I agree that human hearing curves need to be a factor when analysing room response. This doesn't negate the act of measurement, but affects the act of interpretation. Even in its simplest linear understanding, room measurements can be useful to qantifying very large room modes.
Without appropriate correlation measurements to human hearing you cannot use measurements to predict problems, yet this is what many people do: they measure, look at the graph and think they absolutely need room treatment. Measurements are useful once you have detected a problem while listening to music.
I am glad that you agree and confirm that bigger room are better.
I forgot to mention that concepts like the Schroeder-frequency are based on the assumption that the sound field is diffuse, which it is not:
Meyer, „Definition and diffusion in rooms“, J. of the Acoustical Society of America 1854, vol. 26, no. 5, p.630
Gover et al., “Measurements of directional properties of reverberant sound fields in rooms using a spherical microphone array”, J. of Acoust. Soc. of America 2004, vol. 116, no. 4, pt.1, p.2138
Therefore, the actual transition frequency is higher than what is computed using Schroeder’s formula:
Baskind et al., “Sound power radiated by sources in diffuse field”, Audio Eng. Soc. preprint 5146
…..what practical simple advice and summary would you advise on speaker placement and room effects.
Speaker placement depends on many parameters, so there is no straightforward advice.
Allison effect: small speakers without good bass extension can be placed closer to the walls/corners than large speakers that go down to 20 Hz. Large speakers can be placed close to walls/corners provided they have in-built equalization. If not, out-board equalization is needed to avoid bass boom.
Early reflections: reflections coming within 2-3 ms after the direct sound produce high interaural cross-correlation and are less preferred than later reflections, so if placement to side walls is required, absorption may be beneficial.
Ando, “Subjective preference in relation to objective parameters of music sound fields with a single echo, J. of the Acoustical Society of America 1977, vol. 62, no.6, p.1463
Room modes: since common placement rules are based on circumstances you generally do not encounter in domestic situations, they cannot guarantee optimum results, so simply don’t rely on them. You may use them as starting point, and it may turn out that that position is the best in your room, but equally well you may ask your spouse where she’d like to have the speakers placed in the living room, maybe that that position is the best in that room. If you have a dedicated listening room, you may use one of the many placement rules or just place speakers and listening chair somewhere and look how it sounds.
In rooms with hard reflective walls (brick ‘n mortar, concrete) those placement rules may give better results than in rooms with flexing walls (plasterboard on studs). Since all room modes have a maximum in corners, corner placement will excite all modes equally loud, but you’d have to use equalization.
Furthermore, in order to become annoying room modes have to be excited first and then they have to be perceived. Whatever placement method is used, play music (not test signals) and listen while being seated. Move loudspeakers and listening chair if necessary until it sounds ok to you. In our living room I’ve placed speakers and listening sofa where the wife told me to. When playing test tones the modes are clearly audible but in 7 years I have found only 3 tracks which actually do excite them.
Thanks for some great links and info Klaus, they are of great value and very interesting to me.
If you want to read more, I have all the articles of that list I posted the link to, so feel free to ask for a copy.
One, I clipped a couple of paragraphs out of the original articles and I'm sure I did them a disservice in this, you should read the entire text of the links I posted.
Well, I’ve read both links and it won’t surprise you that I have some more comments:
>The following comments apply to truly phase coherent speakers, the majority of which tend to have simple (or no) crossovers.<
It’s perfectly possible to build truly phase coherent speakers with complex crossovers. The answer is FIR-filter. The result looks like this:
> If the positioning causes standing waves or "suck-outs" you can bet the whole tonality of the presentation will be off…<
Any position of a loudspeaker in a room will generate standing waves!! A standing wave is the result of the superposition of an incident wave with its reflection from a room boundary. Happens all the time and at all frequencies.
> In this position [speakers at the two center points of an ellipsoid touching the walls of the room, best listening position 1 to 3 feet from the rear wall.] the sound from the speakers reaches the ears before any reflections coming from the side walls resulting in better soundstaging and an unaltered perception of the speakers tonal balance.<
The direct sound comes first, yes, but the reflections do come and will be merged with the direct sound (precedence effect), the sum will contribute to sound stage and tonal balance. If soundstage and tonal balance are altered by the reflections is not known. I looked at the literature relevant for this issue and while there are indications that some imaging parameter are affected by reflections (blur (Kuhl); phantom image width (Ringlstetter); phantom image location (Linkwitz), there is no thorough and detailed research on this. I've prepared a literature overview, if interested, drop me a mail.
Kuhl, „Effect of a Loudspeaker Radiated Diffuse Sound on the Hearing Event” (in German) Acustica 1978, vol. 40, no.3, p.182
Ringlstetter, „Investigations on the directivity index of loudspeakers „ (in German), Fortschritte der Akustik, DAGA ’96, 22nd Annual Conference on Acoustics (German Acoustical Society), Bonn 1996
Linkwitz, "Room Reflections Misunderstood", Audio Engineering Society preprint 7162
> Imagine the situation of being in a somewhat noisy public place and conversing with the person next to you.<
This is known as “cocktail party effect” and I’m not sure that it is relevant in the present context (direct sound – early reflections). http://www.media.mit.edu/speech/papers/ ... effect.pdf
> Our brains do this automatically all the time to, for example, filter out the annoying natural resonance of a room to facilitate speech…<
It has been shown that early reflections actually improve speech intelligibility:http://scitation.aip.org/getabs/servlet ... yes&ref=no
> The point is that, by placing the speakers and/or listening chair at an even division of the room, you will get natural bass reinforcement from the room.<
When I look at the two first axial modes, the 1st order mode has pressure minimum at room center (i.e. an even division), the 2nd order mode has pressure maximum at room center. At any even room division I will have pressure minima of some modes and pressure maxima of other modes. How this results in “natural bass reinforcement” is a mystery to me.
Letitroll wrote: Unless the signal is modulated as in musical signals, then ILD is a factor up to 4,000hz. This is noted in the article you linked.
I think the following passage nicely sums it up:
> To summarize the matter of binaural differences, the physiology of the binaural system is sensitive to amplitude cues from ILDs at any frequency, but for incident plane waves, ILD cues exist physically only for frequencies above about 500 Hz. They become large and reliable for frequencies above 3000 Hz, making ILD cues most effective at high frequencies. In contrast, the binaural physiology is capable of using phase information from ITD cues only at low frequencies, below about 1500 Hz. For a sine tone of intermediate frequency, such as 2000 Hz, neither cue works well. As a result, human localization ability tends to be poor for signals in this frequency region.<
Both ILD and ITD are hence used for source localization, so the statement “The way we locate sonic events in space is by the brain measuring the time delay of the sound between the two ears” is wrong.
Klaus R. wrote:The former, caused by mounting hardware on the speakers’ front baffle and by cabinet edge diffraction, may disturb summing localization, the latter are merged with the direct sound and are said to have negative effects on timbre (comb filtering) and sound stage/imaging. In neither cases there is evidence that negative effects actually do exist.
This would be in direct opposition to many successful speaker designers who go to great lengths to reduce early reflections. And in opposition to my anecdotal experience in speaker design and modification. But you of course are entitled to your opinion and have every right to state it.
Like all reflections also these very early reflections cause comb filters. To the best of my knowledge there is no research available on the effect of these reflections and hence no evidence, other than anecdotal, that these reflections are a problem. One obviously can always choose to play safe and design cabinets without very these early reflections.
You are quite correct here, the Haas effect is included in the group of theories called in general the Precedence effect. The article should have been more precise. From my reading of it there is no exact hard and fast rule that applies in all conditions, and the effect varies in intensity and delay times depending strongly on environment and type of signal.
The upper limit (or echo threshold) of the time window of the effect depends on signal type. What is relevant for the context of music listening in small rooms is that first reflections generally come within the 80 ms time window where the effect is operational. On rare occasions like hitting the rim of the drum with the stick the time window is possibly closed before the reflection arrives so that this reflection might be perceived as echo. So far I haven't noticed anything resembling that.
From my reading of the entire article and from the diagrams posted, as well as my own experience, it is very important to be as absolutely close to the rear wall as possible, within inches. And if there is any separation it should be filled with highly absorptive material.
Our listening sofa is actually placed against the rear wall so distance between ears and wall is about 20 cm, unless I decide to slide down into a more relaxed position:-) I tried absorbers I had at hand behind my head and could not hear a difference.
Thanks again for taking the time to post so comprehensively Klaus, I'm very grateful for your interest. You have provided me with many hours of engaging reading for the next few weeks.
Back in 2007 I came across Floyd Toole’s paper “Loudspeakers and rooms for sound reproduction – a scientific review”, J. of the Audio Engineering Society, p.451. Some of his conclusions are exactly opposite to what is considered as acquired truth and wisdom in the field of small room acoustics, so I went and read the literature he cites and many more. And yes, there are lots of myths, misconceptions, half-truths out there in audiophile circles.