A comparison of T60 values obtained in a certified NVLAP reverberation chamber using conventional 1/3-octave decaying sound pressure level (SPL method) measurements and time-delay spectrometry (TDS) is presented. The TDS measurements were obtained from the least-squares slope of a backward Schroeder integration of the total energy density versus time for 16 fixed-bandwidth (1333 Hz) energy-time curves (ETC method) at 1/3-octave center frequencies and from a Peutz regression analysis of 1/3-octave averaged time-energy-frequency 3-D curves (TEF method). The SPL method utilized a rotating microphone and a rotating vane diffuser. The ETC and TEF methods were conducted with all combinations of rotating or stationary microphone and vane diffuser, to evaluate their effect. The best agreement between the SPL and ETC method was obtained using a spatial averaging of stationary microphone measurements with the rotating vanes stationary. The rotating vanes introduce the amplitude and frequency modulation interference which caused discrepancies at frequencies of 1000 Hz and higher, resulting in excessively large apparent T60s. On the other hand, the best agreement between the SPL and TEF method was achieved using a spatial average of stationary microphone measurements with the vanes rotating. Stopping the vanes in the TEF method caused large discrepancies at low frequencies of 500 Hz and below. This results from a decrease in the number of excited modes which occurs when the Doppler effect of the rotating vanes is removed. No advantage was realized infusing the moving microphone technique in the TDS procedures. Since efficient broad-bandwidth wide-angle fixed sound diffusers are now available, i the form of reflection phase gratings, it should be possible to create a uniformly diffuse sound field without rotating vanes, thus creating an environments where all T60 techniques could be performed accurately. A significant difference between our integrated total energy density curves (TETC) and those of other researchers using the integrated impulse response squared (IIR), is the absence of spatially dependent fluctuations, even at 125 Hz, in our results. A comparison between the IIR and IETC methods at 125 Hz, 250 Hz, and 500 Hz, for the condition where the microphone and vanes were stationary, revealed that the IIR curves were slightly more irregular but the overall backward integration envelopes were similar. T60s obtained from the IIR for these frequencies were approximately 4% lower than the IETC. Based on this comparison and the smooth linearity of the decay curves, we conclude that the sound field in the chamber was adequately diffuse.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.