Acoustic Diffusion is not an easy thing to test or describe. In fact in the academic acoustics world there is a big controversy about the best way to describe and measure diffusion: the Scattering Coefficient or the Diffusion Coefficient.
“You will know it works when you hear it,” is true but probably not a good explanation. I am going to avoid this whole dispute by publishing the actual data of all frequencies and amplitudes in my tests.
I did several rounds of acoustic testing on the Ramps and here is what I found:
The Acoustic Ramp scatters frequencies horizontally from about 300 Hz to about 4000 Hz (4kHz), but doesn’t do a lot below or above this bandwidth
The Acoustic Ramp scatters and reflect frequencies vertically from about 300 Hz all the way up to 20000 Hz (20kHz)
The Ramp does both of these things at the same time. So if you install the diffusers on their sides, it only changes which orientation scatters and which one reflects and scatters.
Why Do We Want To Scatter the Sound Anyway?
When sound strikes a hard flat surface most of the sound bounces off the wall as an echo. That echo can interfere with the sound that is headed towards the wall by cancelling out certain frequencies and emphasizing other frequencies. This effect is called comb-filtering. You just cannot hear sound accurately in a room that has comb-filtering problems. Scattering the sound breaks up echoes into many little tiny echoes diminishing their effect. The Ramp also reflects sound away from the sound source, further minimizing the effect of comb-filtering.
Reflections can also be controlled with absorption of course, but absorption-only treatments tend to yield rooms that sound dead and lifeless.
How Do you Read Those Diagrams?
Those diagrams are called “sonograms” and they were generated from data gathered using the software package ARTA. There will be a later post about the testing procedure, but for right now let’s focus on the diagrams.
The Sonogram shows Amplitude (color), Frequency (x-axis), and Directivity (y-axis). The above diagram shows what happens when you place a loudspeaker directly in front of a flat panel. The bright yellows and reds show loud levels and the blues and dark blues show quiet levels. The stripe of brightness that goes horizontally across the center of the diagram shows that sound is bouncing right back towards the speaker. At about 500 Hz the sound is bouncing at about 30 degrees and 50 degrees off-center. And from about 300 Hz and below the reflection pattern is pretty wide.
If there is pretty good diffusion at a specific frequency the diagram will show a vertical band of similar colors. Horizontal bands of similar color means that there is an intense specular reflection at a specific angle. For instance, in the above flat reflector model, there is a very strong reflection at 0 degrees, meaning that sound hitting the reflector is bouncing straight back to the loudspeaker. If the reflector had been angled at 45°, we would expect there to be an intense band at around 45° as well.
In the horizontal test of the Acoustic Ramp™ you can see that nearly all of the energy is directed away from bounding back towards the sound source because the center of the diagram shows darker blue colors.
In the vertical test of the Ramp you can see the scattering of frequencies from about 300 Hz to around 4 kHz.
There are several reasons why the Acoustic Ramp™ works better than traditional diffusion treatments, but we’d like to focus on two major improvements in this post.
The Ramp scatters acoustic energy horizontally while at the same time it reflects the energy over 4 different angles vertically.
The depth of the wells of the Ramp are continuously variable leading to increased functional bandwidth.
In order to understand why these factors are important we need to discuss a few things about diffusion in general. There are two different types of diffusers that have been thoroughly tested and evaluated in the academic and in the ‘real-life’ community: the 1-Dimensional (1D) Diffuser and the 2-Dimensional (2D) Diffuser.
1-Dimensional Diffusers
1D Diffusers are probably the most commonly used diffusers in studios and critical listening rooms and offer very predictable results. These diffusers scatter energy in a semi-circular pattern horizontally. Some examples of this type of diffuser are the following:
RealTraps Diffusor (http://www.realtraps.com/diffusor.htm) This diffuser combines the QRD (Quadratic Residue Diffusion) math from Schroeder’s work with low frequency absorption.
Primacoustic Razorblade Quadratic Diffuser (http://www.primacoustic.com/razorblade.htm) The diffuser uses a sequence of depths that doesn’t appear to be strictly a quadratic residue sequence, in that it is aperiodic and the width of the zero-depth reflectors isn’t consistent. The diffuser does offer a tremendous amount of phase-complexity and probably works very well indeed.
2-Dimensional Diffusers
2D Diffusers are most commonly referred to as Skyline diffusers after RPG Inc’s model with that name. They scatter acoustic energy in a hemispherical pattern, both horizontally and vertically. Here are some examples of 2D diffusers:
RPG Inc’s Skyline Diffuser (http://www.rpginc.com/products/skyline/index.htm) This diffuser is based on a primitive root number sequence instead of the more commonly used quadratic residue number sequence.
Auralex Acoustics’ Wave Prism (http://www.auralex.com/sustain/waveprism.asp) The Wave Prism uses a grid of dividers to separate the blocks of different heights which, according to Schroeder, offers better diffusion properties.
According to Schroeder’s research, both types of diffusers work on the same basic principles. The width of the wells (or the blocks) determines the upper frequency limit. The depth of the wells (or the blocks) determines the low frequency limit. The math behind this is as follows:
The wavelength (λ) of a frequency is equal to the speed of sound (c) divided by the frequency (f).
The low frequency limit of effectiveness is defined by frequency whose wavelength is four times the depth of the deepest well (or tallest block).
The high frequency limit of effectiveness is defined by the frequency whose wavelength is 2 times with width of the wells (or blocks).
Let’s use RPG Inc’s QRD 734 as an example. The wells of the 734 are roughly 9 inches deep and roughly 4 inches wide providing the following bandwidth:
Or roughly from 375 Hz to 1700 Hz.
In independent tests performed on RPG’s QRD, the actual bandwidth of diffusion extended to approximately double what Schroeder’s theory predicted. We also found in our testing that the upper limit was significantly higher than the formulas predicted.
The depth of the wells is another area where the Acoustic Ramp™ has a major improvement over previous diffusive treatments. The wedge-shape of the diffuser means that the depth of the wells are continuously variable, which means that the frequency response is much wider than a typical 1-D diffuser.
Additionally, the Acoustic Ramp™ varies between a depth of 12 inches to a deep of less than a half inch. The deepest part of the diffuser is usually installed in the corner where the ceiling meets the wall, taking advantage of typically unused space in the room. The diffuser tapers as it descends the wall allowing racks, furniture and other equipment to be placed against the wall without trapping space.
The angles formed by the wedge shape of the wells allow the installer to direct reflections away from the sound source. In the most common installation, the wedge shape would direct reflections down toward the floor. In alternative installations, and array of Ramps could be used to direct reflections towards the side walls. In both of these cases, directing sound energy away from the sound sources helps to significantly reduce the effects of comb filtering.
The following slideshow is from ECHO Boston (http://echoecho.us), a studio in Brighton, MA, that was designed by XIX Acoustic’s Hendrik Gideonse. The photos were shot by Townsend Colon and Art Directed by Farida Amar. Makeup was done by Leila Rivers.