An acoustic analysis of the World’s ‘longest echo’

I’ve been asked questions via email and twitter about the acoustic measurements in the Inchindown oil storage tank, and so this blog is meant to answer some of those. If you have any questions, feel free to post them below and I’ll try and answer them.

Update June 2016

I’ve just returned to the tanks for a BBC 4 recording. This gave me an opportunity to measure the temperature and relative humidity in the tanks. At high frequency the reverberation time is dominated by the effects of air absorption, and to calculate this properly you need to know how much water vapour there is and the temperature. In the middle the values were 8.8 ºC and 83.8 %RH. When I wrote a scientific paper about the measurements I had to estimate the values based on the reverberation time measurements. My estimates of  10 ºC and 90 %RH were pretty good!

Why is the record for the ‘longest echo’ when the correct technical term is reverberation?

Guinness World Records calls it the ‘longest echo’ because that was the name of the record I broke. I did suggest they should change the name for the record to be more correct. The guidelines for the record asked me to measure the broadband reverberation time. So I did a measurement using the international standard ISO-3382.

What equipment did you use?

See my previous blog entry where you can also download the impulse responses.

How does the reverberation time vary with frequency?

Centre frequency (Hz)	125	250	500	1000	2000	4000
T30 (s)	112 ± 1	70.4 ± 0.3	43.2 ± 0.2	28.8 ± 0.2	17.8 ±  0.1	7.89 ± 0.07

Reverberation time vs frequency in Inchindown [1]

How did you calculate the record time of 75 seconds?

Guinness World Records wanted a single decay time across all frequencies. There are no standard methods for ‘averaging’ reverberation times across frequency, consequently a value was just calculated from the broadband impulse responses. Averaging all six measurement positions gave a value of (75±2) seconds.

Acknowledgement

Thanks to Paul Malpas, Director of Engineering Acoustics Design for doing an independent analysis of the recordings to check the calculations.

Notes

[1] The measurements were stopped when the decay could no longer be heard. For the 63Hz and 125 Hz octave-band the background noise floor of the measurement was never reached. Unfortunately, measurements were stopped too soon so there is an underestimation of T₃₀ in 63Hz octave band. (The T₂₀ =163±4 and EDT = 163±3s).

Advertisement