Acoustic Reflection Patterns

We’ve been doing a lot of work recently on acoustic reflectors and diffusion surfaces for a School of Music. The challenge has been understanding whether or not what we’re doing is going to work. Our designs are analyzed by our acoustical consultant, but we can’t realistically ask them to analyze ever single iteration we dream up so we find it helpful to do our own form of analysis before sending things to them for the OK. We’ve got the visual analysis covered (does it look good or not), but recently we were looking to study the acoustic performance, in particular acoustic diffusion or how sound is scattered in a space.

Sound can be analyzed with raytracing even though it actually moves in a wave.  There are inaccuracies with taking this approach because the wave behavior isn’t being taken into consideration, but the raytracing still serves as a good starting point.  When setting up this study we start with the geometry we want the sound to bounce off of, a point source for the sound to originate from, and a series of rays that are emanating from that point.

Next the intersections between the rays and geometry are found and then the real work starts.  The normal of the face that each ray intersects must be found which is then used to find the reflection vector.  The simulation can either stop or continue on for another bounce after the reflection vector is found.  In this example we were calculating 5 bounces which was typically enough to get all the rays to reflect off the diffusive geometry.

As with any type of simulation it’s important to be able to quantify or at least visualize the results of the simulation in a way that makes it possible to make observations about what is happening.  For this study we placed a vertical dome in front of the diffusive geometry and then intersected the reflected rays with the dome to see how the reflections were being distributed.  Below is a series of examples showing how the reflection patterns change with different diffusive surfaces.

The color of the dots at the end of each ray represents the time it took for the ray to reach the dome.  That color will likely come to represents another property as we continue to refine this study but for now it is helpful to see what rays are reflecting off the diffusive mesh multiple times.  All of these studies kept the source in the same location which isn’t going to tell us the whole story, but we wanted to share some of what we’ve done so far because some of the patterns have been really surprising.  Check out the video below to see a few more of the studies.

7 Comments

  1. David Tomlinson says:

    Pretty sweet, and great to see a visualization of something so complex as sound. Thank you for posting this. I’d be curious to hear your thoughts about taking the reverse approach to the above: determine a desired resultant reflective pattern, informed by this study, and reverse-engineer a diffuser array to generate a similar set of potential scatter patterns?

  2. scrawford says:

    Hi David,

    We actually have a definition that does just that. I’ll put a post together about it in the near future.

    Scott

  3. Joanna says:

    Fascinating! I look forward to the update on the closed loop system you are working on…

Leave a Comment