Post-Processing to get Sound Pressure Level " SPL "

Meaningful Metrics – By Pat Brown

Pat Brown shows  the procedure for recording Multichannel wave files to provide the sound pressure level (SPL) history of the event.

So how loud is a college football game in an open-air stadium with 70,000 people in attendance? That information can be derived from the same recording that was used to produce the multi-channel playback files.


The A-Format file can be decoded to Omni using the software

provided with the TetraMic™. This is the file I will analyze to determine the sound pressure level. It was transferred to a hardware media player (Sound Devices 722™) and then played back into the sound card of the PC. Arta™ was used to perform the noise level analysis.


The simplest way to calibrate the analyzer is to measure the noise with a sound level meter while making the recording. I did this with the iPhone™ for a selected segment of the recording (Figure 1). The internal iPhone mic is adequate for casual investigations. I adjusted the analog output level of the 722 until the A-Weighted Slow SPL indicated in Arta matched what I measured on the iPhone for the same time period. I then analyzed the entire 4-minute segment (Figure 2).

Mono WAVE file to be analyzed for Sound Pressure Level

Analysis of a 4-minute segment

It is also possible to calibrate the Tetramic using a properly fitted external calibrator. Just record a few seconds of the calibrator tone onto the recorder to use a level reference for setting up the analyzer program. After decoding to omni, use this segment of audio to calibrate the analyzer. Steady music/chant – 95dBA on iPhone The method selected depends on the desired degree of accuracy. A calibrator should be used if you are doing measurements for the zoning board.

The Next Step

The only thing missing from the session at Commonwealth Stadium are some impulse responses. With prior planning (and knowing the right people) one could have a sweep or two played over the system during the pre-game. These could be decoded into impulse responses to provide some acoustical performance data on the space with an audience present. If repeated without the audience present, one could easily assess their effect on the sound reproduction. The collection of acoustical data need not be time consuming nor cumbersome. I did all of this with a small bag of gear and without stringing cables or AC power cords. Aside from some strange looks from those seated nearby, there’s little price to pay for getting some multi-channel recordings and SPL data at large events. pb


Speech over PA System - Unexpected pyro (and clipping!)

Speech over PA System – Unexpected pyro (and clipping!)