Ever since I moved to the retirement community where I have lived since 2013 I’ve been struggling with the bad acoustics of our meeting hall. There is so much reverberaton that the sound from our PA system gets muffled and is prone to annoying feedback.
We are seniors, many with hearing loss. The staff does not include a crew of AV engineers to keep things tweaked up, and few residents are versed in how to use a microphone properly. So it’s not surprising that many meetings and performances were marred by bad sound.
Today audio engineers have hi-tech solutions to these problems: a digital toolbox of enhancements for the sound that compensates for the problems just mentioned. They have technical names like “equalizer”, “compressor”, “notch filter”, and “limiter.” They allow a public address (PA) to become a sound system tailored to the acoustics of the room. The categorical name for this technology is “Digital Signal Processors” or DSP.
A professional sound system for a medium-size theater can cost in the mid-five figures. We hired an acoustics expert to make a recommendation. He brought instruments and went about the hall firing a starter’s pistol and making other loud noises and in due course told us that acoustic wall treatments and such would not be much help, but that we should consider an advanced audio system featuring DSP for each of the speakers in the room. This would allow each speaker to be custom adjusted for each of the several different arrangements of chairs and performance areas.
These engineering adjustments would be saved in the audio console as presets so that house staff could simply select the preset that corresponds to the room layout for any given event. If successful, the result would be loud clear sound that was below the threshold of feedback.
We already had a DSP device capable of presets. It had been purchased and installed around the year 2000 and was a cutting-edge device with a price to match – about $1,200. The Sabine Graphi-Q converts the sound from an audio mixer to a digital stream, splits it into 31 one-octave “bands”, each of which has its own volume control. In addition is has 12 filters that can trap and tune out a squeal from feedback. But the processing doesn’t stop there, next it detects peaks of loudness in the stream and tamps them down (compression) essentially normalizing the loudest blasts. This has the effect of making the average audio less loud, the stream is amplified (gain) back to the original average level. And just in case the compression isn’t enough to prevent a super loud blast, there is another filter that imposes a loudness ceiling (limit).
All of this processing of the stream happens in less than two-thousandths of a second so it’s imperceptible to the listener. But if a room has more than one speaker, the transit time for sound is perceptible. Sound travels about 1,100 feed a second. If you are 50 feet from the source, the sound reaches you with a delay of about 50 thousandths of a second. When the sound from a second speaker close to the listener combines with the delayed sound, the interference between the sounds can make it fuzzy. That aggravates hearing loss issues. To fix this, the Graphi-Q has an adjustable delay setting that is set for the distance that the loudspeaker it controls is from the stage.
The delay is the last link in the chain of processes accomplished by the Graphi-Q processor. Installing and adjusting these devices is a lab course in digital signal processing techniques. It allows a novice technician to experience hands-on what he has learned about in textbooks.
The Graphi-Q is obsolete. On eBay you will find them offered for between $60 and $200, though Guitar City has one listed at the original $1,200 price. Sabine, the manufacturer, has closed its doors. But the company website has been preserved and archived (link below). That site is a gem of useful information because it includes whitepapers on signal processing techniques, acoustics, and the manuals and software for the Graphi-Q. This means that for about $100 you can buy a full featured laboratory to learn about DSP. It’s the lab course for Audio Engineering 101.