Audio Lesson for Your Dyno Room
It’s no secret that one of my other hobbies is audio. Car audio, home theater, reference stereo, live music, it all ties in. I was actually more interested in car audio than engines when I first got to engineering school. My Big Brother in my fraternity was an electrical engineer who had been playing with crossovers and speaker enclosure design for years when I met him. He taught me a lot about how the physics tied in to what we hear (and feel), along with how best to start making adjustments to correct what we hear. Fast forward a few more decades and I still play with sound. Sometimes I’m making it (playing music or movies) and sometimes I’m helping mitigate it (dyno cells, listening rooms). The science behind it is the same either way.
We’ll start with my recent project: a reference quality 2-channel audio setup in my listening room at home. While the Digital Signal Processor allows us to flatten the frequency response and align the flight time of music from each driver to the listener’s head at the optimum seating position, there are some things it cannot do. The DSP can’t fix diffraction caused by early reflections from the speaker baffle. These speakers were designed with minimum surrounding area to create that unwanted interference similar to the Bowers and Wilkins 801, which were my inspiration ever since I heard a pair back in the 90’s.
But great drivers will only get you so far. The room itself can color the sound with reflections and standing waves. What you can’t see in the photo above is the foam panels on the walls directly opposite the speakers. These help tame the primary reflection off the back wall, but looking deeper at the measurements, we’ll see that something is missing. Even with careful measurements and recalibration of the DSP, the room still sounds very “lively” compared to what I know is recorded in some tracks. Time to go back and look at the measurements taken with our calibrated microphone using the Room EQ Wizard (REW) software. This is starting to feel a lot like comparing ECU datalogs and calibration files now.
Yes, that looks like a lot of noise. The truth is that moving the mic (or your head) just an inch or two can significantly change the frequency response at the new position as some reflections become either constructive or interfering. That’s why we take multiple recordings from multiple positions and average them before making decisions. With an 18′ sub and over 1kW of power, I can make all the low end bass I want, but the room seems to over-deliver around 50Hz no matter what I do. Looking at the RT60 plot shows how various frequencies decay in the room. The low end clearly likes to hang around for longer than the speakers dictate.
Digging deeper, we look at the Waterfall plot and find something interesting. I saw peaks around 50, 100, 150, 200, 300Hz. This can’t be a coincidence. These are all multiples of the same fundamental frequency (~50Hz) with diminishing intensity as the order is increased.
THIS is something our brain will recognize, and should be addressed. But why is it there? Knowing that sound waves have a predictable length based on their frequency, it was time to break out the tape measure and look at the listening room again. This room is a bit irregular, so I took a bunch of measurements and looked up what frequency each measurement would correlate to for wavelength. The results jumped right out at me:
Seeing peaks at multiples of 46-50Hz is NOT a surprise in this room. Now we’re getting somewhere. We have definitive proof that our room contributes to 50Hz (and multiples thereof) noise. We just need to treat it. Treating low frequencies largely depends upon absorbing energy as the waves approach the walls. Bass Traps can be used to great effect here. After doing lots of reading, the biggest lesson I can transfer to you about bass trap effectiveness is the equivalent of “no replacement for displacement”. Studies have been done on various designs, shapes, densities, and placements. The biggest influence on low frequency attenuation is the thickness of the element. 2″ panels do just about nothing this low, 4″ panels are marginal, and 6″ really starts to work well. This means that either a substantially thick panel on one of the walls, or a bunch of corner elements could be used to have an impact upon the waves in question here.
The easiest is to add are corner traps. We can always stack additional panels in front of these or along the walls later, too. We’ll start with these corner traps that are up to 7″ deep and put them on some of the surfaces touching our 50Hz waves. Do we expect this to cure the problem? Probably not completely, but it should make a dent. We’ll take some more measurements after they’re in and see how we did. Since this is a reference listening room, I’m fully expecting to get or make some larger 4-6″ panels for some of the offending walls too.
So how does this relate to your dyno cell?
I’ve helped many shops reduce unwanted noise in their dyno rooms. Ideally, I can get involved before the room is made to help create a room with favorable geometry. This can avoid standing waves in the first place, and has the biggest impact. After that, we look at where we can either absorb reflected sounds, or break them up to reduce their intensity. Properly done, we can have a significant impact on noise emitted from the dyno room, even if the car is incredibly loud. Reducing the impact on your hearing, or neighbor complaints is a nice benefit too.
by 
