Room acoustics

Hmm, your ears can tell you a lot, but sometimes there are some specific issues that are audible but not easily remedied. These peaks and nulls, which seem to be persistent once the easy room treatments are done, are so specific in their frequency band, that I think you'd be hard pressed to identify and fix them by ear. At least that's my experience. A little bit of data can be useful.
But your mind is still the ultimate decision maker on sonic performance. So it does have to sound good, whatever it measures

My secret methodology for treating room anomalies. You need to find and map out all the sound pressure peaks in the room, incluing the 3 dimension space of the room. The most effective way to do this is using a test tone (s) and SPL meter. I select test tone(s) around 315 Hz or thereabouts since that frequency is very revealing, and effective, but you can use other frequencies, too. This will allow you to find all sound pressure peaks that are say, 6 dB (i.e., 4 times higher!) above the average sound pressure in the room. Generally speaking, these peaks will be found in room corners, at first reflection points and on the wall between the speakers, on the wall behind the listener but also at random and unexpected locations around the room, including in the 3D SPACE of the room. Sometimes right where the listening chair is located! 😧 All of those peaks 6 dB or higher act as "speakers" that interfere with the primary sound produced by the speakers. The objective is to reduce those peaks as much as possible - without producing side effects. But trying to treat the room by ear is a fool’s errand. It’s like trying to solve n simultaneous equations in n + x unknowns. Sure, you can do better than nothing but you will only find a local maximum, not the real maximum.

There are a great number of devices that can be employed to defeat the peaks once you’ve located them. I can vouch for Traps, Hemlholtz resonators, tiny little bowl acoustic resonators, crystals, Mpingo discs, Room Tunes Echo Tunes, Golden Sound’s Acoustic Discs, Skyline diffuser, but not (rpt not) SONEX.

This may be the moment to introduce the so-called Schroeder frequency, named after the German physicist Manfred Schroeder. The Schroeder frequency denotes the crossover frequency between the chaotic behaviour of sound waves above it and the discreetly spaced peaks and dips/nulls below it. Above the Schroeder frequency sound waves bounce around the room, producing many small spikes that are closely packed along the frequency range. Below the Schroeder frequency the peaks occur at the resonant frequency of the room's dimensions, and create big peaks at particular frequencies and again at their upper harmonics. These so-called room modes are large and pretty far apart, so they are quite obvious to the listener, with a boomy bass that often lingers on at particular frequencies. See here for more explanation: https://www.soundandvision.com/content/schroeder-frequency-show-and-tell-part-1
The Schroeder frequency can be calculated approximately, and depends on a room's dimensions. The larger the room, the lower the Schroeder frequency, from about 200 Hz in smallish rooms down to about 100 Hz for a large room. See here for a calculator: http://www.mh-audio.nl/sg.asp
Knowing the Schroeder frequency of your room is important, because treatment above it has to be quite different from treatment below it. See here for more discussion: http://www.linkwitzlab.com/rooms.htm
To put it simply, treating the frequencies above the Schroeder frequency is a matter of adding damping material like rugs, bookcases etc. It can often be achieved without too much intrusion into the style of your home decor, although there are also more visually imposing solutions from the world of studio design. The exception to the idea that dealing with these higher frequencies would be quite easy is if like me you prefer a modern minimalist interior. Such rooms have a hard acoustic and softening the acoustic without changing the style of the interior is a challenge.
Conversely, adding rugs and the like to reduce peaks below the Schroeder frequency is pointless. What you are dealing with here is big resonant peaks that can be tamed in only three ways. The first is so-called bass traps. Unfortunately at the frequencies we are talking about these bass traps are necessarily large and ugly. The second is multiple subwoofers. Main speakers have to be located for best mid range response and imaging, and that is not necessarily best for bass reponse. So separating bass response from the main speakers allows you to locate the bass speakers at the best spot for them. Traditionally HT subwoofers were used alone, but it is now increasingly understood that using multiples smoothens their response because their response peaks and dips do not coincide. See here for some explanation: http://www.acousticfrontiers.com/20101029using-multiple-subwoofers-to-improve-bass-the-welti-devanti... You get twice as many peaks, but of much smaller amplitude, and that sounds a lot better. To be sure, this is still not an argument in favour of stereo subs. It remains true that at these low frequencies sound is not directional, so dual subs are mostly still connected as mono subs for a somewhat smoother response compared  to stereo subs.
Finally, remaining peaks may be equalized by dsp eq units like the Antimode 8033 for subwoofers. The results can be quite stunning with a far tighter and more tuneful bass. The limitation of room eq is that it works best in only one listening position. The smaller the room and the higher the frequency that has to be equalized, the more localized the good result. Using two or more subwoofers gives a good result over a much larger listening area than with a single sub.
It should also be obvious that a larger listening room is highly beneficial. Its Schroeder frequency will be much lower. And the lower the room mode frequency, the less obtrusive it feels. Moreover, equalizing a lower fequency works over a much larger listening area than equalizing a higher frequency.
So, to be honest, good bass reproduction in a small listening room of, say, 10x14 feet is not really feasible. There is no space for bass traps, room modes are at too high frequencies (and their upper harmonics even more so), and equalizing them only works for a very small listening position. In my view, the simplest solution for small rooms is to just use little monitor speakers without too much bass output. The brain is pretty good at imagining there is bass when there really isn't much of it.

(It seems I had nothing else worthwhile to do today, so I decided to submit this novel I had originally written for publication. I haven’t proof read it yet, it’s just too damn long for me. ;>)


I can understand why no one here might fully have reason to understand the nature of the original problem given the unique perspective that I’ve been able to glimpse at it over the last several years. I don’t pretend to know everything on the topic, but I’ve been seeing it from a different angle somewhat.

First, the underlying problem, as I see it, is really bandwidth.

We may pay for our equipment and speakers and we tend to think of them as they’re spec’ed. But their frequency response measurements are usually not measured in-room. Speakers may seem to us to be something of an exception since they’re measured with a mic, but the frequency measurements are usually anechoic. IOW, those measurements most closely reflect only what the gear is capable of, in mathematical terms, under the best of conditions, frequency-wise, and those particular measurements certainly are not taking your room into account.

But, none of these specs in the real world will end up having much of a direct correlation to what we hear and what we measure in-room...the measurements are typically way off from what is spec’d. This tends to cause us audiophiles to sort of go off in all directions and start looking for anything and everything that may influence the problem. Room treatments of countless flavors, EQ/auto correction, speaker placement, vibration control, etc...

At this point anyone who may be only half seeing what I usually post about will likely be annoyed and thoroughly unsurprised when I mention Alan Maher Designs, whose electronic noise reduction devices I’ve been buying for the last 7 yrs now...to the tune of, now, about $10k. That’s $10k’s worth being thrown at a $6k system...(yes, I know, I’m certifiable, but all that is another novel I will bore you with some other time).

But, in the course of steadily improving the sound and presentation (in all regards) in-room over that time, I began to notice that, yes, the frequency response in-room was becoming more extended at both ends and all that, but what made me pay attention was that it was Flattening the overall response as well. This is one of the things AMD gear is touted to do, but until I’d seen it in action under my own roof, it hadn’t fully dawned on me yet what this actually represented.

Alan’s claim all along was that if you flatten in-room response with this approach, then you’re reducing frequency peaks and actually filling in the nulls...without EQ...just by getting rid of noise. IOW, room effects can (to a very large degree - surprisingly large) be thought of in terms of the audible effects of electrical noise. To restate it: what we’ve been audibly attributing to room effects all along (what we are hearing and measuring in our rooms) may actually be more correctly thought of as the audible impact on our systems of ordinary amounts of electrical noise (yes, everybody has it, so don’t think that it somehow won’t apply to you). To put another way: by assuming that in-room bandwidth irregularities are directly resolved by the whole room-treatments/EQ/placement concept only, and that there was no other real factor involved, then we all may have been barking up the wrong tree. Not that room modes don’t exist, they do. It’s just that the majority of what we are hearing may actually be a rather extreme exacerbation of that problem due to nothing more than electrical noise. Remove enough noise and maybe 90% of the audible effects of the room mode problem go away - IF you go far enough to take the AMD approach to a logical conclusion that is, as I have already.

But, as I became more and more convinced that this approach actually works well in the real world (through adding purchases, primarily for other sound improvements, though they all contribute to an overall improvement to the sound), I steadily realized how much less and less dependent I was becoming on EQ and room treatments...these days, a modest EQ boost of no more than 3.5 dB in the low bass (15" woofers), another, more narrow, 3.5 dB boost at 13 kHz and no EQ at all in between...despite OB speakers that are 5ft out from the front wall in an open room...yet I get perfectly sounding lower mids (and, with the EQ’d extremes, everything else so far). (I haven’t measured yet, but I must say that currently I’ve no audibly discernible reason to at the moment).

But, then I realized that that meant there were now a multitude of methods that could be brought to bear on the problem - room treatments, EQ, and now electrical noise reduction (like conditioning here, but without adding any sonic ill effects). But, when I considered all that, I began to see that starting from the room itself first and trying to work backward to the electronics was possibly getting the cart in front of the horse. It all seemed to make more sense to do all the electrical noise reduction first, then whatever minimal EQ was necessary and, lastly, whatever room treatments would then be called for...also, by that point, presumably minimal.

Room treatments are a great thing, but I think its application presumes that all your system ducks are in a row before you start. Change anything: your gear, placement, EQ, noise levels, and theoretically you’re potentially calling for a revised room treatment scenario...one that may even have you starting over from scratch with it if you were to do it right. But, the directional approach starting from ENR, to EQ and then to (hopefully) a less involved room treatment array might make the whole thing a little more doable than if we start with the room first and then try to head upstream.

The takeaway here is that, fundamentally, the best way to resolve the bandwidth problem is to try and resolve it Before it reaches the speakers. Once it becomes audible, then technically it becomes much harder to correct with accuracy.

Buying equipment with better specs is not the answer. The specs only show us what the gear would be capable of IF it were operating at spec. Unfortunately for anyone concerned, I’ve discovered that that is Never the case and that it can’t be without solid ENR - there’s just too much of it in our everyday lives, no matter what our circumstance. I have AMD for all that and it works really well, but, along these lines and many others with AMD performance, it did not begin to start sinking in just How well all the AMD pieces were working together until I was into it for at least a couple thousand bucks or so. But, at the $10k level, my gear is now all operating at several times above spec and the audible improvements have continued to be dramatic in every respect.

EQ is great, too, but the only problem with it is that it can be compromised from normal amounts of electrical noise, as I have discovered. The 2 biggest problem areas for it are the colorations in the lower mids within the "mud" frequencies and the grit/grain/glare in the upper "hash" frequencies. With the noise removal here, I’ve been able to squelch the effects of both, so in my case there are no more ’best compromise’ settings between tone and noise.

Hopefully at some point more audiophiles will discover the relevance of the link between electrical noise, measured in-room response and room modes...and, by discovering AMD, or something like it, will realize that something effective can be actually done about it.

Regards, John



And, yep (you guessed it), all that’s yet one more reason why I like AMD...(this was reason #4217). My apologies once again for the long post. Carry on.

Great thread! Another ++ for Room EQ Wizard, great software. I don't focus so much initially on the frequency response charts. I begin with the waterfall or spectrogram looking for base that is hanging in the room too long. Even a few 6" base traps from GIK start to help this.