Room Treatment 101

[Keith_W]

Since there have been more threads asking about room treatment, I thought it would help if I wrote a short article to explain. 

 

PART 1: THE SCIENCE OF REFLECTIONS

 

First, a quick primer on how we hear. When we place speakers in a room and sit down to listen to music, we hear the direct sound of the loudspeakers and reflections from the room. The nature of these reflections is very important, but also very controversial. In general, there are people who believe that "the soundstage is in the recording". These types of people will want more absorption, and use controlled directivity monopoles. This approach creates a "they are here" kind of sound. There are also people who believe that "the room creates the soundstage". These people will want less absorption, more diffusion, use dipoles or omnis to create even more reflections, and this results in a "you are there" sound - in other words, the room creates the ambience. 

 

There is FAR from universal agreement about which approach is "better". All I can say is that either approach, when taken to extremes, is bad. If you put an omni in a small, reflective room, it will be bad. If you put a narrow radiating speaker into a psedo-anechoic chamber, it will be bad. The key is to understand where the middle ground is. I firmly believe that personal taste plays a big part in this. 

 

These are the important aspects of reflections: how early, how loud, from where, and how long.

 

How Early, and How Loud? 

Reflections that arrive within less than 20ms of the direct sound are within the Haas fusion zone and are perceived as part of the direct sound. If it is more than -15dB in relation to the direct sound, it has the effect of smearing the sound, reducing clarity. If it is between 20ms to 150ms (depending on frequency), it is perceived as ambience and spaciousness. If it is > 150ms, it is perceived as a separate event - an echo. 

 

"How early" can be estimated by measuring the distance between the loudspeaker and the first reflection points (don't forget the ceiling and floor in addition to the side walls and rear walls!) and calculating it using the speed of sound. The formula is t = d/c * 1000  where t is time in milliseconds, d is distance in meters, and c is speed of sound (343m/s). For example, a 5m distance will create a reflection that is 14.5ms delayed with respect to the direct sound. As a rule of thumb, every 1m adds 3ms of delay. 

 

 

It is far better to measure the timing of reflections. The relevant measurement is the energy-time curve. The image above shows how to read the ETC. We can see that in my room, the main reflection (1) that arrives in the first 20ms is more than -15dB attenuated with respect to the direct sound, which has been normalised to 0dB. The second reflection is -25dB to the direct sound. 

 

From Where? 

Note that the ETC does not tell you "from where". Psychoacoustic research into the perceptual effects of the direction of reflections is surprisingly lacking. It is generally agreed that reflections that arrive from the side, if within reasonable limits, are beneficial (Toole and Olive 1997). Reflections that arrive from the front, and especially from the rear, are deleterious because they correlate with the direct sound (Imamura 2013) although this was a small study and no other studies corroborate it. Reflections from below are required because if they are absent, it will sound unnatural. 

 

Research into diffusion is somewhat lacking. A number of studies now suggest that diffusers create "mini reflections" and the sum of these "mini reflections" is perceptually the same as a single large reflection, suggesting that diffusers are a waste of time. Again, few studies, not much evidence. 

 

How Long? 

The last aspect of reflections to be discussed is "how long" - also known as reverberation or ringing. What is reverberation? All wavelengths form room modes and standing waves. Short wavelengths form thousands of room modes which are so close together that they overlap and form a reverberant field. As wavelengths get longer, the room modes start to separate out and they become more apparent, so they no longer form reverberant fields. You may get a reverberant bass field in a basketball stadium or concert hall, but certainly not in a typical domestic listening room. We need to bear this in mind when we look at the relevant measurement, which is the RT60, T30, or T20. 

 

The RT60 stands for "time for a reverberant field to decay by 60dB". This has little relevance in our application because (1) the noise floor is typically about 40dB, meaning you have to measure at ear splitting levels to obtain a 60dB decay, and (2) reverberant fields do not form at long wavelengths as already discussed. Fortunately, sound decay is linear. So we measure the T20 (time for sound to decay by 20dB) or T30 (decay time of 30dB) and extrapolate it to 60dB by simple multiplication. The T20 or T30 given by your software program can be used interchangeably with the RT60. Note that the reverberation requirement is ignored so the T20/T30 needs to be interpreted with caution! A single spike in the T20/T30 at low frequencies IS PROBABLY A ROOM MODE and may not actually be prolonged decay time! 

 

 

RT60's have a target. The target depends on room volume and application. Here you can see Acourate's RT60 display which shows the upper and lower RT60 tolerance for room volume and different standards. 

 

In general, the RT60 target should be 250ms - 500ms depending on taste. A lower RT60 creates a more "dead" sounding room, and there is evidence that it paradoxically reduces clarity if it is too low. Symptoms include having to turn the volume so high to obtain clarity that you have to shout at someone else to be heard. A higher RT60 creates a more "lively" sounding room, but there is a high risk that if the speaker does not have controlled directivity, spectral distortion will result. This can cause speakers to sound too bright or too thick depending on where the spectral distortion occurs. When done right, music can be heard clearly at low volumes and you will still be able to hear normal speech without having to raise your voice. 

 

TAKE-AWAYS

Thus the research suggests that the goals should be: 

 

- reflections within 20ms of the direct sound need to be -15dB or more to avoid smearing the sound, 

- reflections arriving between 20-150ms are beneficial if you want to create a "you are there" experience. 

- late arriving reflections are bad. Fortunately our listening rooms are too small to create loud late reflections. 

- rear reflections are especially bad. 

- both over-treating a room and under-treating a room are bad. There is an optimum in between. 

 

Thus concludes Part 1 of this exploration. 

Edited November 14, 2024 by Keith_W
Fixed typos and errors.

[Keith_W]

PART 2: DIFFUSERS 

In my opinion, diffusers are a waste of time for most rooms. This is for the following reasons: 

 

1. Toole says that the perceptual effect of a series of mini-reflections from a diffuser is the same as a large reflection, 

2. Diffusers need to be very deep if long wavelengths are to be effectively diffused, and 

3. Most normal room furnishings (shelves, furniture, curtains or blinds, general life clutter, etc) already provide effective diffusion at higher frequencies. 

 

The exception is an utterly bare bones room that nobody wants to live in. 

 

Diffusers need to be prohibitively large if they are to work at low frequencies. The depth needs to be 1/8 of the wavelength. A 100Hz frequency has a wavelength of 3.43m, meaning that the diffuser needs to be 43cm deep. This means most commercially available diffusers are ugly room decoration only, or a flex to other audiophiles. Shallow diffusers are products that you do not need, especially if you have shelves or storage along the walls. If you really need to have one, the best type to use is a Quadratic Residue Diffuser (QRD). You can use QRDude to help design your diffuser. 

 

 

PART 3: ABSORBERS 

There are two types of absorbers: pressure absorbers, and velocity absorbers. 

 

3.1. Pressure Absorbers

First, pressure absorbers. There are two types - membrane absorbers, and Helmholtz resonators. Both types of absorbers are more effective for low frequencies, have a narrow band, and need to be tuned to target the frequency in question. They both need to be placed where the pressure of sound is maximum to for maximum effectiveness - preferably the corners of the room, or if not, the walls. Dry walls and glass windows are effectively untuned membrane absorbers, so most listening rooms already have a degree of bass absorption, though it is narrow band and can't be tuned. 

 

Pressure absorbers can target bass ringing, but can not phase or timing issues between subwoofers and mains. The biggest problem is how difficult they are to tune. Membrane absorbers are tuned by changing the tension of the membrane and the absorbent material behind the membrane, and by deploying more absorbers. Helmholtz resonators are tuned by changing the surface area of the inlet, the length of the throat, and the volume of the enclosure. There are some variable Helmholtz absorbers on the market, but you need a lot of them. 

 

For this reason, DSP is the superior tool for tackling low frequency bass problems. DSP ticks all the boxes - finer control of amplitude, able to correct phase issues, and it can even tackle ringing. Dirac ART and Trinnov Waveforming are multi-subwoofer schemes that cancel ringing, but there are DIY options such as a DBA, and to a lesser extent, a VBA. 

3.2. Velocity Absorbers 

These work by increasing resistance to air particle movement and converting sound to heat through friction. The most common type is acoustic foam. Unlike pressure absorbers, velocity absorbers are broadband and absorb a wide range of frequencies. They are also more suitable for tackling high frequency anomalies and almost useless for low frequencies unless one is willing to put up with excessive room intrusion. The important variables are: 

 

- thickness of the absorber: the thicker it is, the longer the wavelength that can be absorbed. As a rule of thumb, the absorber needs to be 1/8 the wavelength of the lowest frequency. Its effectiveness drops below that. 

- density of the material: the more dense the material, the more absorption will occur, but up to a limit. If the density becomes too high, it will reflect sound. Denser material is better at absorbing low frequencies. 

- shape of the material. Some acoustic foam has ridges to increase its effective thickness and direct sound to be absorbed. 

- surface area. The more % of a room's surface area that is covered, the more absorption will be provided. 

- spacing. If the foam is spaced from the wall, it lowers the effective frequency. But this is not the same as "free thickness" because it is not accompanied by more density since we now have air occupying the gap. 

 

The absorption spectrum is specified by the NRC or Noise Reduction Coefficient or Absorption Coefficient. An NRC of 1 indicates 100% absorption (an acoustic black hole), and NRC of 0 is 100% reflection. IDEALLY, companies that are serious will publish the NRC at various frequencies, like this: 

 

 

Sadly, most acoustic foam companies gamble on the ignorance of consumers and publish inadequate data with the assumption that consumers do not know what to look for, or how to interpret the data if it was presented. At most, some will say something like "NRC 1.0 (200Hz)". 

 

Even if the NRC is published, bear in mind that measurement of the NRC is challenging and prone to errors. The standard required is ASTM C423, and requires a special reverberation chamber, humidity and temperature control, control of background noise, proper mounting of the panel to a fixture, specific microphone positioning, and proper test tones. If the chamber does not form a reverberant field at some frequencies, it might make the product look as if it performs better than it actually does (for obvious reasons, companies do not mind this). This is why you might see an NRC of greater than 1, where absorption exceeds 100%. Most average sellers of acoustic foam products do not have access to an ASTM C423 compliant test chamber. 

 

Some room furnishings, e.g. carpet, cushions, curtains, and tapestries will absorb sound. These are usually thin and low density, meaning that they only affect shorter wavelengths and do not provide much attenuation. This is a good thing, because gentle attenuation is usually enough. 

 

The effect of foam absorbers is profound especially if you have a lot of them. As you can see from the curve above, absorption across the audible spectrum is not uniform. These things have a real potential to colour your sound, selectively absorbing some frequencies more, and others less. What is worse, nearly all of them leave bass frequencies alone and only attenuate higher frequencies resulting in a lifeless top end and boomy bass when deployed carelessly. 

 

Because foam absorbers have so much potential to change the sound and the multiple pitfalls, selection and placement needs to be done carefully and with acoustic measurements to avoid spectral distortion. If you think you can buy a random well reviewed brand name absorber and tune it by ear, you are wrong! At the very least, you need a microphone and you need to know how to interpret acoustic measurements. 

 

Fortunately, most rooms reach the ideal RT60 target (albeit, the higher side of the target) with normal room furnishings alone so acoustic treatment is not needed for most people. 

 

At this point, many people on SNA will recommend a professional to measure and install acoustic products. This has its own pitfalls. They usually have a direct financial interest in selling you acoustic products, and they are usually overzealous. I have been to listening rooms that have been "professionally treated" but are actually overtreated, unpleasant spaces to be in. They take a room that measures and sounds OK and turn it into an overtreated space with spectral distortion. Some of them know that DSP is a better tool for low frequencies, yet they don't recommend DSP. Probably because they don't understand it and because there is irrational resistance from some consumers. And for the consumers who want to use DSP, it is so difficult that it turns many people off. 

 

My recommendation is to be an enlightened consumer. The best thing to do is to obtain a microphone and take your own measurements. Visit other hobbyists and do acoustic measurements of their rooms so that you know your preference. "RT60 = 500ms" sounds abstract because you don't know what it sounds like. If you engage a professional, ask them what their target RT60 is going to be, and what their recommendation is to tackle bass problems because these are especially challenging. 

 

Disclosure: I do not sell acoustic services, nor am I involved in any way in this industry. 

Edited November 14, 2024 by Keith_W

[Keith_W]

PART 4: PUTTING IT ALL TOGETHER 

 

I have made this point several times in the past: your room is divided into different zones based on room volume and reverberation time T20/T30/RT60. 

 

 

The Schroder Frequency defines these zones. It can be calculated with this formula: 

 

 

Below the Schroder Frequency (Fs), wavelengths are too long to form reverberant fields and form room modes instead. The important thing to realise is that as long as the speaker produces bass, it is the room that dictates the frequency response. 

 

As wavelengths become shorter, the behaviour of sound transitions from waves to rays. This is the transition zone, between Fs and 4Fs. Above 4Fs, sound behaves as rays - the diffuse field. In this zone, it is the speaker that dictates the frequency response. 

 

The other relevant zone is the pressure zone, which where a half wavelength is longer than the longest dimension of the room. Modes can not be formed, so the room is pressured instead. This is usually pretty low - for example, a 30Hz sound has a half wavelength of 5.7m. This is not to be confused with the pressure zone adjacent to walls - where reflections (which are in phase with the incident wave) cause an SPL build-up of up to +6dB which extends to 1/4 wavelength from the wall. This is why pressure absorbers work best when placed against walls. But anyway, I digress. 

 

What is relevant is that > 4Fs and <Fs should be thought of and managed differently, since the physics of sound is different. Everything is affected - from how we take and interpret measurements, to our strategies to deal with problems. 

 

Below Schroder Frequency - the Modal Zone

There are broadly two ways we can tackle peaks and dips that are found in every room - room treatment and DSP. Neither option is particularly easy.

 

There are some people (notably Ethan Winer) who advocate for a room treatment only approach. As we have seen, pressure absorbers need to be specifically tuned to tackle problem frequencies and this is not easy. How do you know how much tension to apply to a membrane absorber? The performance of Helmholtz resonators can be predicted with a Helmholtz Resonator Calculator but note that the amount of attenuation can not be predicted since it is room dependent. Velocity absorbers need to be unbelievably thick, and even then the attenuation is not even. Accurate bass control will never be the outcome of this approach. 

 

By far the best outcome will be achieved with DSP and multiple subwoofers. This really ticks all the boxes - it is not as physically intrusive as room treatment, it is far more predictable, and it can be tuned to way below audible limits. There are many different schemes and strategies for multiple subwoofer deployment, even some that do not use DSP like Earl Geddes multi-subwoofer method. The disadvantages are obvious: cost of multiple subwoofers, having to reconfigure your system for DSP, added complexity, and the learning curve. 

 

Above the Schroder Frequency 

The first step to listen. If you like what you hear, then chances are it's probably OK. Toole made the point that we become acclimatised to our listening rooms. Over time, repeated listening makes this our preference. I like the sound of my listening room - to me it sounds spacious, tonally correct, and clear. 

 

Do not forget that everything you put in your listening room is also "room treatment". I made the point earlier that the RT60 target can be achieved with normal room furnishing, albeit at the higher end of the target. 

 

If, after all this, you still have a problem, then the solution of choice is room treatment and not DSP.

 

This is because DSP can not deal with short wavelength reflections. The only reason DSP is suitable for long wavelengths is because microphones measure only one point in space. A long wavelength means that that particular measurement is likely correct over a larger area (like how much your head moves when you are listening). I typically correct for an area of 50cm to give leeway for some head movement during listening. The typical head is about 15-18cm wide, so I account for 3 head widths of movement. A long wavelength is many times the dimension of my head - a 20Hz wave is 17m long. On the other hand, a 20kHz wavelength is 17mm. This means that DSP would be correcting for a single extremely specific point in space. It does not represent reality! 

 

DSP can be used for speaker correction. For upper frequency room correction, DSP should be applied very broadly. And note that all that it does is change the tone - it does not change where the reflections come from, or how long they ring. Neither can DSP lower the noise floor of the room. 

 

Room treatment can only attenuate reflections, it can not change the delay. The delay is set by the distance the reflections have to travel - it is a function of room size and choice of speaker placement, toe in/out, and speaker type (monopole vs. dipoles or omnis). The advantage is that it can selectively attenuate the reflections from some directions and leave others alone. 

 

So where should we place room treatment? We need to find out where the problem is coming from. It is best to follow these steps: 

 

1. Take a measurement of L and R speakers individually and look at the energy time curve. 

2. Look for peaks arriving within the first 20ms which are more than -15dB to the direct sound. 

3. Note the timing of the peak and calculate the distance from the delay. Let's say the delay is 9ms, this corresponds to a distance of about 3m. This means the reflection travelled 3m longer than the direct sound. 

4. Get out a tape measure. Measure the distance between listening position and speakers. Then measure the distance between the listening position and all the first reflection points including the ceiling and double it. That will be your culprit. 

 

 

Edited November 15, 2024 by Keith_W

[parrasaw]

These explanations are brilliant Keith!

Thank you.

[JkSpinner]

Great work Keith.  You have helped many of us on our journeys.

[Keith_W]

You're welcome guys. I have added Part 4. 

[basscleaner]

Thanks for so detailed repetition of well-known acoustic theory. However allow me to correct some statements to up the discussion. Talking about first reflections role, it must not be overlooked the aim of acoustical room treatment. When we talk about recording studio, it means, that we have to listen recorded music material maximum close to how it was recorded initially. Than, we must damp first reflections, because they paint the timbre and change SPL differently for some parts of sound spectra. These first reflections are our enemy in this case, don't you? So, if we have our room as control room, than we need to use more absorbent materials close to transforming our room to "echoless" chamber. For many of us it is unacceptable way, because we are forced to spend a lot of time into harmful media, which damages our hearing. That's why there is other way - to agree with some changes, but to correct them further by some methods. The same appears, if to consider this room as room for acoustical comfort. Here I agree with Keith, that the decision will depends on personal taste. It easy to illustrate by stereo scene example. The stereo scene width to a large extent depends of first reflections control. Do, please, a simple experiment: seat at listening position with to big ceramic plates in your hands and turn on only stereo front speakers with control music stereo signal. Choose directions of first reflections such a manner, that sound of right speaker (like a light beam) first reflection from the right plate to be oriented at right ear by a little turning and remember such a position. After that the same do for left ear. Evaluate the wide of scene, timbre brightness with this position, without the position, some changes of it. After that try to do the same, but with cross orientation, i.e. right speaker reflection to left ear and analogue for left speaker. Try to see the difference between two cases and to evaluate a stereo scene changings. So, if you will sense good changes, you need to get the same by use absorbing and reflecting panels, hanging to walls at necessary positions with corresponding areas.

Next I disagree, is dividing sound zone by some effects, concerning to low frequencies. I know for sure, that not only modes determine this range. For your information, it depends with a room volume. For instance, we have competition between modes and comb filtering effects for volumes less than 50 m3. Many of acousticians neglect in vain this! 

As concerning to room treatment acoustical technology, in general, any acoustical room treatment must begin with goal of it as a base, than low frequencies control (including Room Acoustical Dimensions search as a mandatory option or justification for their uselessness), then with or not the accordance with RAD, - soundproofing things, then first reflections control, then reverberation balance between absorption and reflecting for mid and high frequencies ranges.

[Keith_W]

Thank you basscleaner. I hope things are going well for you over there. 

You bring up some very good points. In fact the same points that Ethan Winer makes in his book and his webpage. I did mention that "more reflections! less reflections!" is controversial, and in fact there is disagreement between Winer ("less reflections") and Toole ("more reflections").

 

With regards to comb filtering, Winer says that we are so accustomed to comb filtering that we accept it as part of the listening experience. 

For those who do not know, comb filtering occurs due to periodic phase cancellation when two identical signals arrive with a time delay between them. Comb filtering creates a metallic or hollow coloration depending on the frequency range affected. Here is a good example on Youtube: 

 

 

 

I created the above simulation by taking a sweep and making a copy. I delayed the copy by 15ms, then summed it with the original sweep. 

 

Acoustic comb filtering (sound mixing in the air) does not sound as dramatic as this example where sound is mixed in a DAW. This is for two reasons: 

 

1. The amplitude of the reflection should be lower than the amplitude of the direct sound. I mentioned this a couple of times in my discussion of the ETC, and I did say that if the reflection is louder than the target of -15dB it should be attenuated by room treatment. 

2. Comb filtering is time critical, and therefore dependent on the length of the reflection path. Very short delays result in comb filtering at inaudible frequencies, whilst very long delays (>20ms) fall outside the Haas fusion zone. 20ms has a reflection path of 6.5m which is probably achievable in larger listening rooms. But no doubt a lot of comb filtering still occurs. 

 

You brought something up I forgot to mention - sound proofing. Many of us notice that our systems sound better at night. This has nothing to do with "clean power", it has more to do with lower ambient noise. This is not so easy to tackle because ambient noise depends on our suburb, proximity to traffic, house construction, etc. But there are some things we can do, like finding all air gaps and plugging them. The other benefit of plugging air gaps is lower heating bills in winter and air conditioning bills in summer. 

 

I hope I did not come across in a manner that definitively said that one approach is more correct than the other. I have my own biases and I guess it is inevitable that some of it creeps into my writing. 

[Keith_W]

BTW someone PM'ed me for help on another forum. So I will share his Energy-Time Curve. 

 

 

In short, this is very bad. 

 

1st reflection: arrives at 1.45ms, -15dB. This is 0.52m distance. 
2nd reflection: arrives at 2.55ms, -5dB. This is 0.86m distance. 
3rd reflection: arrives at 4.73ms, -6.3dB. This is 1.62m distance. 

 

The presence of multiple early and loud reflections tells me that this is a small listening room. Comb filtering will surely be the result. 

[basscleaner]

On 15/11/2024 at 3:18 PM, Keith_W said:

BTW someone PM'ed me for help on another forum. So I will share his Energy-Time Curve. 

 

 

In short, this is very bad. 

 

1st reflection: arrives at 1.45ms, -15dB. This is 0.52m distance. 
2nd reflection: arrives at 2.55ms, -5dB. This is 0.86m distance. 
3rd reflection: arrives at 4.73ms, -6.3dB. This is 1.62m distance. 

 

The presence of multiple early and loud reflections tells me that this is a small listening room. Comb filtering will surely be the result. 

Thank you, Keith. This is quite interesting. However let me point out, that I payed attention on low frequency range only. For this range Comb Filtering has some difference from mid and high, due to much less angle directivity. Multiple interference into small space leads to emergence of peaks and dips picture, with strong changes depending from location in particular (and other factors). It's easy to simulate by FEM, for instance, or even REW. That's why  there is competition between modes and CF as a room volume decreases, while direct sound value may be even less, than SPL at some points of listening. To my opinion, there must be the critical room volume value, at lower values of which modal distribution doesn't have the first role.

[Keith_W]

2 hours ago, basscleaner said:

To my opinion, there must be the critical room volume value, at lower values of which modal distribution doesn't have the first role.

 

Yes, it's called the pressure zone, where half a wavelength is equal to the longest dimension of the room. Modes and comb filtering are no longer possible when wavelengths are this long in comparison to room dimension. 

[Sean Perth]

Hi all 

must say this topic goes way above my head however I recently purchased from the classifieds a minidsp and after setting it up have a heap of wavey lines that make absolutely no sense to me 

wondering if anyone can interpret them into English for me  

thanks 

sean

[basscleaner]

To help you'd better describe your task, room and goals. And pictures with wave lines too.

[Sean Perth]

Hi 

so my room is around 4m x 3.8m 

I do have some acoustic panels at reflection points and rear wall 

no other treatments 

I had read something on using minidsp to correct any issues with a room so thought I would try it 

I am happy with how my system sounds before using dsp so this was really to see if any changes were going to be noticeable 

so below are pics that make no sense to me are the measurements from my dsp software 

I haven’t enclosed all measurements but about half 

[Keith_W]

On another forum, I wrote an extensive post on how to take measurements and ask for help. SNA does not allow me to link to that post directly, and I am not even sure if SNA allows upload of .mdat or .zip files. I think it might be time for me to write another post on SNA on how to take measurements, because I am 99% sure that every beginner screws it up. The guys on that other forum routinely screw it up, and they are far more scientifically minded than this community. 

[Sean Perth]

Hi

i have no doubt that I probably didn’t do it correctly happy to be guided in how to correctly well at least try to anyway 

 

sean 

[basscleaner]

It seems to be useful, anyway.

[Sean Perth]

13 minutes ago, basscleaner said:

It seems to be useful, anyway.

Can you expand on that please 

[basscleaner]

3 hours ago, Sean Perth said:

Hi 

so my room is around 4m x 3.8m 

I do have some acoustic panels at reflection points and rear wall 

no other treatments 

I had read something on using minidsp to correct any issues with a room so thought I would try it 

I am happy with how my system sounds before using dsp so this was really to see if any changes were going to be noticeable 

so below are pics that make no sense to me are the measurements from my dsp software 

I haven’t enclosed all measurements but about half 

 

It's hard to take this into consideration with so vague data. To understand your room reaction better to know construction (including masonry, if you know) elements - finishing and significant dimensions (and height too), because, in particular, they determine low frequencies behavior. What is your system position into the room volume, what furniture elements are there? Then you need to measure the noise level and how it distributed along frequency scale. If you have noisemetering device, try to find the power point for low frequency range (maybe more than one).

However you may drop all this out, if you like your acoustics and music to your opinion sounds good. Pay attention to the fact, that narrow peaks and dips at your pictures you may not even notice during listening.

[basscleaner]

Look at previous post, please

[playdough]

Informative reading, thanks  to the OP.. Good job.

Photo is of a lounge with 4 x 21" bass system, after bass trapping installation.

 

 Fundamental is still a little wonky, however sounds "high and tight" with no real noticeable pumping or colouration at the listening possi.

[DSharp]

1 hour ago, playdough said:

Informative reading, thanks  to the OP.. Good job.

Photo is of a lounge with 4 x 21" bass system, after bass trapping installation.

 

 Fundamental is still a little wonky, however sounds "high and tight" with no real noticeable pumping or colouration at the listening possi.

 

Nice! Amazingly flat under 130Hz, all the way down to below 10Hz too.

That is not easy to achieve.

[Keith_W]

Sorry to criticise your graph @playdough but it looks as if you did not use an SPL meter when you took that waterfall measurement. If you look at the long "finger" at 100Hz, you will see that it is 55Hz. It is likely noise, because it does not decay. Noise at 55dB at 100Hz is easily audible and it would cause me to reject the measurement. Also, the waterfall terminates at 300ms and the bass is still decaying. Suggest you extend the measurement out to 1000ms (1 second). You will easily be able to see the noise floor. 

 

 

I am helping someone else on another forum at the moment and I happen to have his measurement loaded on my PC. You can see that I extended the waterfall out to 1 second and you can easily spot the noise floor. And you can tell from how "loud" the noise is that this guy did not calibrate his measurement with an SPL meter either. 

 

 

This is what noise typically looks like. The volume was muted for both measurements and the MMM recorder was turned on. The red curve has the mic in position with no movement, and the green measurement is the added noise picked up by the mic when you start moving it. If you look at the red curve, you see that noise typically has a rising bass response. This makes sense because high frequencies are more easily attenuated than low freqs. 

 

So whenever I see a waterfall graph, I ask myself whether I am looking at noise or decay. And if it is in the bass region, whether I am looking at a room mode. For that graph to be meaningful, I suggest you extend the window to 1000ms and adjust the gain on the graph (in the SPL&Phase tab, right click and choose "align SPL") until the "noise" is at 40dB. Here is the same measurement as above, but with a more realistic noise floor: 

 

 

Also, when did you move to Tasmania? 

 

 

[playdough]

3 minutes ago, Keith_W said:

Sorry to criticise your graph @playdough but it looks as if you did not use an SPL meter when you took that waterfall measurement. If you look at the long "finger" at 100Hz, you will see that it is 55Hz. It is likely noise, because it does not decay.

Hi Keith

Constructive criticism is always welcome, I'm fairly new to these types of measurement in REW, learning and enjoy the banter.

Correct about the fingers of noise, the area is a kitchen/lounge and has a fridge and computer fan. Computer fan I can only hear it in the dead of night, but the fridge, 100Hz@55dB  and the other peak at 100Hz, yea, can hear it now actually. Yes, it's a reject reading however it's still a reading  

 

You have a relevant point about actual SPL to mic actual reading.  and performing a noise floor measurement.

I've done voltage at speaker and some verification with an older measurement standard, an Audio Control SA 3050A, with calibrated mic.

Overall fundamental SPL accuracy would not be far out of actual calibration (but not perfect) , however taking on board your advice for next set of measurements. I've some Azura horns and B&C phase linear coaxials to test next, that'll be fun.

The crossover was changed from first order to third order LR beyond those measurements, which was just after the 2 pairs of 21's were moved in and substantial bass trapping treatments installed.

(should do more measurements although I listen to music as well !)

 

Tassie, yea, building a home in Hobart, have been working on the block for 3 months. A competent hut has been built, no power supply yet but that's coming. Hoping to base new  speaker testing there. Otherwise it's a long story,,,,,

Cheers

Matt

 

 

 

[playdough]

3 hours ago, DSharp said:

 

Nice! Amazingly flat under 130Hz, all the way down to below 10Hz too.

That is not easy to achieve.

Cheers Darren

The big surprise was the 24vdc 2 x 10w amplifier on a 50w SMPS power brick running the speakers at the time running at 4 ohms.,,,,,,,,

Compromise for that type of response is physical size. 2 pairs of 21's in 650 litre reflex enclosures/DSP. the drivers rarely excursion beyond a couple of mm for 100+dB at strikingly low distortion figures. Uncoloured and tight actual response if not peaky around 20Hz. 10 to 300Hz is it's normal service set up,  crossover to a 300Hz, 2 way point source elliptical horn. My best work yet.