The ECIRS Dataset – Ecophon Classroom Impulse Response Set

Accurate dataset of monoaural impulse responses (in. WAV) measured in a real classroom, capturing incremental changes in absorption and diffusion by the application of absorbing and diffusing acoustic materials. In total, 23 different acoustic treatment configurations are presented in the database. Each configuration is measured for two sound source positions and nine monoaural receiver positions, totalling 18 IR wave files, for each of the 23 acoustic treatment configurations. The dataset accounts for a total of 414 different IRs for the same classroom, for different source-receiver combinations and acoustical treatment configurations. The ECIRS dataset is free availability under a CC BY license, thanks to the support and altruistic contribution to science of Saint-Gobain Ecophon. This dataset is presented in a communication at the Inter-Noise 2025 conference, whose abstract is presented below: "Recent studies have shown that classroom noise and reduced speech intelligibility negatively impact concentration and learning. Additionally, poor room acoustics can cause vocal strain for teachers. These issues can be mitigated with appropriate acoustic treatment, but determining the level of required treatment for a significant improvement remains a challenge. While room acoustic standards define just-noticeable differences (JNDs) for various parameters, research has shown that JNDs vary based on room usage, sound emissions, and specific acoustic conditions. Therefore, perceptual studies remain essential for assessing changes in classroom acoustics that significantly affect intelligibility.To support research in this area, this communication presents a freely available accurate dataset of impulse responses measured in a real classroom, capturing incremental changes in absorption and diffusion. This resource is intended to help researchers conduct perceptual tests evaluating the effects of small acoustic modifications in classroom acoustics. Unlike simulated data, this dataset—derived from actual material changes in a real classroom— reduces uncertainty and inaccuracies in perceptual experiments. Finally, an example application of the dataset is presented, demonstrating its use in a perceptual study on the influence of early reflections on perceived sound impression at different classroom positions."