It is not a mode, and the confusion is costly
SBIR, for speaker boundary interference response, is often filed alongside room modes. They are two distinct phenomena, and confusing them leads to treating the wrong one.
A mode is a resonance of the whole volume: its frequency depends on the room dimensions, not on where the speaker sits. That is the subject of the page on room modes.
SBIR is an interference between the direct sound and a single reflection, off the boundary nearest the speaker. Its frequency depends only on the distance between the speaker and that boundary. Move the speaker and the frequency changes; move the wall and it changes too.
A practical way to tell them apart on a measurement: an SBIR null moves when you move the speaker, a mode does not.
The mechanism, in one sentence
Sound leaves the speaker towards you, but part of it also travels to the wall behind, reflects, and reaches you late. That delay corresponds to a path longer by twice the distance to the wall, there and back.
When that extra path equals half a wavelength, the two contributions arrive in opposite phase and subtract: that is the null. When it equals a whole wavelength, they add: that is a peak.
The first null frequency follows directly, with c the speed of sound and d the distance to the wall:
f = c / (4 × d)
The phenomenon repeats at odd multiples, so a second null at three times that frequency, with peaks landing on the even multiples.
The table that changes how you see the problem
Here is the first null frequency according to the distance from the speaker to the wall behind it.
| Distance to wall | 1st null | 2nd null | In the sensitive band? |
|---|---|---|---|
| 0.10 m | 858 Hz | 2,572 Hz | no, but in the midrange |
| 0.20 m | 429 Hz | 1,286 Hz | borderline |
| 0.30 m | 286 Hz | 858 Hz | ⚠️ yes |
| 0.40 m | 214 Hz | 643 Hz | ⚠️ yes |
| 0.50 m | 172 Hz | 514 Hz | ⚠️ yes |
| 0.75 m | 114 Hz | 343 Hz | ⚠️ yes |
| 1.00 m | 86 Hz | 257 Hz | ⚠️ yes |
| 1.50 m | 57 Hz | 172 Hz | no, modal territory |
| 2.00 m | 43 Hz | 129 Hz | no, modal territory |
The band flagged here, say 80 to 300 Hz, is where the null is heard most: it is where the bottom of male voices, the body of brass, the foundation of an orchestra and the punch of an effect all live. A hole of several decibels there impoverishes everything else.
Why « move it off the wall » is only half an answer
Read the distance column from top to bottom. Moving the speaker back does not make the null disappear: it lowers its frequency. From 0.30 m to 1.00 m you go from 286 Hz to 86 Hz, staying inside the sensitive band the whole way.
In other words, intermediate distances are the worst, and they are precisely the ones adopted spontaneously when a speaker is moved « a bit » off the wall. The advice, applied halfway, puts the null in the middle of the most troublesome band.
Two directions are coherent, and they are opposites.
Very close to the wall, under 0.20 m, the null rises above 400 Hz. It becomes narrower and the ear tolerates it better, but it enters the midrange, where timbre is at stake.
Clearly far, beyond 1.50 m, the null drops below 60 Hz, where room modes and the subwoofer take over. The problem is not removed, it is moved into a region already treated another way, as described in the page on subwoofer placement.
The choice therefore depends on the space available and on what you accept degrading. It is a geometric trade-off, not a recipe.
The one layout that removes the problem
There is one case where the null does not exist at all: the one where there is no rear reflection.
If the speaker is flush-mounted into the boundary, the wall becomes an extension of its baffle. There is no longer a there-and-back path behind the speaker, so no interference at that frequency. That is why professional rooms soffit-mount their front stage, and it is not a matter of appearance.
The gain is twofold: the null disappears, and half-space radiation brings a level boost in the bass. The cost is heavy, since it means building, and the side walls and ceiling still produce their own interference.
There is more than one wall
The calculation applies to every nearby boundary, and each produces its own null at its own frequency.
| Boundary | Distance, example | Null at | Note |
|---|---|---|---|
| Wall behind the speaker | 0.40 m | 214 Hz | the one everybody considers |
| Side wall | 0.70 m | 122 Hz | often closer still in a narrow room, hence higher in frequency |
| Floor | 0.90 m | 95 Hz | the distance depends on the height of the speaker's woofer, not its footprint |
| Ceiling | 1.50 m | 57 Hz | room modes take over at these frequencies |
| Wall behind the sofa | 0.50 m | 172 Hz | the same phenomenon, but at the ear: the distance is measured from the head |
Those four frequencies are computed with the same formula, on one example. The point is not their values but their spread: four different distances produce four staggered nulls, where four equal distances would have piled them up in the same place.
A useful consequence: different distances to each boundary are better than equal ones, since that spreads the nulls instead of stacking them. It is the same reasoning as for modal coincidences, applied to reflections.
The point no setting can fix
It is the same as for the subwoofer, and it is worth repeating here because the temptation is strong.
An SBIR null is a cancellation, not a lack of energy. Sending power into it by equalisation raises both cancelling contributions, consumes headroom and improves almost nothing. An absorber on the wall behind the speaker helps a little, by attenuating the reflection, but the thickness needed to act at 150 Hz is far greater than that of an ordinary panel.
The effective lever remains distance, and therefore geometry. It is decided on a plan, before installation.
Where the answer is different
The speaker is highly directional. Less energy travels backwards, so the reflection is weaker and the null shallower. The frequency, however, does not change.
The boundary is not reflective. A lightweight partition, a filled bookcase or a glazed bay do not return the same energy. The null still exists but its depth varies a great deal.
The speaker is far from you. The simple formula assumes the listener is much further away than the distance to the wall. In a small room, the exact geometry of the two paths changes the result somewhat.
You cannot move the speaker. As with the subwoofer, the variable then becomes the seat: the depth of the perceived null also depends on where you are.
What HTM computes
HTM computes, for each declared speaker, the distance to nearby boundaries, the corresponding cancellation frequency and an assessment of severity. You therefore see which speaker is a problem, at which frequency, and which one is most exposed, without doing the arithmetic by hand for four boundaries and seven speakers.
An interactive SBIR view, with a test point to drag through a section of the room and live readings for the four boundaries, is announced for version 1.2. See the Features page for the exact status of each function.
These are geometric predictions: the frequency is reliable, the severity depends on assumptions about directivity and boundaries. The full approach is to predict then verify by measurement, as described in the guide on measuring with REW. The Testimonials page reports a case where a null predicted around 160 Hz was found on measurement; that is one case, not a guarantee.
Sources
- Interference between direct sound and a reflection off a nearby boundary is classical acoustics. The relation
f = c / 4ddescribes the first null for a listener much further away than the boundary. - Every frequency on this page is computed for a speed of sound of 343 m/s, the usual value at 20 °C.
- The so-called sensitive band, 80 to 300 Hz, is a listening reference and not a standard.
Further reading
The page on room modes covers the phenomenon most often confused with this one, and the one on subwoofer placement applies the same reasoning to the bass. The glossary defines the terms used here.