Do we want side wall reflections?

[Flaesh]

In a normal home audio room

Norms may vary. Ipsilateral reflections are virtually nonexistent with typical corner installations; the effect of contralateral reflections is more interesting. In my room, the random placement of windows and doors makes it difficult to use panels in these areas.
I'd be interested in @Chris A 's opinion regarding contralateral reflections.

 

[Flaesh]

Just buy..

.. a couple of decent 2-3 foot waveguides and a couple (or one) subwoofer arrays. The range in between is more complicated.

 

[Duke]

Norms may vary. Ipsilateral reflections are virtually nonexistent with typical corner installations; the effect of contralateral reflections is more interesting. In my room, the random placement of windows and doors makes it difficult to use panels in these areas.
I'd be interested in @Chris A 's opinion regarding contralateral reflections.

Obviously I'm not @Chris A, but (at the risk of over-simplifying) those first contralateral reflections result in low interaural cross-correlation with respect to the direct sound, which in turn makes a more beneficial contribution to spaciousness than the first ipsilateral reflections normally do. From the standpoint of spaciousness, the ears like low interaural cross-correlation in the reflection field. The relatively long time delay before the arrival of the first contralateral reflections does a good job of preserving the image precision inherent to the direct sound, which can be degraded by the arguably too-early arrival (and high interaural cross-correlation) typical of the first ipsilateral reflections.

 

[Chris A]

I'd be interested in @Chris A 's opinion regarding contralateral reflections.

First I want to verify that we're talking about the same thing: contra-lateral reflections...

In my experience, these crossing reflections are the reason why listening rooms can be too small. I've found that the human hearing system doesn't do very well with these type of reflections if they come too early--less than ~20ms from the direct arrivals. These reflections from the back of room, I've found, severely interfere with the illusion of soundstage and imaging.

This is an approximation based on the size of a 13' (4m) room size with the listener fairly close to the middle of the room). The "Haas interval" to detect two sounds is a minimum about 20-25ms (based on a couple of factors). If the sounds come too soon, then I find these reflections interfere with the sense of envelopment and spaciousness of the room's acoustics. The room is "too small" acoustically, and even if the loudspeaker directivity is quite good down to the Schroeder frequency of the room, these early reflections from the opposite loudspeaker interfere with the sense of envelopment..

If the room's dimensions/reflection times are significantly longer than this 20-25 ms, the situation is much different. The reflections from the side and back of the room then add significantly to the sense of envelopment.

Chris

 

[STC]

I've found, severely interfere with the illusion of soundstage and imaging.

In my experience with a real room using a 26-speaker array to create controlled virtual reflections and ambient fields for more spaciousness, I haven’t found early arrivals — including those contralateral/crossing ones — to severely interfere with soundstage and imaging the way the original post described. These reflection delays can be controlled to be in sync or in staggered delay spanning up to 80ms and probably more.

The core stereo phantom image stays remarkably stable as long as the reflection levels don’t get too high compared to the direct sound. It’s only when the virtual reflections start to mask or overpower the direct arrivals that I notice smearing or shifting.

This matches how I understand the precedence effect (Haas effect). The system basically prioritizes the first-arriving wavefront’s interaural cues (ITD/ILD) for direction and localization — that’s what holds the stereo image together. Later arrivals within the usual fusion window (~5–40 ms, depending on the signal) mostly get fused in rather than heard as separate sources, which helps keep the phantom image intact even with added reflections. A 2014 review paper by Brown (more of a medical rather than audio engineering paper) on the precedence effect describes this nicely: localization stays robust because we rely on the direct sound’s cues while the brain largely suppresses the conflicting ones from early reflections.

They’re related mechanisms but not exactly the same — one is about pulling position from binaural differences, the other about deciding whether later sounds get blended in or treated as new events. In small rooms, strong early contralateral reflections can still create practical problems like comb filtering, image broadening, or tonal shifts. But it’s usually not because they “violate the Haas interval” and wreck localization. The precedence effect is mostly doing its job of protecting the direct-sound imaging. The real issues tend to show up more as masking or level imbalance when the reflections get too strong.

ST

 

[Chris A]

The core stereo phantom image stays remarkably stable as long as the reflection levels don’t get too high compared to the direct sound.

Here lies the problem...

In my case, the greater portion of the acoustic energy is sent into the room, away from the adjacent walls just around the front of the room. If you are in a listening room as small as I identified above (13' or 4m on the longest linear dimension), I have found that the back of the room needs absorption to attenuate the main reflections bouncing off the back and adjacent rear side walls to keep the early back-of-the-room reflections under control.

However, if you are using loudspeakers that do not control their polars well (especially those reflections below 800-1000 Hz that attenuate much less than much higher frequencies) and that fill the front of the room with early reflections, then you have the same problem that a Bose loudspeaker has: no real soundstage image--even if those loudspeakers are mounted on elevated stands and are placed as far away from the walls as you can get them. The little bookshelf and "little speakers on top of lollypop stands" are included in this category, as do any loudspeakers having single direct radiating woofers, which lose their directivities below ~800 Hz, e.g.,

Chris