what is a diffuser ?

Definition:

• A diffuser is an acoustically hard and reflecting surface which,
• because of the shape and design of that surface,
• is capable of homogeneously diffusing soundwaves
• in space and time,
• within a pre-defined working range.
  
  

Following this definiton you may safely draw the conclusion that a diffuser is acoustically hard and reflective for soundwaves striking it, and consequently not absorptive. Absorption is usually not the purpose of a diffuser, rendering the role of this acoustic aid quite unique: there are countless forms of dedicated sound absorption,
yet there are only a few that can offer dedicated sound diffusion.

The main purpose of a diffuser is to realise a locally diffused sound field with maximum homogenity within the working range of the device, while preserving the acoustic energy of the sound field.

'Locally' specifically means: within one of the 1st reflection zones that are present in any music room.
The meaning of 'homogenity' will be explained as well.

A number of simple physical principles underly this definition. In fact, we already know these by practice. A description of these qualities will add some more significance to the role of the diffuser .

Fact 1:
When a soundwave strikes a flat and acoustically hard surface, it will reflect with almost identical frequency and amplitude (SPL).
The incoming angle equals the outgoing angle of reflection, which is about the only 'acoustic diffusion' that has taken place here.

The image on the right is an attempt to visualise the kind of reflection described above.
In this case the soundwave strikes perpendicular to the flat surface.
The largest part of the acoustic energy is reflected outward with the same angle it struck.

When the angle of the soundwave striking the wall would have been 45° the largest amount of acoustic energy would reflect back into the room with a 45° angle as well, like a massive beam of acoustic energy. Which is exactly what it is, in this context!

Such "high-energetic" reflections easily occur in the first reflection zones of an untreated room.
They will have a clearly detrimental effect on the spatial imaging qualities of music brought into the room by means of any given loudspeaker system or musical instrument.


TOP

Reflection and dispersion of a straight incoming soundwave on a flat and hard surface

Most of its acoustic energy is reflected into one direction.

Fact 2:
When a soundwave strikes a non-flat surface it will, depending on the 'three-dimensional pattern' on that wall, partly desintegrate into a number of smaller reflections with lower amplitude, while the remaining acoustic energy manifests itself through some high-energetic reflections that are identical to, but smaller in strength than a full flat surface would generate.
The nuance is in the indication, "a number of smaller"...

The visualisation on the right shows what happens to the indirect soundfield.

Quite often this 'Fact-2' is commercially misused to claim an incorrect simplification of the diffuserprinciple.

In this context the claim is that a simple repeating and curved (cylindrical) shape pattern completely omits the need to apply more complex exponents of QRD- and PRD-diffusion.
This claim is incorrect and probably the best proof that a person claiming this has never experienced optimised reflection control in a music room at home.

The crux is that excessive repetition of a cylindrical or otherwise simple pattern is equally wrong and even more detrimental for the indirect soundfield as the complete absence of such a pattern (the hard and flat surface from fact-1). It will create a phenomenon called "lobing", clearly visible in the image on the right.

This may lead to the remarkable phenomenon that a wall, treated (with best intents) with a much repeated and simple wave pattern, whether cylinder-shaped or otherwise, will increase the already existing problem of untamed reflections, when compared to a normal flat wall.
You may perhaps remember the "eggbox rooms" that mark the start of home acoustics in the seventies...

Because of the excessive repetition of simple 3D-patterns, unwanted reflections are only "pampered" within a very narrow frequency band while the rest of the spectrum responds as if it strikes a normal flat wall.

This single, narrow but well pampered part of the total music spectrum is not wide-banded enough - physically spoken - to do much more than produce some sort of an effect that will soon start to work against you.

An excessively repeated pattern, whether simple or complex, can never be successfully applied for the diffusion of a wideband soundfield. "Excessive" is the key here.

This is also applicable to 'real' diffusers (QRD, PRD) and for the very same reasons! There is a minimum and a maximum number of periods to apply on a given diffuser surface.
Too little and too much diffusion (repetition of a pattern) will both produce unsatisfactory results...

Reflection and dispersion of a straight incoming soundwave on an optimized curved and waved hard surface

The acoustic energy is powerfully reflected into a limited amount of directions, creating some, in itself less powerful, reflections that remain clearly audible.

As a result of this rudimentary dissipation by the wave pattern on the wall, these "less powerful reflections" now also originate from different directions!

Out of the frying pan and into the fire...

Fact 3:
When a soundwave perpendically strikes an optimised diffuser surface, it will desintegrate into many reflections of much lower amplitude.
Moreover, these weakened reflections will be dispersed into countless directions.

This fact reflects two fundamental characteristics of successful sound diffusion:
"homogeneous dispersion of a soundfield in space / and in time."

• Dispersion in space implies that the reflection based coherent directions in the soundfield (coherence as shown in fact-1), with its individual reflections, are replaced by spatial 'randomness' (incoherence).
  Soundwaves are homogenously dispersed into 'countless' directions, as opposed to the images already shown with the facts 1 & 2. Thus the direction-sensitive and spatial aspects of the whole indirect soundfield will alter.
  Dispersed reflections arrive, spatially changed, at the listening position.
  
  
• Dispersion in time implies that the runtimes of and within individual reflections are altered. This happens in conjuction with, and is a logical result of the above mentioned dispersion in space.
  As a result of these spatial changes in direction, fully new runtimes for reflections emerge, but also for parts of these reflections.
  So reflections are not only processed as a whole but are also disassembled. Parts of a reflection will end up 'here' while other parts end up 'there', causing dispersed reflections to arrive at the listener with a randomised change in runtime.
  
  

We've already noticed that any spherical or irregular shaped surface is capable of randomly scattering or breaking some of the soundwaves that strike it.

Bearing this in mind you may say that a diffuser is nothing more than a carefully calculated irregular and reflecting surface, that has to be applied by limited repetition.

Such a 'careful calculation' will result in a predictable working range of almost 4 octaves for a QRD-7 and well over 3 octaves for a PRD-7 diffuser.

This wide, connected and predictable working range sets the 'real' diffuser apart from a random irregular or a (semi-)spherical shape.
The latter simply cannot have a working range that contains 3 to 4 subsequent octaves.

Some reserve is wise here and by now you will understand why:

Quite some "irregularly shaped" panels are offered as "diffuser" by acoustic webshops while, in fact, these products will not work according to the definition of a diffuser.
Ignorance with regard to this definition enables these shops to make money selling relatively cheap products and claim them to be "diffusers".
They indeed do something with the soundfield, as you know now...

These products mainly have a pyramid shape or another simple irregular or wave pattern, such as the shallow white panels and pyramids at the ceiling in the image above.

These are sold as 'diffusers' but they're only capable of dispersing a narrowband section of the soundfield and are not really up to the requirements for successfull wideband diffusion.

As explained, they only spatially disperse a narrow band of the spectrum. After living with them for a while, you will start to judge them more and more as an effect, not as a constructive change of the entire indirect (reflected) soundfield.

The white column-panels however are PRD's (RPG Skyline) and these certainly are real diffusers. And they're not really cheap...

above: QRD-7 diffusers: homogenous diffusion
within a working range of nearly 4 octaves

Reflection and dispersion of a soundwave perpendicularly striking an optimised diffuser surface.

Spatial dispersion of acoustic energy is much smoother than with earlier surfaces.
The amplitude of the individual reflections has been minimised as well.

Many reflections also received a different runtime.

Equal dispersion, as shown above, is identical to "homogenous diffusion".

Homogenous diffusion is possible when the optimised diffuser surface consists of at least 3, rather 4, and a maximum of 8 "periods".

A correctly designed diffuser will inherently meet the two criteria generally set to define 'good sound diffusion':

• The working range is sufficiently wide, connected and predictable through calculations.
  The diffuser should operate between pre-defined lower and upper frequency limits, which constitute "the working range". This range should be wide enough to avoid "generating a sound effect" and to deliver a constructive contribution to a significant part of the indirect soundfield.
  
  
  

• The diffusion pattern within this working range should be as homogenous as possible throughout the range.
  Equal dispersion - homogenous diffusion - theoretically expands as the complexity of the pattern increases.
  A QRD-7 is the simplest exponent of Schroeder's diffuser principle. With it homogenous diffusion within a predictable working range of almost 4 octaves is possible.
  
  
  
  TOP

why diffusion ?

There are usually a couple of good reasons to apply diffusion in a serious listening environment. One 'practice driven' article is "A matter of diffusion" (still in Dutch): "Een kwestie van diffusie" on this site.

The labour intensive nature of diffuser assembly definitely determines the high price of commercial products. Have a look at the images in the right column...
The price gets even higher if more exclusive finishes are chosen. Add this to the monopoly position on the market a manufacturer can exploit to even higher price limits.
And finally: diffusers are heavy stuff. Shipping them is costly...

Apart from RPG there is, even in 2020, no manudacturer around producing serious acoustic diffusers. As we've already seen there are a lot of webshops claiming they sell diffusers that are not really diffusers.

It goes without saying that DIY can be a huge advantage in the assembly of diffusers, provided you can meet the criteria for home assembly.

The main motive to apply diffusion in a music room is reflection control. Reflection control manipulates the indirect (reflected) soundfield. This field consists of sounds that reach the listener's ears after touching one or more surfaces in the room.

When such early reflection can be successfully tamed, which is yet another synonym for "homogenous diffusion", it will reward the listener with both ambient and focussing qualities hitherto unheard within the indirect soundfield.

Reflection than turns out to fulfill a usable and constructive role in music reproduction, instead of being a continuous source of unease, harshness and smearing in space and time.

The impact of a well dispersed indirect (= reflected) soundfield includes more than ambient characteristics. Order in reflections also causes a much tighter placement and stable focus of instruments and musical events on the virtual stage in front of you.

A pattern of correctly tamed and dispersed reflections gives spatial realism and a wider scope to the sound image (the ambiance). It also introduces stability and "image sharpness" in the musical picture. And finally the amount of coloration, inevitably added to the soundfield by strong reflections, decreases or disappears. The neutral character of the room increases while, at the same time, the characteristics of the recorded space will get more manifest, at least if these are present in the recording.

In and around the listening zone acoustic diffusion should enable a homogenous soundfield, capable of decoding spatial information hidden in the recording, and do this in a way far superior to flat and hard 1st reflection zones.
This information has always been there, in the recording, but was never able to sufficiently emerge, mainly because of 'faulty' 1st reflection zones. Reflections no longer arrive at the listeners position via specific wall- or ceiling areas, but create an ambient soundfield that literally serves as a stage or carrier for the direct sound (the sound that reaches the listener without first touching some object or surface in the room).

Psychoacoustics rightly state that ease and stability in a spatial image is one of the basic conditions enabling you to listen to artificially reproduced music without experiencing listening fatigue.

Listening fatigue comes into play when our brain consistently has to fool itself, in order to undergo - at least to some degree - 'the appearance of music'.

Of course our brain is capable of processing correct spatial information without a large effort or constant misleading of the mind!

Concinving power automatically emerges
and there will be no listening fatigue at all.


TOP

profile of a beautiful but very costly commercial diffuser; these images will probably show you why such diffusers cost a lot...

above:
the dimensions of this diffuser: 100 x 200 x 20 cm
price (2017): 2300 euro
home building should be cheaper...



below:
various exponents of diy diffusers and bass management in a recording room

absorption and diffusion !

Acoustic optimisation at home is not a matter of applying either diffusion or absorption somewhere in the room. Such a consideration is actually a non-issue that may only become an issue if you yourself make it one.

Acoustic optimisation at home includes both reverberation control and reflection control.
Reverberation control requires controlled absorption. Reflection control requires controlled diffusion, yet can also be done by means of additional absorption.

Nevertheless absorption is not a primary, nor is it a very suitable tool to solve all reflection problems, although it is definitely better than doing nothing!

The basic distinction between these two is quite simple:

• absorption removes acoustic energy from the room
  
  
• diffusion alters acoustic energy in the room
  
  

Reflections can be absorbed (= removed as much as possible) when the reverberation character of a room and/or other practical considerations plead for such a method.
However, when reflections must be tamed and their acoustic energy should be preserved, a diffuser is the only tool that does this job.

Normally the acoustic optimisation of a music room should, first of all, start with widebanded reverberation control. Have a look at the dedicated article on this site HERE (still in Dutch).
In other words: it should start with absorption.
As you should not render all surfaces in the room absorptive, enough reflections will remain that need some additional treatment. This is even the case when the wideband reverberation time of the room is already brought under control.

You actually want to preserve the acoustic energy carried by the remaining reflections instead of making it disappear by applying even more absorption.
Preserving this energy means preserving vividness in the ambiance and the spatial environment a music room can offer.
Absorbing this energy will lead to a reduced ambient presentation and a smaller portion of spatial information, while the direct soundfield simultaneously becomes louder and more manifest.

All of the above will hopefully reflect the need of searching and finding an acoustic balance between the two basic qualities - a balance between direct and indirect sound, which equally implies a balance between absorption and diffusion.

Diffusers are usually positioned at a suitable location against a wall or other surface in the room. This location usually corresponds with one of the 1st reflection zones, where incoming soundwaves benefit most from dispersion in space and time. Such diffusion will remove coloration and 'reflective coherence' from the sound image, yet preserves the acoustic energy that is part of the indirect soundfield.

In a relatively small music room, as is usually the case at home, creating a balance between diffusion and absorption is one of the main objectives. Absorption regulates reverberation times and diffusion prevents early reflections and flutter-echo from introducing coloration, distortion and other artefacts that create reflective coherence.

This way diffusion enables spatial information, being constantly distilled from the soundfield and processed by the brain, to be equipped with the natural width and depth of an unlimited, large room that wants to be 'colored and filled' with the recorded space itself.

The balance between sufficient absorption and good diffusion will automatically remove the sonic impact of limiting surfaces in the room.

You may say that - in a small room - diffusion and absorption are the complementary elements in the hands of an acoustic architect. The correct balance between vividness and absorption is created by them.

absorption and diffusion are two faces of one entity called 'acoustic optimisation'.

their mutual balance totally determines correct presentation of highly valued musical criteria for reproduction quality:
spatiousness, presence, focus, ambiance, directness, convincing power, non-coloration and dynamic contrast...

above:
a fractal QRD diffuser;
a fullscale miniature QRD-7 diffuser
is embedded in each well...

below: a QRD-7 series
these were in 2005 the first prototypes of
a QRD-7 with a folded deepest well -
a so-called 'folded-well diffuser or FWD








profile of a QRD-19

building them...

The QRD-diffuser, aka "Schroeder-diffuser", is not that difficult to construct at home provided you have correctly cut wood at hand and are able to work with sufficient accuracy.
You also have to have sufficient patience and time. When these are met nothing stands in the DIY way...

When building a diffuser, regardless of type and size, the end product has to live up to 2 crucial demands:

• To avoid unwanted absorption and unwanted resonances in a diffuser it is crucial to avoid cracks and air gaps between the parts and in the wells.
  Have your material cut at a sawmill where they can and are willing to do this with sufficient accuracy!
  Avoiding unwanted resonances in a diffuser is merely a matter of structural integrity. This is the reason diffusers are fully glued during assembly. It is also a major reason commercial diffusers have such a high price tag.
  It requires much work.
  
  
• With regard to the completed endproduct (the diffuser), multiple experiences by a number of different builders made it clear that untreated mdf is not an ideal finish for a diffuser, even though the material seems hard enough.
  The reason can be found in the selective absorption characteristics of untreated mdf. Application of the required number of diffuser surfaces in a given room may generate too much unwanted midhigh and treble absorption. This may lead to two separate problems:
  
  First of all it may color the soundfield reflected by the diffusers, by holding back (absorbing) certain mid- and high frequencies. The impact of the diffusers will lack vividness.
  A second problem might occur in an already well treated room with optimised reverberation times. The added absorption generated by the untreated mdf diffusers will add unwanted absorption, which also leads to dullness in the ambient soundfield.
  You can easilty understand why this needs to be avoided. Both problems severely compromise the very reason you want to use diffusers in the first place!
  
  It is therefore crucial to apply some sort of a hard finish to your home built diffusers, which makes them even more labor intensive.
  A couple of primer and paint layers, sprayed or manually painted, would be ideal. A finish with a transparent parquet lacquer might even be harder and very beneficial too.
  
  The use of an alternative for mdf may offer possibilities to reduce and eliminate the inherent absorption. You may consider natural wood, although that would also require a (transparent) lacquer finish.
  In Holland mdf with a white lacquerfoil finish is available. This foil is meant to be a paint primer and is already very reflective by itself. You only have to apply a final layer of (color) paint or varnish to finish.
  I have used it several times and I found it worth the added price tag.
  
  
  
  
  
  
  

QRD-7 anatomy

From this point on the QRD-7 diffuser is used to clarify the anatomy of a diffuser.


A QRD diffuser is an element that can have any height you want and has a fixed width.
It is built according to a fixed pattern of vertical, diffuser-high wells of equal width.
These wells have different well-depths.

Wells are always separated by a thin separator, also called 'fin'.

As strange as it may appear at first sight, it is those fins make a diffuser into a real diffuser.
No fins, no diffuser. It is what makes them complex and labor intensive to build...

Have a look at the profile of a QRD-19 on the left.

I assume you will be able to deduce that these fins in particular make a QRD into a labor intensive product. Without them a diffuser would be quickly assembled, yet the addition of fins to the design turns it into a more complex and time consuming job.

The diffuser is built on the basis of a backwall and joined with two side segments that have been assembled separately. Suitable materials would be multiplex, mdf, hard wood, tempex, hard foam, pvc, stone.

 

Below you find an image of the anatomy of a QRD

The terms used in this image are consistently used throughout all diffuser articles on this website.

anatomy of a QRD

TOP

N-number

Below and on the right you find examples of QRD diffusers.
QRD = Quadratic Residue Diffusor



The simplest yet very effective QRD diffusers have 7 'wells'.
That's why it's called 'a QRD-7'.

One complete pattern of [N = X] wells is called 'a period'.

Any diffuser configuration on the wall or ceiling should consist of several periods.
Periodicity is required to generate homogenous diffusion.

Apart from N=7 other N-numbers exist. Have a look at the table on the far right. The word "reeks" means 'sequence'.
In this case: depth sequence.

The N-series, up to number 31, is made up of QRD diffusers with growing complexity : N=7, 11, 13, 17, 19, 23, 29, 31...

These will be diffusers in which 1 period consists of 7, 11, 13, 17,19, 23, 29 or 31 wells.



Below: a QRD-11

above: fractal DIY diffuser - QRD-7 6/7 Mod.2

a scaled QRD-7 is embedded in each well of the bigger diffuser, doubling the top end of the working range;

the 0-well has a negative depth number here (M=2)
it is deeper than the front of the fins and the diffuserfront itself

this results in a QRD-7 with a deeper low end of the working range
when compared to a 'normal' QRD-7 in which M=0.

what does the M-number mean in this table?

The M-parameter defines the (relative) depth of the (odd) 0-well. This 0-well is the left well in both diffuser images on the left.
"M" means 'modified sequence number'.

M=0 always represents the unmodified depth sequence for the N-number in the first column. For the M=0 diy diffusers preposed on this website the physical depth of the 0-well is, of course, 0 cm.

Have another look at the QRD's (drawings + photo) on the left.
A depth of 0 cm will correspond to the front of the fins and therefore with the front surface of the diffuser itself.

In a QRD-7 6/7 mod.2 (photo left) this depth is no longer 0 (front of fins) but 2, a relative, negative number on which the other depths in the sequence are based.
With a QRD-11 8/11 mod.3 this (relative) depth would be 3.

When looking at the M-sequences within a single N-number (i.e. QRD-19), you will see the order of these sequences greatly differs. So wil the modes of construction. An M=>0 diffuser usually requires more labor and has more constructive challenges in the assembly. Challenges further increase if you also want to fold the deepest well, as is the case with the most efficient version of the QRD-7, the FWD-diffusers found HERE on this site.

In the end the 0-well will also determine the diffuser depth, as well as the amount of wood you need to build it.

If you would calculate the working ranges of various M-sequences within one single N-number, relatively small differences will emerge in the lower and upper limits of those ranges.

The total working range (4 octaves in a QRD-7) will NOT change by changing the M-number.
The total working range, as a whole, will shift up- or downwards with the changing M-number.

Practically speaking:
The reasons stated above do not really justify construction of diffusers with a 0-well larger than 0, unless you want to realise the deepest possible low end limit within one single N-number. In that case you should build the diffuser according to the sequence (reeks) that will result in the lowest bottom end frequency.

symmetric or asymmetric design ?

Any QRD can be built symmetric or asymmetric.
This is because the odd 0-well can be divided in two half-width wells - one on both extremes.

In an asymmetrical QRD the odd 0-well only occurs once in the pattern, positioned on the left side of the drawings that were used in this article.
The other wells (in any QRD) are already mirrored symmetrically, as seen from the odd middle fin of the period.

To create a fully symmetric diffuser, the odd 0-well must be halved.
Both half-width wells are put symmetrically at the extremes.
Thus a symmetrical profile emerges with identical specifications.

Have a look at the drawing on the right.



There are no differences in operation and specifications between these designs.
Even the width of the completed diffuser (1 period) can be made identical.

There are at least some differences in the constructive details when building.

Although most diffusers are built symmetrically there is no technical reason to do this. However, when more diffusers are applied together, with a small space in between, it is probably visually more pleasing to look at symmetric diffusers.


The illustration below shows two rows of connected QRD-7 periods
- both represent equally correct diffuser surfaces for practical appliance -

The top row is built asymmetric. The bottom row is symmetric.

Visual differences show up only in the far left and right extremes when asymmetrical diffusers are tightly connected.

TOP

good-to-knows...

We must get a bit technical now, in order to better understand QRD design.

The N=number, the well-width, the maximum well-depth and the chosen depth-sequence (M-number) are essential designparameters for any QRD diffuser.

For a more technical and detailed article on design and designparameters you may read:
"Evaluating specifications and characteristics of QRD-Diffusers".

This article, in conjunction with the one you are reading now, has everything there is to know about simple yet very efficient acoustic diffusion at home.

• The N-number of the diffuser defines the number of wells per period.
  The higher the N-number, the wider the working range will be.
  A correctly designed QRD-7 has a maximum working range of 3,75 octaves, regardless of any choice of other design parameters.
  
  

• Well-width defines the upper frequency limit of the working range.
  Narrow wells will result in a high upper frequency limit.
  This parameter is inextricably connected to the chosen N-number: identical well-widths will result in different upper limit frequencies when a different N-number is chosen.
  
  

• The maximum well-depth defines the lower frequency limit of the working range.
  Deep wells will result in a low bottom-end frequency limit.
  This parameter is also connected to the chosen N-number:
  identical maximum well-depths will result in different lower frequency limits when a different N-number is chosen.
  
  

• The chosen depth-sequence (M-number) defines depth and order of the various wells.
  Each N-number is accompanied by various depth-sequences. Each of them will result in a slightly different working range, which always has a bandwidth of 3,75 octaves max.
  
  Some N-numbers have only 2 or 3 depth-sequences, others have 6 or 7, as you have read in the table with prime sequences.
  A designer chooses one of the available depth-sequences within a given N-number.
  
  
  
  
  

Limitations...

To put it bluntly:
One cannot combine very narrow wells (to obtain a high upper frequency limit) while at the same time make those wells very deep so you can achieve a deeper low-end limit of that working range.

It would be nice to create a very wide working range this way, but it will not work out fine...
The thing is, within those very narrow and very deep wells new unwanted reflections, resonances and coloration will originate, making the whole assembly quite useless for diffuser purposes.

Correct proportions between well-width and maximum well-depth must be resprected to get a succesful diffuser design, and therefore these proportions are a crucial design parameter.
This proportion is related to the chosen N-number.

With a QRD-7 and a PRD-7 the proportion is 1:4 maximum.

This will result in a 100% efficiency benchmark that corresponds, with a QRD-7, to a working range of 3,75 octaves.
A PRD-7 has a working range of 3,25 octaves at 100% efficiency.

For example: if you choose to apply a well-width of 5 cm, the maximum well-depth will than be 5 x 4 = 20 cm.

If 20cm wouldn't suffice to meet a planned bottom-end of the working range, you can choose a higher max. well-depth of, let's say, 30cm.
30 cm divided by 4 results in a well-width of 7,5cm which, in turn, will result in a significantly lower limit for high frequencies.
Yet in both cases we're dealing with a design of 100% efficiency. Only the working ranges shift, defined by either well-width or max. well-depth. Choosing one always has consequences for the other.

Plan - example

On the right an example of a QRD-7 plan.
Reading and interpreting it is probably no problem now, when you have read this article up to this point.

The deepest well in this example has been folded once to save space and materials, yet the specs remain the same, with or without folding.

This model is the FWD-36 diffuser, part of a family of diffusers found elsewhere on this site.
This -36 is a fine allround diffuser!





Anti-flutter strip or Diffuser?

An FWD-36 diffuser has a well-width of 36mm, making it an allround diffuser with regard to the working range.

When the well-width of a QRD-7 drops under 25mm
it shouldn't be called a diffuser anymore. A well-width of 25mm or less positions it in the category of "anti-flutter strips" of which 2 exponents are shown on the right, a bit upward.

These are not really diffusers, because their working range far exceeds the part of the frequency spectrum where the fundamentals are found (which is up to a little over 4kHz / c5 on the piano)

A well-width smaller than 25 to 30mm will position the upper frequency limit of a QRD-7 a lot higher than the top-end of the fundamentals range (a little over 4000Hz.)

Such a QRD-7 can do wonders in eliminating local flutter-echo and other high pitched reflection problems that are often a local phenomenon in a given room.
Such acoustic aids are referred to as "anti-flutter strips".

The width of the fins - 6 and 4 mm combined in this example plan - should never exceed 1/4 of the well-width. Changing one dimension in the design will always have an influence on other dimensions. Some of them can be successfully compensated, like the width of the top- and bottom plates or even the backwall.
Other dimensions, most likely the maximum well-depth, can get compromised, depending on the part(s) you plan to modify.

Better also read the technical article, proposed here earlier, before trying to alter a design.


TOP

D I Y diffusers can be very rewarding...

QRD-7 strips (HDF) to eliminate flutter-echo
or embedding in fractal diffusers

both products have identical specs

 

above: QRD-7 plan and profile (dimensions in mm)

The example plan above deals with allround QRD-7 diffuser FWD-36.

In an allround diffuser the maximum well-depth lies somewhere between 14 and 21 cm, resulting in a bottom-end frequency limit somewhere between 225 and 350 Hz, well into the lower part of the fundamentals range.

The corresponding well-width lies between 3,5 and 5,5 cm, resulting in a top-end frequency limit of 5 to 8 kHz.

There is no real advantage in the application of diffusers with a range far exceeding the upper limit of the fundamentals range (up to 4,1 kHz). 8kHz would really be the practical top-end here.
A higher top-end limit of the working range is of no advantage, and this is even more so because the low-end of the range will also shift upward, which is not a good thing either. The more the working-range corresponds to the fundamentals range, the more "allround" a diffuser will be.
As we've already seen, a very high top-end of the range renders the diffuser more and more useless as a diffuser as it more and more turns into an anti-flutter strip.



The photo below is quite interesting. It shows the only diy-diffuserkit I have ever come across on the internet, apart from the WaveWeaver-kits offered by SoundScapeS between 2010 and 2014. Commercial kits were never that attractive, or so it seems...

The price of this kit below is, still, very high (to Dutch standards), with a considerable three-digit price tag...

However, the price per square meter commercial and ready-built diffusion is even much higher: depending on the chosen finish, 1 square meter of commercial diffusion will cost between 650 and 1500 euros.

And we haven't even discussed fractal diffusion yet...

Fractale QRD-7

I've mentioned so-called "fractal diffusers"several times now.
Have a look at the photos on the right and the drawing below.

A fractal diffuser is essentially a normal QRD. Instead of a each well having a flat well-bottom, a correctly scaled mini-QRD is embedded and serves as the bottom of each well in a fractal diffuser.

Such a "mini-QRD" is usually cut from massive HDF profile, as you can see on the right.

Fractal diffusion is a bit like the russian matrouchka dolls: each doll has a slighly smaller replica embeddid inside, and so on...

Fractal diffusion might be called: 2-way or even 3-way diffusion. The top-end of the larger diffuser should about equal the bottom-end of the embedded element.


The drawing below illustrates multiway fractal diffusion.





A fractal diffuser will have a far extended top-end frequency limit compared to an identical QRD without embedded QRD-strips.


The Diffractal by RPG, shown on the right, homogenously diffuses the soundfield up to 25kHz. because of the embedded QRD-strip.


The bigger diffuser will have a working range that starts as low as 250Hz.


This implies that a "2-way diffuser" is already capable of covering almost the entire audiospectrum!


You may expect that a real 3-way diffuser, as shown on the left, would truly cover the entire spectrum, as the biggest of the 3 diffusers starts working around 50Hz.


The working range of a fractal 2-way diffuser covers almost twice the width of a single QRD: 3,75 versus 6,5 octaves!


But as you know by now, even a single QRD with a well chosen working range will do a very good job, yet at a fraction of the price of fractal diffusion.

above:
allround QRD-7 model FWD-36 and bottom-end diffuser FWD-72 (well-width 36 and 72 mm)
in which the deepest wells have been folded once.

This 'trick' allows you to realise the lowest possible bottom-end combined with minimum diffuser depth and use of materials.

above: a commercial fractal QRD-7 diffuser, the RPG "Diffractal"

size: 61 x 61 cm

in each of the 7 wells a small and scaled "mini-QRD" is embedded.

price, ex works (no transport): ± 500 dollars for untreated mdf finish
other finishes add to the price

 

Folded-Well Diffusers (FWD)

As of August 14, 2017 you may visit the "FWD-diffuserpage" to find a series of 9 detailed plans of QRD-diffusers with a folded well.

In a folded-well diffuser (FWD) the two deepest wells are folded once, and in the two biggest models they're folded twice.

Have a look at the FWD-concept on the right.

As opposed to the classical QRD's mostly seen on this page so far, the FWD concept results in a reduction in diffuser depth of 33 to 40%

The use of construction materials is reduced accordingly too.

And all of this without compromising the specifications of the diffuser at all !





D. I. Y.

As you may know, I was able to make many people happy for over 10 years with WaveWeaver diffusers. These models are no longer available here, as the rights have been transferred to a third party.

To meet the needs of the home builder, I have re-designed and expanded the concept with a range of 9 similar models, all having 100% efficiency within the maximum working range.

The well-width of the 9 models varies between 36 and 98 mm and actually cover the earlier WaveWeaver models with minimal changes to the design.

Nowadays the name of the model reflects well-width.
Model FWD-58, for instance, has a well-width of 58 mm.

These 9 models are: FWD-36 / 40 / 46 / 52 / 58 / 64 / 72 / 84 and -98.
Of course the working ranges of adjacent models have a significant overlap.
This allows you to meet just about every diffuser-need.

Plans, parts lists, a universal photo-story as well as a detailed video-manual on the assembly of these diffusers are all at your disposal for free.

FWD-40 is the earlier WaveWeaver-1 diffuser
FWD-46 is the earlier WW-1½
FWD-52 is the earlier WW-2
FWD-64 is the earlier WW-3


TOP

price tags

Commercial diffusers are, by nature, quite costly. This is partly due to the labor intensive assembly and to the monopolistic position of the few diffuser manufacturers around.

When I built and distributed the WaveWeaver diffusers (2005 to 2015) there were 4 different QRD-7 models available. I calculated a sales price of twice the ex-works price. Nothing extreme or unreal, given the success in that decennium.
I have to admit this generated more than enough honest income. I might even have asked some more, and it would also be paid for...

The reality of a monopolist is quite different.
A manufacturer like RPG is capable of ending up with a smaller ex-works price than I ever could. Nevertheless their sales prices for untreated mdf QRD's have always been 65% higher than mine.

Budgetwise DIY is a no-brainer with diffusion. The price of the end-product is made up of 4/5 labor costs and 1/5 material costs. All you must do is reserve time for a project. You might start off with building 4 allround diffusers (120-140 cm high), paint them and then experiment.
At a price tag of 100-150 euros this is worth the experiment.

Depending on the chosen materials (natural wood, for instance), the price tag of a home-built diffuser can easily increase. It's a (visual) choice, as the material itself has a secondary influence when comparing natural wood and mdf.

Most home builders tend to chose mdf to paint it, during or after assembly or both, in a color of their choice. Natural wood finish, though beautiful, means a high price tag. Spray painting requires a trained painter and will also have its price tag.

The weight of a single QRD-7 may quickly become a thing to reckon with.
Up to a diffuser height of 150 cm a well-width of up to 70 mm will end up in a diffuser that still can be handled quite well.

Diffusers with higher N-numbers or the addition of more periods on one backwall (photo on the right) quickly add weight. They should be built on location.

The home built diffuser on the right weighs 90 kilograms and combines 4 periods on 1 backwall.
Dimensions are 250 x 106 x 25 cm - quite impossible to handle for 1 person.









additional information

In a serious dedicated listening room diffusion is, I think, indispensable for controlling and taming reflections and indirect sound (early reflections), while at the same time ambient vividness remains intact.

On this website you can find additional information on diffusion.
Below an overview of these sources.
When available in English they will be mentioned on the update page

• Plans for classic QRD's:
  downloadpage
  Choose download nr. 7
  
  
• Plans and manual FWD diffusers
  downloadpage
  Choose download nr. 6
  It might be wise to go to the FWD webpage first.
  This download can also be found there, along with other information you shouldn't miss.
  
  
• Webpage on FWD-diffusers
  (English):
  here
  
  
• Information on design parameters, specications and characteristics of diffusers
  (English):
  here
  
  
• Gebruikservaringen met QRD-diffusers:
  (Dutch):
  here
  
  
• Photo gallery "compromiseless music room", with much d.i.y. diffusion:
  here
  
  
• Photo gallery "diffusion in general":
  here
  
  
• Photo gallery "FWD-diffusers":
  here
  
  
• Video-manuals for d.i.y. / assembly of FWD diffusers:
  here

I wish you good luck building and enjoy the experience !

More info, feedback to share or special questions? Please send me an email: readscapes@xs4all.nl


Toine Dingemans, May 1, 2020.


TOP

above: four joined periods of a diy-diffuser
mdf, 250 x 106 x 25 cm / 90 kg

below: QRD-13 under construction