John Darko
When listening to music played through loudspeakers in a room, we hear two types of sound: 1) the sound that travels directly from the loudspeakers to the listening position and 2) the sound that bounces off the walls, floor and ceiling to arrive at the listening position distorted and late. This second type of sound contributes to a room’s reverberation. A loudspeaker’s dispersion pattern can influence how much reverb we hear but generally speaking and for most loudspeakers, reflected sound is a function of the room more than the loudspeakers sitting within it.
When I first took the keys to my Lisbon apartment in January 2023, the amount of reverb heard in the main lounge/listening room was – to be blunt – ‘disgusting’. Immediately, things became interesting. The uninformed onlookers pointed to the tiled floor as the main culprit. “Get a rug down, man”. I did exactly that – and added a sofa to boot – to see the room’s reverb score (its RT60) barely move from one full second. The RT60 registered approximately one second without the rug and sofa in place. And it stayed at approximately one second once they were added.
In other words, it took frequencies above 300Hz coming out of the Zu Soul 6 approximately one second to drop below the room’s noise floor (typically -60db) whether or not the sofa and rug covered large sections of the tiled floor. You can catch up on that story and a more detailed explanation of reverb here.
Seasoned audiophiles hip to the realities of room acoustics will know that the cause of reverb doesn’t just lie with the floor (tiled or not) but all of the room’s bare surfaces. Because bare surfaces are reflective surfaces. It’s why I had my Lisbon lounge’s walls and ceiling covered with Vicoustic absorbers and diffusers in 2023 — and again a few weeks ago. The absorption panels reduce reverb and the diffusers help it remain consistent across frequencies above 300Hz. Side note: below 300Hz the room’s behaviour moves from reverberant to resonant to introduce bass peaks and nulls.
As of right now, my Lisbon lounge room has an RT60 of around 250ms — one-quarter of what it was two years ago. And yet, removing the 2m x 3m rug that sits between the couch and the loudspeakers barely changes the room’s reverb. I told this to a hi-fi friend who immediately called “bullshit” in disbelief. “That tiled floor has to do something,” he opined. I explained that it was doing something but that he was looking at the floor when he should be looking at the rug, which does almost nothing to reduce the tiled floor’s reflective behaviour.
The average rug absorbs very little below 1kHz. It might take a little zing out of the room but that’s about it. Any improvements would be negligible; and we would see a similarly negligible change if we were to hang that same rug on a wall or glue it to the ceiling.
My friend still wasn’t convinced. I showed him the 2023 measurements taken in the empty Lisbon room with the Zu speakers and his only response was, “Yeah, but the room was empty when you added the sofa and the rug. Of course they’re not going to do anything when the room reverb is already that high”. He implied (or did I infer?) that the rug would do more in a less reverberant room.
Today, I got to test that theory. I placed a room measurement microphone at the listening position in my acoustically treated room in Berlin. I connected that mic to a MacBook Pro running Room EQ Wizard (REW) that was in turn connected to the hi-fi system. REW generates test tones that play out the loudspeakers for the microphone to read and return to REW for graphic display. This is where we see my Berlin lounge room’s RT60 hover just above 200ms as we journey north from 300Hz.
More pertinent to today’s discussion, we can see this on two RT60 graphs: one taken with a 2m x 3m rug laid out on the floor between the couch and speakers (as it normally is) and the second taken with the rug rolled off to one side to leave a large portion of the room’s wooden floor fully exposed to sound waves.
Here are those two RT60 graphs:
Whilst we can conclude that the rug does something for reverb at 1.2kHz, that something is negligible. Similarly, this side-by-side shows how the rug worsens the reverb time ever-so-slightly at 400Hz. For context, here’s what two rounds of professional acoustic room treatment did to my Lisbon lounge’s reverb time:
In Lisbon two years ago, we saw how a rug did next-to-nothing for reverb in an empty room with a tiled floor. Today, we’ve seen it do next-to-nothing for reverb in a fully acoustically treated space with a wooden floor.
No wonder professional acoustician Jesco Lohan says, “a carpet isn’t a good strategy for treating a room if you’re serious about improving the sound of your speakers”. Anyone saying otherwise has yet to come to terms with the inevitable: that the only predictable way to lower a room’s reverb time is to treat multiple walls and the ceiling with panels whose absorption coefficient is known.
