This is part of 2 this series of posts about room acoustics improvement for my listening room. In part 1, I’ve provided the background about the room structure, different issues I was having, and how I was planning on tacking them while keeping the room relatively “normal”. In this, part 2, I will go into more details about the modifications I’ve made to the room, with some details about building the treatment, and materials that were used, along with some measurements to keep things more clear and provide some tangible data as to what were the differences achieved. This might be useful to others considering what sort of change they can expect from similar modifications.
Before dwelling into the details, I’d like to add a note or two about the measurements. When comparing the measurements, keep in mind that in between different steps (some of which were weeks/months apart due to lack of free time), there were potentially additional changes. These might include moving speakers closer/further away from the walls, moving the listening position a bit closer or to the side, and other potential minor changes. Therefore, not all measurements can be directly compared, but these changes are limited, and more substantial differences can defiantly be attributed to the treatments being compared. Additionally, not all measurements were done at the exact same level (as they were performed at different times, and I wasn’t too concerned with the absolute SPL for the measurement, as long as its sufficiently above system noise level for a trustworthy measurement).
Lets get to work…
Elevated Ceiling Absorption
First things I wanted to tackle was the elevated ceiling above the stairs. This seemed like a most probable source of many of the issues I was hearing. This is an area of 2.5m*1.6m that is approximately 1.6m higher than the rest of the room’s ceiling. Standing closer to the stairs (or at the top of the stairs) made this crystal clear, that its a problematic part of the room. For the sake of completeness I’m including the room diagram that was presented in part 1, once more.
As a “what if” exploration, I chose to use cushions from one of the old sofa’s we’ve had in the past. To be able to “mount” them to the ceiling, I’ve purchased a pair of telescopic poles that can be adjusted to the ceiling height anywhere in the room. Here’s an example of placing some of these cushions in the area of interest presented in figure 2.
This technique, with different number of cushions at different locations proved very useful for getting a better understand of what are the problematic areas and what are the points of interest to get better sound. This can be useful for both objective measurements and listening tests.
The results from these few measurements showed that indeed, adding some absorbing material at the elevated portion of the ceiling would help improve sound noticeably.
I’ve decided to build a light wooden structure shaped like a ladder that would hold acoustic material next to the ceiling. For the material used I’ve looked into different options taking into account material cost, local availability, and how hazardous it is/isn’t. After considering the different options, I chose to go with Mineral wool. The specific product I went for is “Knauf Earthwool” with a density of 36KG/m3.
To cover the earthwool, I’ve decided to go with the “Hexagon Acoustic Panels” you can find on practically every online sales site from different suppliers. They are practically all similar as far as I’ve seen from using a few of these. In my case they were mostly meant as a visual cover for the material, but they are obviously having their own effect. Comparing measurements with/without these hexagons in place showed what was quite expected. At lower frequencies the additional of the hexagons (12mm thick) improved absorption, but at higher frequency it has diminished it somewhat. Thankfully, lower frequencies is actually were we want to focus our attention more. To be able to cover the ceiling with these panels I’ve first stapled thin cardboard pieces to the wooden pieces to have something the hexagonal panels can grab on to with some glue. Figure 4 shows the result after it was all said and done. As you can see, I chose to combine a few shades of panels (white/gray), but a single color option would have looked great as well. This is purely a matter of personal taste.
Now for some measurement results, as you can see in figure 6, most of the spectrum is lowered compared to figure 5 (reference, same one presented in part 1). This has also made a very noticeable audible difference when listening to music, this is something that is clear right away, with no need for a microphone to confirm it.
As you can see there is quite a bit of difference in the results, both in shape, and in slope. At mid-range frequencies the peak dropped by approximately 150ms, with the mid-range average now sitting closer to 500ms than 650ms as it was originally. This was quite a good improvement for such a small area of coverage of this treatment, just goes to show how much of a problem this part of the room was. It was quite cost effective as well, at slightly over 200$ for everything needed (for a 4m2 area).
Big Window Absorption
On the left side of the room, there is a big window that is approximately 6m2 in area. This was also causing noticeable reflections, which were frequency dependent of course. In case of a window the treatment options are far more limited than other materials, with a curtain being almost the only practical option in this case. I’ve therefore decided to go with this solution, and ordered a curtain that was meant to be both acoustically beneficial, as well as aesthetically matching the room. Another benefit is the removal of light reflections from the TV via the window that was quite distracting when watching movies.
I’ve opted to go for a fairly heavy fabric, with a 2 layer material, with deeper “pockets” which uses more material, and therefore absorbs better.
Figure 8 shows the measurement following this addition, with the curtain halfway closed, as it is typically left on a daily basis.
Interestingly this was very effective, this can not only be heard right away by making sound noticeably clearer (which isn’t really reflected by the RT60 decay measurement directly), it also shows up as a wideband improvement in RT60 decay time. This dropped everything from 250Hz up to 11KHz by as much as 50ms at midrange area. Talk about 2 birds with one stone, both an acoustic improvement of noteworthy value, as well as improving room aesthetics and reducing light reflection from the TV through the window.
Curtains cost can vary quite dramatically by choice of material, and installation type (manual/electric), as well as dimensions and even the place you get it from. In my case it was quite reasonably priced compared with the dimensions, at ~500$.
Acoustic Panels on the Back Wall
Next, I turned my attention to the back wall of the room. This was a fairly big piece of wall that was empty with no absorption/diffusion (with the exception of the 2 doors located on this side of the room). I knew I had to add some absorption there, which was also verified by placing some cushions there as another “what-if” experiment as shown in figure 9. After testing a few commercial foam panels, I’ve decided I’m going to build my own panels for this wall.
As the main absorption material I stuck to the same 36KG/m3 Earthwool I’ve used previously. I wanted to make it look nicer though, and more along the lines of stereo/theater room. I’ve therefore decided to print some film posters onto canvas, and use that to cover the panels. Additionally, I’ve decided to cover the sides with aluminium (painted in gray paint) to add some extra depth to the panels, and improve aesthetics. I’ve seen this done in some office building, and liked the idea a lot.
The aluminium stripes I’ve used were 70mm wide, while the earthwool (and therefore MDF frames I’ve constructed for it) were just 50mm in thickness. This was by design, to allow for an easier hanging of the panels. This is also beneficial in terms of low frequency absorption, allowing these panels to work to a slightly lower frequency. The side view diagram of the planned structure can be seen in figure 10. Such a structure would allow the panels to be taken off easily, some left/right sliding over the mount for fine alignment, as well as switching places between them easily to match visually between adjacent panels.
I chose to build the “wooden” frame out of MDF. It is completely flat, and not too difficult to work with. Its heavier and more rigid than pine which many people choose to use, but this is not much of an issue for this application. I’ve used a small nail-gun/staples to put it all together, and had 5 frames ready and waiting to be used. Note that I am using MDF which is covered with “Formaika”, not because its useful, but simply because that’s the material I’ve had on hand. I’ve used green MDF as much as I can, simply because its more tolerant of water in case it ever gets exposed to it somehow.
Now, I turned my attention to the absorption material, cut it into pieces of the target size (each panel is 60X90cm), placed it inside the panels, and covered them with the canvas printouts I’ve previously ordered. This was once again done using an electric staple gun.
Next, were the aluminium frames. After measuring and cutting to size, I’ve glued the frames together with Den Braven’s “Super 7”, and let it set. Once set, I’ve used spray paint to give the aluminium frames a nice and clean gray shade. Finally, it was time to put it all together, where once again “Super 7” was used to glue it all to place.
The last step was to glue the MDF hangers onto the wall. I’ve used the high-tack variant of “Super 7” for this, and cut some grooves into the MDF to allow it to grab better. I’ve combined some pictures of the process together below.
Once everything has set, I’ve hung the completed acoustic panels on the wall. 4 were placed on the rear wall, and 1 on the left wall next to the back corner of the room. This can be seen in the gallery below.
I’ve repeated the measurements with the panels in place, with the results shown in figure 11. The panels have made a noticeable difference, mostly in the 220Hz-2.2KHz region, with as much as 80ms lower RT60 decay time at 500Hz. This is interesting to see, showing the effectiveness of these panels down to <250Hz owing to their depth of >70mm from the wall.
I think these panels turned out quite nice, and the cost of materials wasn’t too high either. All 5 of these panels ended up costing ~300$ in material. Roughly half that was spent on the material for the aluminium frames, and printed canvas. Therefore, if you dont mind a plain look of the complete panel being covered in fabric, you can get it done for ~25$ per panel.
Seeing as 2 of the first reflection points have been dealt with to some degree (curtain on left side, and closet located further away from the speaker on the right), next I turned my attention to the ceiling. Once again, the telescopic poles, along with the old sofa cushions came in handy for another “what-if” exploration by pressing the cushions against the ceiling. This was, once more, clear from the first listening test that it would be a great improvement. It not only made sound clearer, it also helped bring the sound-stage lower to the height of the speakers, unlike how it was previously where everything seemed to originate from above the speakers.
I’ve considered different options for the ceiling treatment, as well as building my own panels as I did for the back wall. However, after checking different options, I’ve decided to go with pre-made commercial panels for this application. Mostly for the looks. One candidate I was quite interested in were Artnovion’s Myriad F Double SQR clouds. They have nice round edges, and don’t look bad at all. While waiting for these to be available in stock with the local supplier in white color as I was planning on using, I was offered a similar product from another supplier. Its technically identical in all parameters, but a few CM larger, and didn’t have the rounded edges. The product I’ve used was “Style” from “isolspace”. It was available in white color, and I could get it the same day.
I’ve installed it, and was happy to see it didn’t look as bad or as claustrophobic as I was worried it might be. I chose to install them approximately 10cm below the ceiling level, which would make them more effective at lower frequencies, with some trade-off at higher frequencies perhaps.
The measurement results after installation of the ceiling panels can be seen in figure 14 below. With the added space between the panels and the ceiling, it starts being effective at <300Hz, but more pronounced is the difference it makes at higher frequencies reducing up to 50ms of the RT60 decay time at the midrange, and remaining effective all the way up to 6KHz. This in turn not only reduced the RT60 decay time further, but also makes it more even across frequencies.
Cost wise these panels can vary somewhat in cost, but there are typically going for <200$ a piece with mounting HW and shipping included (you will need these delivered to you, they are 120cmX120cm, and won’t fit in most private cars without bending them, which would leave its mark so try and avoid it).
The items shown above amount to the entirety of passive treatment I’ve installed in the listening room to date. I don’t currently plan on additional changes, but as we all know, these thing can change over time. The resulting sound in this room is significantly better than it was originally, and it got better with each an every step along the way. I would dare say that improving the acoustics is as big of a change as replacing speakers, and this is something that cannot be overlooked by anyone who is serious about this hobby, and spends significant time and funds on it. Even counting all costs of the items discussed in this post together, this is noticeably cheaper than the cost of speakers + electronics it is complementing. In terms of value for money, this is a major bang for your buck for any serious hi-fi system.
You can see that the target I’ve set for myself, was mostly met. The figure is at, or below, 400ms for almost all of the spectrum. There is a sharp rise at lower frequencies (<200Hz or so), which is expected, and something I was aware of from the get go. However, thankfully, a rise of RT60 decay time at lower frequencies isn’t too bad as far as acoustic problems go. At 100Hz it is now ~650ms, which is a full 350ms lower than it was originally in this room. For the moment at least, I don’t plan on improving this frequency band, mostly due to the size of the treatment that would be required, and the way it would make the room look/feel, but that might change in the future.
Other than the passive treatment steps covered here, I’m also running an active room correction in my systems. Currently they are 2 different software/algorithm tools, performed by 2 different pieces of HW, as I’ve decided at some point to separate the stereo/theater systems completely. That is, different source, different processor/DAC, amplifier, and loudspeakers.
I quite like the active correction as it gives additional control with limited effort. For instance, it helps correct frequency response, which is especially hard to do passively at lower frequencies. Additionally, it lets you choose different target profiles for different use cases, music genres, or even depending on your mood this evening. Finally, its biggest advantage, is that it can be reconfigured (and even turned off completely) with a flip of a switch.
Until next time (if there will be such), this is all for this post.