dEchorate

Quick Start

from datasets import load_dataset

# Stream to avoid downloading the entire dataset
ds = load_dataset("schismaudio/dechorate", streaming=True)

# Or download locally
ds = load_dataset("schismaudio/dechorate")

Dataset Description

dEchorate is a dataset of calibrated multichannel room impulse responses (RIRs) with echo timing annotations and 3D source/receiver positions. The RIRs were recorded in a cuboid room equipped with configurable acoustic wall panels, allowing systematic control over early reflection patterns.

Each RIR is accompanied by precise 3D coordinates for both the source loudspeaker and each microphone, plus annotated echo arrival times. This makes dEchorate uniquely suited for research in echo-aware signal processing, room geometry estimation, and acoustic simulation validation.

Dataset Structure

Data Fields

Field	Type	Description
audio	Audio	Multichannel RIR recording
filename	string	Original filename
source_position	string	3D coordinates of the source loudspeaker
receiver_position	string	3D coordinates of the microphone

Data Splits

This dataset has no predefined splits. All recordings are in the default train split.

Usage Examples

Load and inspect RIRs

from datasets import load_dataset

ds = load_dataset("schismaudio/dechorate")

sample = ds["train"][0]
print(sample["filename"])
audio_array = sample["audio"]["array"]
sr = sample["audio"]["sampling_rate"]
print(f"Sample rate: {sr}, Length: {len(audio_array)} samples")

Dataset Creation

Source Data

All RIRs were recorded in a cuboid room at the Fondazione Bruno Kessler (FBK) research center. The room was equipped with removable acoustic panels on each wall, enabling researchers to systematically vary acoustic absorption and measure its effect on early reflections. Source and receiver positions were calibrated with a laser rangefinder.

Annotations

Echo arrival times were annotated by the dataset authors using peak-picking on the measured impulse responses, cross-referenced with theoretical echo times computed from the known room geometry and source/receiver positions.

Known Limitations

• Single room geometry: All recordings are from one cuboid room. Results may not generalize to irregularly shaped rooms.
• Small dataset: The dataset contains fewer than 1,000 recordings, limiting its use as a standalone training set.
• Lab conditions: The controlled environment does not capture the acoustic variability of real-world performance spaces.

Related Datasets

This dataset is part of the Drum Audio Datasets collection by schismaudio. Related datasets:

• schismaudio/rirs-noises — Simulated and real RIRs plus noise recordings from OpenSLR

Citation

@inproceedings{dicarlo2021dechorate,
  title     = {dEchorate: a Calibrated Room Impulse Response Dataset for Echo-aware Signal Processing},
  author    = {Di Carlo, Diego and Tandeitnik, Pinchas and Foy, C{\'e}dric and Bertin, Nancy and Deleforge, Antoine and Gannot, Sharon},
  booktitle = {EURASIP Journal on Audio, Speech, and Music Processing},
  year      = {2021},
  doi       = {10.1186/s13636-021-00229-0},
  url       = {https://zenodo.org/records/4626590}
}

License

This dataset is released under the Creative Commons Attribution 4.0 International License (CC-BY 4.0).

You are free to share and adapt this dataset for any purpose, including commercial use, as long as you give appropriate credit to the original authors.