Its been a long (loooooong) time since my last blog post, mostly due to lack of time for any “after hours” projects. Finally, I have something new I’d like to post on the blog. Unlike most previous projects, this isn’t electronics related, but it is Audio/DIY/Measurements related. So hopefully, you will find this post interesting, and perhaps even useful.
A while ago, I’ve moved into a new place, and finally had a space I could use for a stereo/home-theater. This was the main use of the new space I’ve had, which meant I had much more freedom in the setup and room organization than I’ve ever had in the past. One thing that was clear from the get go is that this room has noticeable acoustics problems, as will be detailed later, and therefore it had to be treated to some extent. In this post I’d like to go into some detail about the steps I took, so far, to deal with these acoustic issue. What I chose as my targets, what I chose to deal with passively/actively, how I built the panels I’ve used, etc.
Due to the length of this write-up, I’ll split this into 2 parts, with this post being the first. The first part will include mostly introduction to the topic, as well as introducing the room structure, and initial measurements with no treatment. In the next, second part, I will detail step by step, the different modifications and their effect on the measurements.
Preface/Background:
Before we dive into the details, let me start with some preface. One thing I’ve set as a limit from the get go, is that I didn’t want to turn this room into a “looney house”. I wanted it to be a pleasant room to be in, with the acoustic treatment kept to a minimum, especially esthetically. Now that this is out of the way, lets talk practically, what were the issues that were present in this room, what is the structure of said room, and what should even be the target to aim for when dealing with acoustics?!
Room Structure:
The room I’m placing the sound system in is built inside a basement (partially below ground level). As such, most of the walls are concrete, with only a small portion being cinder-blocks. The room dimensions are approximately 5.7mX5.9mX2.4m (LxWxH). Being built in the basement, there is a stair-case which has an elevated ceiling compared to the rest of the room. There is a big wall closet on one side of the room, and big (wide and high) window on another, along with a few smaller windows. The basic sketch can be seen below in figure 1.
A few small modification that were done to the room mostly for esthetic reasons/practical considerations that have nothing to do with the sound system included:
– Front (top of image) wall covered with bricks on the inside, mostly for aesthetic reasons. The bricks chosen have a fairly deep (~2cm) and random variance to their surface, so they will act as a form of diffuser for higher frequencies.
– Left (window) wall has been thickened slightly (~4cm) with drywall sheets placed on metal rails. This has some benefit over concrete wall in the degree of hardness and therefore reflections it would provide.
Acoustics basics, in the context of this post:
Before we go into the problems in this room, and what can be done to solve them, before we can even talk about measuring them, we should first discuss the basics of acoustics. This is by no means a comprehensive guide to it, this topic alone can consume hundreds of pages and more to even cover at a reasonable level. I fully recommend reading about it more if this is of interest to you. Some good references are Floyd Toole’s “Sound Reproduction” which is considered the de-facto standard and intro to anyone interested in this, and Everest’s/Pohlmann’s “Master Handbook of Acoustics”.
The limited background I’d like to cover here is what I think are the main take-aways for a first level measurement and understanding of the room you are in. These would be:
1 – Frequency response in room. That is, how loud each frequency will sound (in steady state) in the measured position in the room. Whenever you measure in-room response, no matter how good your speakers are, you are most likely to see quite significant deviations from the target curve. This will be most profound at lower frequencies, and with higher (frequency) resolution measurements. This is caused by the interaction with the room and what’s in it. A few “buzz” words you can use to look for more info on the topic would be “room modes”, “Schroeder Frequency”, and “SBIR”.
There are different “optimal” target curves that you can find, but generally they all agree there should be a downward tilting graph. There are different slopes to it, and some with extra elevated bass response, but that is of 2nd order, and can be refined according to preference, and use case of the sound reproduction system (music genre/movies/etc).
This target curve might seem strange, “why no a flat response”? That isn’t something a short explanation would do justice with, but suffice to say, that is what was found to “sound good” to most people. It is ok to deviate from this target curve, especially if its over narrow frequency bands, so there’s no need to be too aggressive in achieving this curve exactly. You can look up the term “psychoacoustic smoothing” for some more info on this if you’d like to get some better understanding of this.
2 – Sound decay time in the room. This is a measure of how long it takes for a sound that is sent into the room to be sufficiently attenuated. There are multiple ways of measuring it, and one fairly common and accepted is the “RT60 Decay” time. In practice, in small rooms (all domestic room for that matter), this isn’t quite the correct measure, but we can use this term as its widely used in this regard as well. That measure is, how long it would take for a sound sent into the room to decay by 60dB.
So surely, we want this RT60 to be in the order of 0 seconds, or as close to it, so there is no echo, right?! Interestingly, we (humans) actually need this to be a few 100’s of milliseconds [ms] for us to feel comfortable in a room. The exact target number isn’t so clear though. It is a function of the intended use of the room (conversing/ music listening/ watching movies/ sound mixing), and can be divided further even by the type of music being listened to, and personal taste.
So what RT60 decay time value should you target? If you will read across online forums, you will see a number of 300ms mentioned by many. This is found to be a good number with a good balance for many use cases. Before using this as a target, if possible, I recommend being in such a room and having a listen for yourself. For instance, I found myself quite disappointed in what I’ve heard in a room treated to this target. I’ve listened to more than one combination of speakers/electronics, with different types of music, and for the most part I must say it was underwhelming. In “audiophile” terms, I would say it was too “dead” for my ears. I’ve asked a friend for another opinion, and he shared similar feedback afterwards. Other target figures can be higher, say 350-400ms, and some people (especially these who listen mostly to classical music) might even enjoy it better with 500-600ms. This measurement is done as a function of frequency, and results can vary quite a bit with frequency.
3 – ETC/Clarity/etc?
There are other measurements you can make, which would provide much info. For instance, they can help you understand where a certain reflection is coming from by translating time delay to distance from the speaker/measurement point. I prefer to keep these measurements out of this post, as its already becoming too long. I will say there is a lot of information about it out there, and I encourage you to go and read about it further if its of interest to you. This can prove very useful, and has implications to “what sounds good” or “how intelligible speech is”.
Performing Measurements:
Nowadays, its amazing simple, and cheap, to make acoustic measurements at home. There are some measurements that call for more costly tools, but for the most part, the information needed to significantly improve your room acoustics can be had from very cheap and easy to use tools.
What you will need is a computer/laptop which you already have, a measurement microphone, and a software to perform the analysis. 2 readily available options for a microphone are a Umik-1 from Minidsp, and UMM-6 from Dayton Audio. Both are USB devices, so no need for any additional sound cards to interface with them. Both companies will provide a calibration file you can download from they website based on the serial number of the microphone, so that you can correct the frequency response measurement accordingly.
As far as software options go, there are a few of these, but the most widely used as far as I’ve seen, is REW. Its a very capable tool that can do quite a lot. Its free, and you can experiment with it even without having a calibrated microphone.
Passive/Active Solutions:
When it comes to fixing acoustic problems, there are a few different solutions. One would be a passive solution. That is, some sort of fix to the acoustic problem at its source. This could be in the form of absorbing/redirecting the sound that is being reflected by the room. There are different types of absorbers (pressure/velocity for instance), and different types of redirectors (diffuse/scatter). There is much information about this elsewhere, but a 1st order estimate in most domestic rooms would be that a velocity absorber would most likely be the most appropriate solution. That isn’t always that case, but more often than not, this seems to be true.
The main downside of the passive correction, is that it takes up space, and affects aesthetics. Its also becoming quite big (and deep) the lower in frequency you’d like it to be effective. Therefore, for low frequency especially, this becomes impractical in most cases.
Next comes an active correction system. With the exception of some very extreme solutions (such as Dirac’s ART which is on a whole different level/concept), most active solutions are limited in what it is they can fix. Frequency response, especially at a single listening position, is something they can fix quite well. Impulse response (phase part of the frequency response) is also possible with some of the solutions. Fixing more than that, becomes much more difficult.
That is why in most cases, its better to deal with things in a passive manner to the extent that is possible by other constraints, and then add active correction on top of that. Some people have a fundamental opinion against adding active correction to a system, for reasons that might or might not be valid. I suggest you give it a try yourself before you form such a solid opinion against it.
Problems Present in My Room and My Goals/Approach:
As can be seen in figure 1, the room I’m placing my system in, is quite difficult. There are a lot of hard surfaces, the room if fairly square, which makes the room modes more pronounced at certain frequencies, and the elevated ceiling at the staircase is causing quite a lot of issues with echoes.
From the get-go, I’ve decided I’m going to fix in a passive manner what ever I can, up to a limit. That is, I want the acoustic treatment to be limited so that the room doesn’t feel too different from a normal room. I wanted it to be a pleasant room to be in, even when not listening to music. The rest, I would fix actively to the extent that is possible. This would be mostly frequency response, and specifically, at lower frequencies where room modes dominate the sound signature.
I’ve performed a measurement after first populating the room (sofa/TV/speakers/equipment stand/carpet mostly for aesthetics). The frequency response is of little interest to me at this point. Figure 2 shows that (Left – green, Right – blue) with 1/6th octave smoothing. As you can see there are quite large variations, especially in the bass region where the room dominates. This is shown mostly for completeness/in case anyone is interested in this, this isn’t a measurements I’m repeating often, as that isn’t what I was trying to fix with the passive treatment.
Figure 3 shows the Decay measurement results, with the 60dB crossing trace overlayed on top of the results. This is shown for the right speaker only here. As you can see its quite high, >600ms at all frequencies below 4KHz, and rising sharply at <150Hz (as much as 1s at 100Hz). This rise at lower frequencies (and fall at higher frequencies) isn’t surprising, and is quite common in most rooms. Some rise at lower frequencies isn’t much of a problem, as is a fall at higher frequencies. However, in this case, these numbers are quite high throughout the spectrum.
My room is used for both stereo, and home-theater. My music taste in music genres is fairly wide, but I do listen mostly to acoustic instruments. Based on my previous experience in different rooms, and with different systems, I knew that 300ms would probably be too low for my taste, but the current 500-700ms was too high, and I didn’t like it. The sound of the room was too pronounced. Therefore I’ve set my goal at 400ms, but this can always change as I get closer to the goal and listen to music and “re-learn” the room at each step.
That is all for the first part of this blog post. In the second (and hopefully final) part I will describe the passive treatments I added to the room, including steps from the build and choice of materials, as well as some measurement results, to go along with it.