking and for exploring new approaches to separation, localization, dereverberation, and immersive rendering of classical music.

Problem

• Advancing music source separation in classical orchestral recordings
• Providing multitrack data for machine learning in music information retrieval
• Addressing separation challenges in complex acoustic orchestral environments

Innovation

• Multitrack orchestral recordings with 23 microphones setup
• Dataset provides isolated stems for supervised source separation
• Includes room impulse responses for acoustic characterization