Have you ever walked on to a job site where they were trying for weeks to get the speech reinforcement system above a whisper without going into feedback? Sometimes it’s a gain structure thing, but sometimes it just can’t work. You can spend days spinning your wheels trying to fix something that is impossible to fix. It always reminds me of this scene in My Cousin Vinny: https://youtu.be/pWC0sKCS5oA?t=40s
“How do you know it won’t work? What’s…what’s this equation over here?”
“You never heard of PAG-NAGs?”
“Sure…sure I heard of PAG-NAGs. I just never actually seen a PAG-NAG before.”
*pauses…looks at lead field tech*
“Go ahead, honey. You gonna try it?”
Potential Acoustic Gain vs. Needed Acoustic Gain, PAG-NAG, equations allow us to tell if the level a system can deliver (Potential Gain) is more than the level it has to hit to be useful in the space (Needed Gain). If the Potential Gain is more than the Needed Gain, we have a stable system. If the Potential Gain of the system is less than the Needed Gain, the system will go into feedback before it gets loud enough. The system will be unstable. It’s as simple as that.
You might think you’d need to model the space, take into account the absorption coefficients of the walls and look into the pick-up pattern of the microphones used. However, it is simply based on geometry. The important measurements are stuff we would expect from experience, distance:
- from the talker’s mouth to the microphone
- microphone to nearest loudspeaker
- talker to nearest listener
- talker to furthest listener
- loudspeaker to furthest listener.
There are a few other pieces to the puzzle to make it complete.
The calculation takes into account the Number of Open Microphones (NOM) in the system, which, if you have an automatic mixer, will be NOM = 1, and have no effect on the stability. It also looks at whether the system can be equalized.
If the system can be equalized, a Feedback Stability Margin (aka “fudge factor”) of 6dB is included. If the system cannot be equalized, the FSM is 12dB.
The point is, once you put these easy to figure out numbers into a spreadsheet, you can immediately tell if a system will be stable. Not only that, but because the numbers are so easy to measure and imagine, you can adjust them in the spreadsheet until the system becomes stable (PAG > NAG). How? Again, these calculations just prove what we already know. The three “no-duh” geometric changes we can make to improve gain:
- Distance from the Talker’s Mouth to the Microphone
- Decrease this distance, and you get a lot more potential gain.
- If you go with gooseneck microphones on the table, instead of ceiling microphones, your potential gain will skyrocket.
- Distance from the Microphone to the Nearest Loudspeaker
- Increase this distance, and you get more potential gain.
- If you can set a “presenter area” and zone the loudspeakers, you can effectively move the microphone further away from the nearest loudspeaker.
- Distance from the Loudspeaker to the Furthest Listener
- Decrease this distance, and you get more potential gain.
- If you require less of a throw for the loudspeaker, they don’t have to work as hard.
- Instead of putting speech audio out of program loudspeakers near the front of the room that have to throw all the way to the back row, think about distributed loudspeakers.
- Instead of having to reach 50 feet from the front of the room to the back row, they only have to throw 6 feet, from the ceiling to the listeners’ ears.
Like most things with audio, putting a number on it just proves what we already know. The power is found in using these equations to predict audio performance before we get on site (or selling the idea to the client). If we spent time and effort proving the validity of the design before making promises to the users, we’d have a lot less explaining to do. Without a design review, we’d just be spinning our wheels like a ’64 Buick Skylark with a regular differential (no Positraction) stuck in the Alabama mud.
“This is your opinion, or is it a fact?”
“It’s a fact!”