DataRES
Authors/Creators
- 1. Aalto University, Department of Information and Communications Engineering
- 2. University of York, Department of Physics, Engineering, and Technology
- 3. L-Acoustics, UK
- 4. Friederich Alexander Universität, Department of Electrical-Electronic-Communication Engineering
Description
Dataset for research on Reverberation Enhancement Systems
A Reverberation Enhancement System (RES) is an active system for variable room acoustics. It uses microphones to capture the sound in a room and loudspeakers to reproduce the processed microphone signals, thus enhancing the reverberant field in the space.
DataRES comprises room impulse responses measured in rooms that host a RES. These rooms are used for research and development on RESs, either in the academic or industrial environment, or for hosting live performance events. The spaces differ in volume, geometry, and reverberation properties, and the multichannel audio setups differ in the number and position of transducers.
The data is first divided between the collected rooms. For each room, the data is further divided per processing block according to the model depicted in SignalModel.pdf. The processing blocks are the impulse responses between:
- stage musicians and audience - H_SA
- stage musicians and system microphones - H_SM
- system loudspeakers and system microphones - H_LM
- system loudspeakers and audience - H_LA
DataRES reports information for each room in the form of json files:
- datasetInfo.json contains the list of the collected rooms and their high-level information
- roomInfo.json contains the low-level information of one room. There is one for each room folder
- rirsInfo.json contains the list of recorded room impulse responses for the related room. There is one for each processing block
PyRES is the accompanying python package which provides and api to easily work with DataRES.
The work leading to producing the dataset and related research was supported by the Research Council of Finland, project No. 357391.