[ Home | Staff & Contacts | DIY & Tweaks | Listening tests | HiFi Playground | Music & Books ]

Room Acoustics - Part I

What you read is mostly irrelevant tosh.
and what should you do about it?

[Italian version]

Product: The space around your loudspeakers and your ears
Price: your comfort at home and your enjoyment of music there
Ranter: Mark Wheeler - TNT UK
Typed: Spring 2009

Most of what you read about hifi room acoustics in the audio comics and online is complete bunkum. Rubbish. Not worth the time spent reading it let alone the time and money spent following the 'advice' offered.
"He's off again," assert plebs, stage left, "he's off on another rant disagreeing with established custom and practice!"
Fear not fair reader, replieth ye olde scribe, it is not established custom and practice that I challenge, but some of the misapplication of theory to practice, which I abhor.

There are almost as many dubious theories and daft suggestions written about domestic audio room acoustics as there are about cables. Most of what is written on the field of domestic audio is based on research into acoustics designed and funded to find solutions to commercial premises, and that is on those rare occasions when it is based on any research or academically tested theory. These commercial premises are concert halls, recording studios, post-production suites, conference centres and other places where there is a single purpose in mind. That purpose is reinforced by construction in proportion to that intended purpose in both scale and materials; the concert hall may have to be large enough to seat 800 or 4000 and still achieve and acceptable sound from the stage performers for each of those ticket buying seat occupants, while providing fast enough entry and egress for them all. The post-production suite might only have to accommodate less than a dozen, but in an environment where they can work effectively for long hours and evaluate sounds that will be replayed (and hence heard by paying public) in numerous diverse environments.

The domestic audio room usually has to fit a room carcass already present in the house, and also serve as a room to sit and watch TV, or sit and converse in groups of family and friends, and the audio room might also serve as a room to dine. The former requirement to fit a pre-existing shell dictates the dimensions of the room. The existing structure often also dictates the materials of the floor, wall and ceiling, and the piercings of these materials for doors and windows, dictating both their size and position.

In Western Europe and North America (USA & Canada) the types of construction of houses and apartments varies dramatically in the same town, let alone between continents and other parts of the world. Let us imagine an audio guru, whom we will name Guru San out of respect for his wisdom; a sage whose wisdom we may read in a comic or webpage. Guru San may live in a room that is fundamentally different in construction techniques from our own. The webpage or magazine might include some sketch plans or a photograph they may look superficially similar to ours in those little low-res images with a few dimensions recorded. The differences may be such that the two rooms actually present opposing acoustic problems. In the interests of brevity, clarity and charisma, the author Guru San might not have mentioned that his room is built with a solid stone floor, masonry walls, a lath & lime-plaster ceiling with tiny windows high in the walls. Guru San's 4x3.5x2.5m room will have a completely different acoustic from his reader's identically dimensioned 4x3.5x2.5m room, because the reader's room has a suspended wooden floor, stud partition walls and ceiling and full height glass sliding doors covering the whole of one 3.5m wall. Addressing the audio needs of each of these rooms will be very different, even if they are identically furnished (right down to the pastel shades) and housing identical audio systems and record collections. Hence when reading an author who has obviously expended considerable effort fine-tuning their audio rooms and trying both DIY and commercial acoustic solutions every reader should remember that these results did not happen in their own rooms even if they share an abiding love of the music of Jimi Hendrix with the author.

Your humble scribe has completed just 3 dedicated audio rooms in freehold owned detached houses (hence with the freedom to rebuild or alter the fabric of the buildings) and with each project learned to trust received wisdom less and to trust ears more. I've also advised, contributed to or set up systems in several more and learned more than I've offered with each situation.

An enthusiast emails the following to his favourite audio guru, Guru San

  1. The dimensions of his room including size, position and depth of window and door openings
  2. Usual positions of his furniture
  3. Reverberation time tested from his left and right loudspeakers measured at the listening position using a nice piece of freeware on his pc
  4. His present system, including speakers and their position
  5. Types of music he prefers
"What room treatments," asks the enthusiast, "will improve my bass and stereo image?"

What will Guru San recommend?
Guru San is obliged by courtesy to recommend some room treatments because that is what the enthusiast asks. If Guru San fails to recommend something the enthusiast will be disappointed; we all wish to offer to people what they say will make them happy (unless we're afflicted with antisocial or psychopathic personality disorders of course), so we all tend to do this, even people as wise as Guru San. However, Guru San has not been given enough information to suggest too much with confidence. In his near infinite wisdom, Guru San proposes some form of bass traps in the corners nearest the loudspeakers.
"Why?" demand angry plebs, feeling short-changed, "Why does Guru San propose bass traps in the corners if our old fool of a scribe says there's insufficient information to assess the problem, let alone start suggesting solutions?"
Because, replieth ye old scribe patiently, whether the bass problem is too much or too little or whether it's a matter of definition, in a rectangular room the nodes and antinodes built up from standing waves are addressed well enough by corner placed bass absorber-diffuser arrangements with much less domestic disharmony than treatments placed mid-room or on the ceiling. It is a safe suggestion that is unlikely to do much harm. Equally it is very unlikely that it is the optimum suggestion that the guru might have made if he'd descended from his mountaintop and inspected the room for himself.

Knocking out some early reflections might also help the 'stereo image' from the enthusiasts point of view. However, this enthusiast fails to mention what he means by this term stereo image. If he (and it usually is a he who worries about things like stereo image) means pin-point accuracy of individual sound-sources in the sound stage then it is a fools errand. Such phenomena are the predictable creation of two channel audio heard by two channel listeners. Real audio (like and orchestra, a street scene or a band with individual instrument amps) do not create this needlepoint tableau strung between two sources. So pursuing such a holy grail is the knight-errant's quest. However, if our enthusiast means by his term stereo image, believable scaled consistently placed virtual 'sources' where instruments and their players might be imagined to be positioned, the guru will know that early reflections (especially those from small objects that therefore only reflect significant quantities of higher frequencies) will invite the listeners brain to decode the aural information with a distorted projection of source: drums so far behind cymbals that Animal from the Muppets would require a couple of metres sewing onto his arms; the triangle in front of the orchestra rather than in the back row near the tymps. Anything that absorbs or diffuses the otherwise integer reflections local to the loudspeakers will help, knows the guru.

If the guru had the chance to visit the room he might notice:

  1. Modern high thermal insulation lightweight construction of walls and ceiling
  2. Timber suspended floor on springy joists and heavily carpeted
  3. Argon filled triple glazed windows
  4. Speakers spiked directly to this floor through the carpet
  5. Very soft furnishings including many scatter cushions
  6. Lightweight curtains
  7. Radiator behind the speakers
  8. CRT TV between speakers instead of the flat LCD TV on a stand he'd imagined from the pictures
  9. Frequency dependent reverberation time
"The case is altered" suggests the guru

The reader's room is likely to be very inert acoustically. The reader's room structure will absorb low frequencies early on every surface (from the suspended floor, through the glazing, to the low-mass thermally insulated walls and ceiling). The carpet and comfy soft furnishings will absorb middle and high frequencies very effectively too. However, the fixings and spacing of battens and noggins supporting the wall materials (let us assume 12mm gypsum board backed by 25mm insulation foam) will affect what frequencies are absorbed and what frequencies excite resonances as the gypsum mass resonates against the foam's elasticity and batten mass. What might superficially seem like a very inert room may have very specific frequency dependent problems.

Ironically, despite what you might have read elsewhere, the large glazed area may not have to be a problem. Glass below 800mm in the UK must be toughened glass by law and this does tend to ring like a bell, but in this instance it may be less of a problem if the frames are both substantial and wooden, the sealed units well taped and held by flexible silicone and the argon fill imposes an unequal load on each side of the glass as it moves due to its different density. Glass mounted above 800mm may be even less of a problem. However, it is just as likely that the full height glazing comprises old aluminium frames or modern UPVC covered aluminium frames with large unsupported areas of minimum thickness toughened glass, which might resonate with a bloom almost as loud as the source. It is not glass or glazing that's the problem, it is how it is used.

The guru is likely to have struggled to tame standing waves, reflections, flutter echoes and lumpy bass in his heavy masonry room with parallel walls and tiny windows. He may have learned much on his journey to audio nirvana in that room, but the lessons may well be of no use in the reader's room. Even next door neighbours in a modern estate of identical houses will have different sounding rooms because they install different furnishings in different arrangements; most notably in terms of curtain size and weight, pictures on walls and floor coverings.

[nice room north speaker end]

The room above:

  1. Central section 4.9m x 3.9m x 2.2m maximum plus opening for stairs through ceiling
  2. Open tread stairs in very solid piranha pine
  3. 2m x 1.6m openings at each end into smaller rooms adding 3.6m to total length making clean low E possible
  4. Heavy curtains over openings with 150mm steps up at each
  5. All floors 150mm concrete slab with ceramic tiles over
  6. Wood framed windows with small double glazed lights
  7. Lime plaster over stone walls with wavy surface
  8. Pine 'bead & butt' boards glued over 12mm birch ply below 950mm wainscot with unequal spaced reclaimed hardwood frames
  9. Lath & plaster ceiling between 100mm beams
This room has no completely evenly spaced parallel facing walls

That room was a surprising success. It broke many of the rules laid down by audio writers and yet it worked really well. However, it was one of those rooms where loudspeaker placement was critical. 5cm too far back would collapse the soundstage due to early reflections (the LEDR test is an invaluable tool to quickly identify this problem without endless playing of whole recordings with a good acoustic after every speaker position adjustment). The loudspeaker position had to be on a line parallel to the end wall, the alternatives parallel to the other 3 walls were irredeemably unpleasant to hear. The loudspeakers could not be equidistant from the side walls to avoid serious bass cancellation down the length of the room.

The experience of that room was that despite careful choice of construction and treatment from the planning stage, designed to make the best acoustic use of the space available, on installing the system and testing alternatives, the situation could be improved. The original hypothesis had been that the bets result would be to orientate the room diagonally. This would have enabled each speaker to be positioned in a similar environment (equal distances from nearby boundaries and early reflections) and listeners to be placed on an optimal triangle's apex. No arrangement could be CAD modelled thoroughly because of the large openings into other spaces, but a diagonal arrangement had modelled well for the central space. However, it did not work in the real world.

[nice room south listener end]

The audio spur had been wired to suit the intended diagonal arrangement, in the alcove formed between the stone fireplace and the corner. The lime plaster walls were uneven enough to minimise standing waves. The timber panelling below the wainscot was fairly inert to the knuckle rap test (the planks being backed by birch-ply and insulation) and 50mm within the plaster, reducing further the number of coincident distances. Against all logic the best speaker position was aligned along the beam visible in the top picture, but with front baffles entirely before the near face of the beam.

The open tread stairs immediately behind the left speaker might be expected to cause early reflections and a collapse of the soundstage. This was tested by covering them with heavy absorbant blankets and polyester quilt filling, to virtually no effect compared with exactly the same mass and area of blanket 1.5m behind the speakers. There was no difference with either music recorded in a convincing acoustic, or with the LEDR test. However, moving the speakers back towards the wall by 20cm, so that the front baffles were directly below the Lower surface of the beam caused a soundstage collapse in all 3 dimensions.

The solid cast concrete floor is paved with Platts 11mm ceramic 'quarry' tiles.
This is another contradiction to the dictats of audio journalism. Heavy carpet on heavy underlay seems to be the universally accepted wisdom. Again this seems to be borne of thinking that runs like this:
Enthusiast, "I need to minimise reflections and standing waves in my listening room"
Plebs, "You'll need to stick loads of sound absorption on your floor, walls and ceiling then!"
Enthusiast, "I can't do that, it'll look too weird as a sitting room and my partner won't like it and my friends will think I'm a sad geek"
Plebs, "Well you've got to get that absorption in somewhere..."
Enthusiast, "I know, I'll put it all on the floor, so every second or third reflection simply will not happen. And my partner will be impressed by my willingness to lash out serious money on attractive soft furnishings rather than more overpriced wire."

The plebs have believed all they have read about home hifi acoustics and our enthusiast has probably also read articles condensed and oversimplified from acoustics text books that seem to imply that it does not really matter where one absorbs the reflections as long as one does it properly. In terms of total reverberation time in a rectangular empty space this is at best an oversimplification; text books tend often to compare the total area & depth of absorption treatment in a given room volume and plot these against reverb time at a particular frequency; this is to minimise the experimental variables of course, not to imply that position is unimportant. Just as it is far easier to control loudspeaker standing waves with central filling than wall covering, but far easier to execute wall covering than central filling, pragmatic solutions are in order in our sitting room.

Despite reams of written words to the contrary, I have never experienced a domestic listening room that worked well that was not equally comfortable to sit and talk in with the audio system silent. A typical British Edwardian ground floor room, with high ceilings (typically over 2.5m), parallel walls, often almost square except for deep alcoves either side of a fireplace, is often an uncomfortable room in which to hold a conversation. This is because of frequency selective problems throughout the audio spectrum. There are standing waves, flutter echoes, curious comb-filter and boom effects caused by the alcoves; talkers in such rooms often find themselves applying their own feedback to modify their speech patterns to minimise room effects which further impedes intelligibility. I speak from bitter experience trying to conduct family therapy sessions in such rooms. I speak from equally bitter experience trying to make audio systems work in such rooms.

Equally, the over furnished, heavily insulated, low ceiling modern sitting room (often 'through-lounge' in Britain) saps all the rich timbre from speaking voices as much as loudspeakers. In such rooms people often compensate by over emphasis and exaggerated vocal resonance (however unconsciously) and similarly their occupants often choose larger-than-life audio systems in an attempt to overcome the room's inertia.

A room that is comfortable to sit in and talk for hours for the sheer joy of human exchange is a room that starts with an advantage in audio terms. This is likely to be a room with dimensions that are in proportion so that no one dimension seems out of scale with the others. No dimension is likely to have an integer relationship with another dimension (a room 5m x 5m x2.5m being the opposite of this). Soft furnishings are likely to be evenly distributed and pictures and taller furniture are likely to break up reflections in the top 1m.

Most importantly, the British domestic living room usually has slightly more absorption in it than would be ideal for audio. In Britain large heavily stuffed sofas are currently popular, as are extra scatter cushions, heavy carpet on underlay on suspended wooden floors and heavy floor length curtains. Advice to British enthusiasts should bear this in mind. The untypical room described above, featured a futon sofa, a rug 3mx3m (50mm thick knotted wool) directly in front of the loudspeakers, a small table (1m x 1m) behind the right speaker, two easy chairs and the audio support furniture. The curtains over the openings to the other ground floor areas did improve the sound when closed. However, any more absorption caused the sound to lose life and the room to feel uncomfortably dead; Extra floor length curtains were tried in various quantities and with any more than those covering the openings, the sound deteriorated. Tubular diffusers were tried in various places and these made absolutely no audible difference, probably because the uneven walls and beamed ceiling created sufficient diffusion already.

Irrelevant theories like Live End Dead End (LEDE) do not belong in discussions on domestic audio any more than caterpillar tracks on family cars; the relationships in this comparison are equally relevant and equally irrelevant. Special acoustic products for domestic audio settings do have a place where the limitations of a room are fixed and intolerable.


Domestic audio is to be enjoyed in domestic settings. A room that is comfortable to sit in and hold an easily intelligible conversation is more likely to make a good audio room than one designed using theories developed for concert halls, lecture theatres and recording studios. There is often adequate acoustic absorption provided by seating and soft furnishing in many European styles and the audio enthusiast should pay special attention to the arrangement of these elements to exploit their full potential. In a future piece we'll be hearing how a really BAD room may be improved with minimal expense but some woodworking skills.

Music enjoyed during this editorial

All of it! My latest audio room is complete and almost fully furnished.
I have slung together a quick system of Garrard 401, SME 3012 series 1 (pre-bias), Audio Technica AT95E, Yaqim phono pre-amp, modified Assemblage SET300B (various NOS valves/tubes), modified Hammer Dynamics speakers, to listen to:

  • The Rise of Dancehall volume 1, reggae compilation
  • The Rise of Dancehall volume 2, another reggae compilation
  • DJSS presents: Jazz & Bass Sessions III
  • LTJ Bukem: Logical Progression
  • Greatful Dead: Axomoxa on vinyl

Go to Room Acoustics - Part II - How to tame standing waves and get better bass

© Copyright 2009 Mark Wheeler - www.tnt-audio.com

[ Home | Staff & Contacts | DIY & Tweaks | Listening tests | HiFi Playground | Music & Books ]