Pre-amble: Don't worry about SBIR until you have handled more critical things like room reverberation (RT60) with room treatments to clean up your sound & proper subwoofer setup to solve the most egregious room modes. But once you have done that, if you are able, tackling SBIR is a good next step.
What is SBIR: Sound from your speakers travels in all directions, especially in the lower frequencies. That sound goes behind the speaker, and bounces off the wall behind the speaker -- and then combines with the speaker's output creating peaks and dips in the frequency response. This can happen with any boundary, so if side walls or the ceiling are within six feet of the speaker, it is happening. The frequency at which is happens changes based on the distance between speaker and wall. By carefully placing the speaker close enough to the wall that the frequency is high enough it can be killed by an acoustic panel, we can largely eliminate this source of distortion.
The Steps
1. Make sure we are not mistaking SBIR for Room Modes, or vice versa. If the frequency dip stays in the same frequency range when you move the microphone around the room, that is likely SBIR. If the frequency dip changes to different frequencies when you move the microphone around the room, then it is NOT SBIR.*
2. Okay, it is SBIR. What frequency should you worry about?
85/x=y hz
where x=meters from the wall to the front of speaker (the baffle).
For example: for a speaker whose baffle is .8 meters from the front wall, the math looks like this: 85/0.8=106hz.
In a typical system with an 80hz crossover, this 106hz SBIR dip represents a problem to address with an acoustic panel. (See step 4 for how thick and what kind of panel.)
3. The same math should be done for each other wall, floor and ceiling. Anywhere that the dip occurs above 80hz, in a system with an 80hz crossover, place a panel, if possible.
4. Then Bob Gold will tell you how thick a panel you need on that same wall (between speaker and wall, the distance you measured) in line with the speaker, to kill that problematic frequency.
5. The end.
TLDR: You have three options:
1. Put your speakers IN the wall, or
2. very close to the wall WITH thick panels between the speaker and the wall, or
3. c.8' from the wall.
-----
A very fine video:
How to Fix Speaker Boundary Interference (SBIR) Issues
In tonight's livestream at 11pm EST, we discuss Speaker Boundary Interference (SBIR) with Anthony Grimani and Matthew Poes. What is SBIR and how do you overc...
www.youtube.com
------
*Sometimes when you measure a room with REW, it is hard to tell: Is it SBIR or a Room Mode?
**Does insulation behind an acoustically transparent screen kill SBIR? Sometimes:
There are three main reasons you see lots of absorption behind acoustically transparent screens:
1. Comb filtering. This is reflections from the screen back towards the speaker. It can create small cancellations at some frequencies that impact the accuracy of the audio. You can kill this with one or two inches of absorption.
2. Bass trapping. The area behind a false screen wall is often a great deep (as in two or three feet deep) space that is otherwise unused for much more than the speakers. Hence it is a great space to build substantial bass traps out of layers of loose batts of insulation. This is the flexible cheap stuff you see at home centers.
3. SBIR. This is bass and mid bass energy from the speakers themselves that is largely omnidirectional that bounces off the wall behind the speakers and messes up the sound when it combines with the direct sound of the speakers. (It can also bounce off the side walls if they are within c. 5 feet of the speakers.). The exact frequencies can be calculated because it is based on the distance between the speaker baffle and the wall. Once you know the frequency of concern, you can make sure you address it with absorption that is thick enough....which might be anywhere from a couple inches to nearly a foot. To calculate it, read: this post but note, of course, that if you are doing the bass trapping mentioned in #2, you have probably solved for #1 and #3 as well.