The DIY Path to Better Acoustics

Note: Byte Gremlin uses Affiliate Links, which generate income to help support the site.

I’ve recently completed my first pass at transforming the acoustics of my theater, and let me start by saying, the difference was mind-blowing. Roughly $300 later, movie dialogue is easier to understand, the operatic tones of John Williams are richer, and action sequences are more enveloping.

In short, I feel like I have a whole new theater for less than I spent on a single speaker.

$300 later, movie dialogue is easier to understand, the operatic tones of John Williams are richer, and action sequences are more enveloping.

While the actual process of treating my theater only took about a week, many months were spent researching different treatment methodologies and implementations. Ask a dozen acoustics nerds how to treat a room, and you’ll get a dozen different answers. My intention with this article is to thread the needle, providing a clear approach with concrete steps you can take to figure out your rooms problems and improve your sound, without exploding your head or your wallet.

That said, there is no “generic” treatment for a room. The acoustics of your space are unique, and any good treatment plan is going to start with some basic measurements of your space. You can’t fix your rooms acoustic problems until you know concretely what those problems are.

That’s where we’ll begin.

NOTE: Acoustic treatment should be the LAST thing you do to your home theater. What treatment you make, and where you place it is totally dependent on the placement of your speakers, and your main listening position. A common rule of thumb is, adding or moving anything lamp sized or larger will alter the sound in your space. Make sure you’re set for all the basics before ploughing ahead! If you’re not, check out the ByteGremlin Home Theater guide.

The Dual Nemeses of Sound Quality:

Before we get into buying equipment, making measurements and treating your space, you need to understand the two underlying problems room treatment aims to solve.

Uneven Frequency Response:

Just like light has different wavelengths we can see (i.e. all the colors of the rainbow), sound waves can be broken down into different frequencies. A bass drum will sound in a range of low frequencies, a flute will play at high frequencies, and human voices occupy a range in the middle-high region of audible sound. In a perfect world for a given volume, speakers would reproduce all frequencies of sound to be the same volume.

In the real world, the objects in your room (and the design of the speaker itself) cause some frequencies to be louder than others. A room with uneven frequency response can result in unnatural or muddy sounding audio, where the louder frequencies drown out the quieter ones.

Untreated Early Reflections:

The second denizen of home theater sound lies in the many paths sound travels as it makes its way from the speakers to your ears. You may visualize the sound from your speakers traveling in a straight line to your ears, but actually, sound emits from your speakers in all directions, with each part of the wave bouncing off whatever surface it runs into first. Many of those reflected sound waves bounce around the room before eventually hitting your ears. Your brain takes the earliest sets of these reflections and merges them with the sound waves that travel directly to your ears, which has the effect of muddying the clarity (and sometimes directionality) of the sound.

 

Part One: Measuring the Room

Acoustics is literally a science in and of itself, but for the purpose of treating a room, I’m just going to tell you how to take measurements of your room acoustics, and what to do with the series of squiggly lines that represent that data.

Gearing up for measurements:

For this guide, we’ll be using the go-to room measurement software for acoustics, Room Equalizer Wizard (REW). It’s free, it’s fantastic, and you can get it here. Note to Mac users: if you have any issues getting the software to load up, try using the betas.

For my room, I used the UMIK-1 USB microphone. These are pre-calibrated at the factory, and include a digital calibration file that’s custom to each mic sold. It’s affordable, hassle free to setup, and reliable.

You’ll also need some sort of stand to hold the mic in the correct position. The UMIK-1 includes a rotating microphone clip that screws onto a standard 3/8″ camera tripod mount, so if you’ve got a camera tripod in a closet, that’ll work just fine.

If you’re a millennial who scoffs at the idea of dedicated cameras (shame on you) and lacks access to a tripod, the best option is to buy a boom mic stand. If you go this route, you’ll also want a 5/8″ to 3/8″ adapter screw in order to connect the UMIK mic clip to the larger threads on the stand.

Pro tip, if you’re using a home theater receiver with built-in calibration software like Audyssey, I highly suggest buying an angled camera mounting arm along with the mic stand. This will allow you to more precisely and easily line up your receiver’s included microphone (a process you need to complete seven or eight times per calibration on most receivers).

Now you’ve got the gear, we’re ready to take some measurements.

 

Taking measurements with REW:

Setting up your equipment with REW is pretty painless. First, connect your USB microphone to your computer, and connect an HDMI cord from your computer to your receiver, then launch REW. You’ll be asked if you have a calibration file for your mic. The UMIK actually has two files, one for measuring with the mic offset vertically by 90 degrees, and a standard file calibrated for measuring with the mic pointed right at the sound source. Pick the standard file.

Next, get your mic mounted on your stand. Place it in the center of your prime listing position, at ear height. If you’ve got a high backed sofa, you’ll want to cheat the microphone up and forward a few inches so it’s not right next to the fabric, as this can affect the measurement.

Power up your system, select whichever receiver input your computer is connected to, and head into the REW preferences panel. You may need to adjust your input and output. You want to make sure the program is sending multi-channel sound out to your receiver (as opposed to stereo), and that your Usb mic is selected (as opposed to measuring from the internal mic on a laptop).

For the most accurate results, you’ll want to crank the volume up loud, and minimize ambient noise as much as possible (turn off the HVAC, and politely ask your significant other to can it). Definitely grab some ear plugs for this.

For the most accurate results, you’ll want to crank the volume up loud

If your receiver includes room correction software such as Audyssey, YPAO, or ARC, make sure you disable this, as otherwise you’ll be measuring a skewed frequency range rather than the raw uncorrected sound. Also make sure you disable any EQ presets that may be enabled (such as “Vocal Boost,” Sci-Fi, etc).

In REW, use the level check button to make sure you’ve got the volume up nice and high without clipping (aim for a headroom of -5 DB) The level meter on screen will turn red if you’ve got the volume too loud.

You’ll want to grab an individual measurement from each speaker with your subwoofer unplugged, and then one final measurement of your front left and right channels with the subwoofer connected. For the individual measurements, ALWAYS point the microphone at the center of the speaker driver array. For the final fronts measurement, point the mic dead center.

REW makes it easy to zoom in and out of your sound range, and to view one measurement on any graph you want, so always take a measurement of the full frequency range your speakers emit (10hz to 20,000 is a good typical range if you’re unsure).

Finally, select “measure” in REW, adjust your settings as per the below screenshot, and hit measure. If you’re not bothering with earplugs, you can set a start delay for the measurement, allowing you to click measure and haul ass out of the room before the big SWOOSH.

By default, each measurement is named for the time it was made. It’s helpful to change these names to the name of the speaker being measured. Otherwise, the half dozen plus measurements will be impossible to differentiate, and therefore useless.

Once you’ve made your measurements and gotten a sore back from bending over to adjust your microphone a few dozen times, we’re ready to dive into all that data. Ahh… SCIENCE.

Finding your Room’s Problem Areas:

REW offers up a dizzying assortment of graphs, but for the treatment of a typical space, there’s only three data points we need to worry about. I always start by pulling up this data for my main left, main right, and center channel speakers, as the sound quality from these channels is going to define your soundscape more than any speakers behind (or above) your ears.

I always start by pulling up this data for my main left, main right, and center channel speakers

An important note: The measurements you’ve already taken hold all the information REW needs to generate a variety of different graphs. If you’re reading this as research before you dive into the process, you don’t need to do specific measurements for the below data. Just measure as recommended above, and you’ll be able to generate each of the graphs we’re talking about when you’re ready.

REW is complex software, and the interface isn’t super user friendly. I’ll offer some basic steps to follow to gather the information we need, but if you get lost or are looking for more in-depth REW instructions, check out the addition links at the end of this section.

The Data:

There’s three graphs we need to generate to understand our room’s problems. The first, the Time Frequency Curve, reveals any issues with Uneven Frequency Response.

My Right Speaker’s TFC Graph after some treatment

Generating this graph is easy. First, open REW and open up your set of saved measurements. In the center view panel, select the “All SPL” Tab, and uncheck all the channel selections in the lower panel, save for one (I’d suggest starting with your main Left, or main Right speakers).

This will pull up a spiky frequency chart, but we need to make some adjustments to make this view useful. First, pull up the “limits” on the righthand side, and set them as below:

We now have a useful range of frequencies and volume to look at, but the raw graph is a bit too spiky to draw conclusions from. Instead, we’ll smooth out the data, to give us a more practical graph. Head into “controls,” and select “1/12” under smoothing, and then select “apply smoothing.”

This chart gives you a baseline to judge the frequency response of your speakers and your room. Ideally, you want as much as possible of the frequency range to be within 10 DB of one another, but your room, your speakers, and your living companions will set limits on just how flat you can really make this graph. In my case, I didn’t see a huge change to my frequency response after treatment, but my biggest problem was related to untreated early reflections, and that’s where my room was dramatically improved.

To assess reflections, we need to pull up an ETC graph.

My Right Speaker ETC Graph After treatment. Believe it or not, this is BETTER

Instead of dealing with frequencies, the ETC graph shows the volume of different sound waves over time. Generate yours by selecting an individual channel measurement on the left, and then selecting the “impulse tab in the center view. Only Envelope (ETC) should be checked. Once again, we need to set limits as below (in this case, we’re limiting the range of time on the graph to a useful level.

ETC Limits Settings

An ETC graph for an ideal room would show the initial peak (which is the sound wave that hits your ears directly), followed by a huge drop-off in volume, as most early reflections would be tamed. A couple of early reflections can add character to the sound of the room, but when you have five or more peaks of similar volume to the direct sound, you’ve got problems.

In my case, I had a series of high peaks, almost as loud as my direct sound wave, arriving shortly afterwards.

Here’s my pre-treatment Right channel for comparison

For more info on how to look at this graph, check out this article from GIK Acoustics:

Unpacking ETC: Time-domain measurements & early reflections

In addition to showing us the state of our room’s reflections, the ETC graph can actually be used to identify which surfaces are causing the reflected sound. This takes some simple math, some string, and ideally, a friend.

Finding Reflection points:

Because the ETC graph is identifying the time it takes the sound to reach the microphone, it can also calculate the distance of the path of that sound. Once you’ve identified a spike (or reflection) in the graph you wish to identify, move your cursor to the 0ms spike (AKA the first and only direct sound), hold down the control key, and right click and drag to the reflection point. The red distance show in feet is the extra distance that sound wave traveled over the direct sound wave.

This is the arts and crafts portion of our experiment.

To identify what object or surface the sound is bouncing off of, you need to measure the distance from the speaker to the microphone, and then ADD the red distance number identified in REW. This is the total distance that particular reflected sound traveled to get from the speaker, to the reflection point, and then to your ears.

This is the arts and crafts portion of our experiment. Measure out and cut a length of string exactly as long as the path of the sound. Tape one side to the speaker, and have a friend hold the other end to your microphone. Next, pull the string taught, moving around the room until you find the point where the string both strikes and departs a surface at the same angle. This is the point that’s causing the reflected sound!

Before you treat the room, you’ll want to take some time and mark the earliest (and loudest) reflection points for each of your front three speakers. If you’re having trouble finding the place where the string lines up correctly, remember that it could be on the ceiling or the floor. We’ll address how to treat these problem areas later.

The final REW graph we’re interested in is called a waterfall graph, and it’s actually just another way of looking at reflections. Instead of looking at reflections over time, a waterfall graph shows you how loud (or how intense) the reflections are at different frequencies. It’s great for getting a visual overview of how much reverberation you’re getting in a room, and it’s particularly effective when diagnosing and treating problems in the bass range of your sound (typically 300hz and below). If you clap your hands and hear a pronounced ringing, this is the graph for you.

Post Treatment Waterfall. This Graph reveals that my room could use more Bass Traps.

We open this graph by selecting our measurement on the left, then selecting the waterfall option in the center view (it may be hidden behind a double arrow). Again, we want to set our limits. If we’re looking for bass specific problems, set as below

An ideal waterfall graph will be fairly flat and smooth left to right. The three dimensional area reaching toward you is the amount of ringing/decay in the sound. Ideally you want this to drop off quickly, otherwise you’ll hear a nasty echo that’ll impact the quality of your audio.

So now you’ve analyzed your room, and you’ve got a rough idea of your problems. You may have lots of early reflections to tame, ringing base frequencies that are out of control, and/or huge troughs in your frequency response. It’s time to get our treatment on!

 

Looking for more detailed guidance on REW? Check out these resources:

Continued on Page 2:

Pages: 1 2 3