[Paper Review] Impulse response of the Bayreuth Festspielhaus

Kai Huang|ERef Bayreuth (University of Bayreuth)|Mar 21, 2017

Music and Audio Processing7 references3 citations

TL;DR

This study demonstrates that smartphone recordings of hand-claps can accurately measure the impulse response and reverberation time (RT) of the acoustically complex Bayreuth Festspielhaus, achieving quantitative agreement with professional measurements within a 500–4000 Hz frequency range. The method provides a low-cost, accessible approach for monitoring room acoustics using consumer-grade devices with proper procedural controls.

ABSTRACT

The Bayreuth Festspielhaus is well known for its architecture because its design is heavily influenced by composer Richard Wagner. Due to the special acoustic design, the reverberation time (i.e., time scale for the sound pressure level to decay $60$~dB) is larger than usual opera houses. Using hand-claps and smart phone recordings, I measured the impulse response of the Bayreuth Festspielhaus in the auditorium, on the stage, as well as in the orchestra pit. The measured reverberation time shows quantitative agreement with the literature values within a certain frequency range, demonstrating the possibility of using this approach to monitor room acoustics.

Motivation & Objective

• To evaluate the feasibility of using smartphone recordings and hand-claps as a low-cost method for measuring room acoustics in large, historically significant performance venues.
• To compare the reverberation time (RT) and center time obtained via smartphone recordings with established literature values from the same month (September 2014).
• To identify key sources of variability and error in amateur acoustic measurements using consumer devices.
• To develop a checklist for reliable data collection using hand-claps and smartphones in real-world acoustic environments.

Proposed method

• Hand-claps were used as impulsive sound sources at multiple locations: the auditorium, stage, and orchestra pit in the Bayreuth Festspielhaus.
• Three smartphones with 44.1 kHz sampling rate recorded the impulse responses, with microphones and AD converters varying across devices.
• Raw signals were filtered into octave bands (250–4000 Hz), and energy decay curves (EDCs) were computed via backward integration of squared envelope signals.
• Reverberation time (T15) was extracted by linear fitting of EDCs in a semi-logarithmic plane, using the ITA-Toolbox for ISO 3382-compliant analysis.
• Center time (ts) was calculated as the first moment of the squared impulse response to assess direct-to-reverberant sound balance.
• Multiple claps and recordings were performed at each location to improve statistical reliability and reduce noise-related errors.

Experimental results

Research questions

• RQ1 Can smartphone recordings of hand-claps produce reliable and quantitatively accurate reverberation time measurements in a large, architecturally complex opera house?
• RQ2 How do variations in microphone quality, distance, and direct sound levels affect the accuracy of RT and center time measurements?
• RQ3 To what extent do the measured RT values agree with professional literature values collected in the same month (September 2014)?
• RQ4 What procedural adjustments are necessary to minimize noise and variability in amateur acoustic measurements using consumer devices?
• RQ5 How does the spatial distribution of center time (ts) reflect the influence of source and receiver positions on early decay and speech intelligibility?

Key findings

• Reverberation time (T15) measured in the auditorium and on the stage showed quantitative agreement with literature values from Garai et al. (2015) for frequencies ≥500 Hz.
• At 250 Hz, RT was overestimated in both locations due to insufficient decay to -20 dB within recording duration and low signal-to-noise ratio (SNR).
• The direct sound amplitude varied by over an order of magnitude between smartphones, indicating strong device-dependent sensitivity and gain differences.
• Center time (ts) values showed high scattering across recordings, with the lowest values observed for the recorder closest to the source, confirming its sensitivity to direct sound influence.
• The average center time was 0.125 ± 0.023 s, suggesting speech intelligibility ≥80% when the speaker is near the proscenium, consistent with prior measurements.
• The study confirms that with proper protocols—such as sufficient inter-clap delay (≥1.5×RT) and multiple measurements—smartphone-based methods can yield reliable acoustic data for monitoring purposes.