About the Project"One of my motives for upgrading my sound system was to have what I considered to be an “endgame” sound system with the intention of flush-mounting. The reason this is important is because, obviously, that sound system is being embedded in a wall, more or less precisely designed to have just enough space to sink your speakers into it. The last thing you want to do is have to revise it for a new sound system. I’d been tossing the idea around for a while as a means to add more deep trapping to my room and improve the performance of my reflection free zone, and now that I had the new PMCs, it was time."
Hacienda Studio Re-design - Flush Mount
Waveguides, Waveguides, Waveguides!
Authored by Head Engineer Matt Davis
One of my motives for upgrading my sound system was to have what I considered to be an “endgame” sound system with the intention of flush-mounting. The reason this is important is because, obviously, that sound system is being embedded in a wall, more or less precisely designed to have just enough space to sink your speakers into it. The last thing you want to do is have to revise it for a new sound system. I’d been tossing the idea around for a while as a means to add more deep trapping to my room and improve the performance of my reflection free zone, and now that I had the new PMCs, it was time.
A conventional “hard” flush-mount relies on massive and rigid boundaries which compose the face of the surface the speakers are sunken into. In doing this you essentially take the speakers “out of the room” from an early reflections standpoint, allow me to explain. The Schroederfrequency is the highest frequency at which modal activity exists, beyond this frequency sound is considered directional within a room. This frequency decreases with room volume (which is a big part of why people say bigger rooms are easier to treat, smaller modal range!) and sabines of absorption. Another element at play is the Baffle Step f3, which is the lowest frequency that a sound system projects exclusively hemispherically (forward into the room), it is positively correlated to the shortest dimension of a speaker’s front baffle. Below the baffle step of a sound system and the Schroederfrequency of the room sound radiates more or less omnidirectionally from a speaker. So what happens if you force all of that low frequency energy to radiate hemispherically into the room? A big bass boost, improved low end phase coherence, and free headroom. You also can’t have a front wall bounce if your speakers are inside your front wall, which is very nice.
This was my first foray into flushmounting a sound system, so I was somewhat trepidatious about the prospect and consulted with a few friends with experience doing this. In my research I discovered that my particular sound system is alleged to be somewhat averse to conventional flush mounting on account of their transmission line design. The inherent phase incoherence created by the transmission lines tends to be exacerbated when they are projected hemispherically into a room, as opposed to radiating in free space unless very specialized and complex routes were taken to mitigate the problem. My room’s modal distribution would shift against my favor by shortening my room length, and I did not like my odds with taking an experimental approach to reconcile the potential phase coherence issues (remember, first time flush mounting!) and opted to design a soft flushmount instead.
Soft flushmount designs are an interesting creature. Instead of a massive rigid boundary creating the facing of the wall, you do the opposite. A minimal structure with an acoustically transparent facing designed to allow sound below the baffle step of the speakers to enter freely…into a ton of bass trapping. Like I said before, the advantages of hard flushmounting are early reflection suppression, a boost in bass level below the baffle step cutoff, and improved low end phase coherence. Do you get that out of a softflush design, not all of it. Obviously sound will still be radiating omnidirectionally around the speaker and will not yield the same degree of room gain, the front wall bounce suppression is not theoretically as effective as a hard flush, ditto the low end phase coherence. So what do you get? All the real estate you need for comprehensive bass trapping and the promise of a highly attenuated front wall bounce, and the improved phase coherence of effective deep trapping of sub-baffle step frequencies. My PMCs are naturally forthcoming in the subrange and have a tremendous amount of headroom, so the loss of potential room gain/headroom was a fair price to pay for a bit more certainty of results. With that and the need for the deep trapping provided in mind, I felt confident to proceed.
While consulting with my friend John Brandt, he advised me to give his waveguided bass traps a whirl. The general concept is creating a series of fixed rigid boundaries surrounded by absorptive fiberglass. These are arranged at specific angles with airgaps in between to trap incident bass frequencies with much higher efficacy, and to a much greater depth before reflecting out than standard deep fiberglass trapping. Even the least dense fiberglass becomes reflective to sound beyond a depth of about 18”. The operative mechanic of this technique is to subvert that limitation through continuous changes of impedance for incident sound, traveling through air, into a bit of fiberglass, bouncing off a boundary back through that fiberglass, into air again, more fiberglass, another boundary, and so on many many times before reaching the front wall absorption, and bouncing through that same series of impedance changes again on the way out. Being an acoustics forum dweller, I was already familiar with Tom Hidley’s famous hanging baffle traps, of which this concept was derived. The hanging baffle concept uses lower density cores (Homasote) hung from the ceiling and wrapped with fiberglass. These traps utilize an additional mechanic, namely sympathetic vibration, to further optimize trapping. John said his trapping method was more effective per cubic foot in smaller studios like mine that can’t give away 6+ feet for hanging baffle arrays, and I’m inclined to believe him, as he’s really quite clever. Being the lunatic that I am, I decided to make my waveguides about four feet deep.
But I figured why stop there, and designed a new and carefully ray traced side wall treatment through the RFZ. This new design utilizes smaller waveguides to redirect early reflections further behind the mix position. Now I had a plan together.
This was a detail oriented job, the tolerances for angles and measurements were tight, as was the timeline, so I flew the absolute legend, contractor Matt Justice to Florida to help me build this thing. Six days of 14 hour days, 500 pounds of fiberglass, six 4×8 sheets of MDF, a bunch of Guilford of Maine fabric, loads of additional lumber, an insane amount of fiberglass exposure later, and it was done. The front and front/side walls were covered in 2” OC703 prior to waveguide installation, and the fiberglass covering the 3/4” MDF waveguides was 3.5″ R13. The fabric facing panels used the thinnest lumber of the job (to minimize potential edge diffraction), and the gap between the speakers and TV was particularly nerve wracking to me and, in the panels surrounding the speakers we ripped 2x4s to the absolute minimum width that would maintain structural integrity. This lumber was then bored in the center and OC703 was introduced to trap edge diffraction.
There were a few other logistics involved in this process, one of which was an inconveniently placed air vent right in the middle of the proposed flushmount facing, which we re-routed to the back of the room with a few extra turns in the ductwork to minimize vent noise. Lastly, the unique design of my MB2s-XBD system afforded me the ability to attach my passive crossovers to the flushmount facing for easy access in the future, so I did just that.
Alright, and the results? Some failed to meet my expectations, but quite a few exceeded them immensely. The new RT60 in the bass range is nothing short of transformative, the waveguide traps worked wonders on this front and my RT60 is on its way towards nominal. A significant number of excess group delay anomalies have disappeared from the upper spectra, signifying an improvement in boundary interference issues. The waterfall plot has also become much more uniform, meaning less modal ringing is present in the room. There also has been a notable amount of room gain introduced, which means free headroom in the subrange, although it is accompanied by a dip in the vicinity of 35hz that I haven’t cracked the code on yet, but which has been amenable to a boosting shelf in the meantime. My theory there is sympathetic resonance in the structure, if I make breakthroughs I will post an update.
THIS CASE STUDY HAS BEEN UPDATED. SEE BELOW FOR MORE DETAILS.
- Date : Updated: Oct 7, 2019
- Category : Studio Design and Build
"And what does the data say? In a word, transformative. The average RT60 of the bass range has dropped from 250msec into the 130msec range, the length axial mode sequence has been greatly attenuated, the 37hz group delay and amplitude response anomaly has completely resolved, and the amplitude response of the entire modal range is now flat."
While the results of flush-mounting the PMCs were overwhelmingly positive, I felt we could still do better by implementing deep trapping from the face of the flush-mount surface to a depth of two feet, across the entire face. To optimize for gas flow resistivity, this called for a foot of R13, then a 4" airgap, followed by another foot of R13...in a very tight space with no ventilation and no existing support structure to mount the fiberglass. Fun. After a lot of strategizing, my contractor Matt Justice and I came up with a plan for retrofitting this quarter wave trap into the design without removing the facing. After a lot of swearing, contortion, and gratuitous fiberglass-on-face action we made it happen in one day, although I'm pretty sure Matt is still mad at me. And what does the data say? In a word, transformative. The average RT60 of the bass range has dropped from 250msec into the 130msec range, the length axial mode sequence has been greatly attenuated, the 37hz group delay and amplitude response anomaly has completely resolved, and the amplitude response of the entire modal range is now flat. But how does it sound? The perceptual bass response has undergone a night and day improvement, the subrange feels weightier and more focused at the same target curve. The flushmount is also giving the impression of a hemispheric projection pattern, as can be heard and seen in SPL plots, resulting in much greater efficiency in the low drivers after implementing a baffle step compensation filter in DSP. Quite possibly the tightest bass decay I've ever heard in a room so far. Bass is more directional than ever, and the decay characteristics of kick drums and basses are much more detailed. I can't speak for Justice, but this one was well worth an itchy day to me.