Simulation of Complex Rooms

As I design my retirement cottage, I would like to simulate the acoustics and correct any mistakes in design prior to construction.
Unfortunately, the simulation tools I have found assume "normal" construction and cannot replicate my plans. Hopefully someone can direct me to more sophisticated tools...
The listening room:
-The cottage will on a bluff above a scenic river so the exterior walls will be largely glass to enjoy the view.
- To address standing waves my idea is for the exterior walls to be inclined outwards, like an inverted pyramid, with approximately 30 degrees of inclination.
---This should eliminate standing waves between parallel walls, as they are not parallel
---Any standing waves must involve the ceiling, and the ceiling joists will be exposed as sound traps and provided with damping material
Are there software tools which can simulate such a room?
Are there precedents which have been empirically evaluated?
 
  • Like
Reactions: 1 user
diyAudio Moderator
Joined 2008
Paid Member
Parallel walls don't eliminate standing waves, although they redistribute them. In this example both shapes share these same modes.

dm1.png
 
  • Like
Reactions: 1 user
Administrator
Joined 2004
Paid Member
I'm working on it now. It's been a few years since I used CARA, but it's coming back.
Started with a 1000 sq foot cottage, all one room. Actually it's 93 sq meters, because CARA works in metric. I made the room all golden ratios L-W-H to give it a fighting chance. The first room has wood floors, gypsum board walls and ceilings, with full glass walls on the short sides. Everything is parallel. Of course it's awful with reflections everywhere and lumpy frequency response.

Now I'm working on a room of the same width and height, but the glass walls slope out 30 degrees. I've also put carpet on the floor, beams on the ceiling, some acoustic foam panels here and there and two book shelves. Just angling the glass makes a huge difference, but adding absorption and scattering helps too. Reflections are much more diffuse and room modes and frequency response has smoothed out. Amazing for a room with two glass walls.

I'll try to get the reflection animations and graphs posted here. BTW, do you have room dimensions in mind?
 
Wow, that is fantastic!
Let's use 8 meters square (far from golden ratio!) for an aera of 64 sq meters at the floor (with a larger ceiling due to inclination of the wall), with all four walls glass and angled out. No carpet, or perhaps some area carpets (a giant old oak just blew down in my little forest and I am having the local sawmill cut it, so I will have huge oak planks on the floor). The ceiling beams will be .4 meter tall and .4 meters apart. Absorbtion can be on them and/or between them. Let's use 3 meters to the bottom of the beams, or 3.4 meters to the ceiling.
 
  • Like
Reactions: 1 user
Administrator
Joined 2004
Paid Member
OK, I'll do a model of a room like that. Below is animation of a Dirac pulse in a room with tilted glass room with carpet, bookcases, ceiling beams and acoustic foam on sections of the wall and ceiling. It's hard to get these to play on the page, but I'll see what can be done.

ANGLED GLASS ENDS WITH ROOM TREATMENT
Angled treated.gif
 
  • Like
Reactions: 1 user
diyAudio Moderator
Joined 2008
Paid Member
but if the waves are directed to absorbent materials I should enjoy some damping?
In a rectangular room, the majority will hit the ceiling before the second reflection. The remaining in this example reaches the back wall before the second reflection where by this time they are diffuse and delayed sufficiently to begin making a positive contribution.

ce.png
 
Administrator
Joined 2004
Paid Member
OK, this is a weird room. It's like the radar room at the top of an airport control tower. :D Great for looking out and down on the world.
I built the room and did the sims, it lives up to its oddness. It's too live in the mids and highs, which will make it sound bass shy. I'll have to work on the beams and ceiling treatment and put a nice thick rug in there. The angled glass does do a good job of spreading around the reflections, even tho there are still too many of them. It might be better and more practical to bring the walls straight up 120cm or so for electrical, shelves, seats and so on. Cover the walls with wood, maybe even some diffusers. Angle out the windows from there.

Once we get the room fine tuned, I should be able to give you some samples of what it would sound like. CARA exports a stereo impulse file that can then be convolved with a recording to get an idea of what the room sounds like. Headphones are best for this.
 
Being bass shy is catastrophic. (my son sings opera, as a basso profunda) And I tried to play tuba in years past. But I know that this was mission-impossible, as both the view and the bass were precious.
What can be accomplished with partial walls (not floor to ceiling) separating the kitchen portion of the great-room from the rest?

I have some excellent headphones as I have designed and built vacuum tube headphone amps without output transformers based on 600 ohm headphones.
 
diyAudio Moderator
Joined 2008
Paid Member
The challenge with good bass is reproducing the upper bass and lower midrange which are in the usual baffle step region. Early reflections there due to the wide dispersion create resonances that are difficult to tame.

I haven't given you a solution there as it is not straightforward and there are multiple ways to handle it.
 
Administrator
Joined 2004
Paid Member
I have some sims ready that might be interesting and should allow us to hear 3 different rooms.

When the power comes back on at my house, I’ll post them. Spoiler, with enough acoustic treatment on the ceiling. I got the crazy glass house to behave.
 
  • Like
Reactions: 1 users
Administrator
Joined 2004
Paid Member
OK, here are some results, I hope that you can make sense of them. I modeled 3 similar rooms
  • 8x8 meters with no walls of ceiling. Carpet on floor. This "room" is ultra dry, no reverb.
  • 8x8 meter room with all parallel glass walls, hard floor, plasterboad ceiling. Very reverberent as you can image. It's almost all reflections.
  • 8x8 meter floorplan with glass walls that slope outward from the floor at a 30 degree angle. Carpet floor and wood beams on ceiling. A lot of acoustic foam in the ceiling.
Here is what the CARA analysis has to say about each room.

Room with no walls:
"The reverberation times are too small within the whole audible frequency range.​
Music reproduction in this room sounds rather dry and is lacking​
in a natural space impression, resulting in a slightly depressing atmosphere."​
Glass box room:
The reverberation times are too long at midrange and high frequencies.​
This reverberation spectrum makes music reproductions seem lacking in bass.​

Treated room with sloped glass walls:
"The reverberation times in this room are well-balanced over the whole audible frequency range.​
There is nothing special about this situation, it is considered ideal."​

Graphics: First the frequency response of each room.

No Walls.
No wall FR.png


The Glass Box
Glass box FR.png


Treated room with sloped glass walls.
Treated toom FR.png


In the No Walls super dry room we can see that CARA has simulated the comb filtering caused by the stereo setup. This causes a dull sound.
You can read more about that here: https://www.diyaudio.com/community/threads/fixing-the-stereo-phantom-center.277519/post-4398895
In the two rooms with reflections, those reflections fill in the dips caused by comb filtering between two speakers and a head. Of course the reflections have their own problems.

Impulse responses of the rooms.​

No walls.
No walls impulse.png


GlassBox.
All glass impulse.png


Treated room with sloped glass.
treated room impulse.png


The impulses show us what the reverb looks like over time. The impulses do not start at zero because we see the time of flight from speaker to ear. I don't know why there is a rough double impulse, it might be the model of the ELAC 516 speaker used in the sims. The impulse looks almost the same looking at a single speaker.
EDIT: The second big impulse is floor bounce. In a room with no floor, that second impulse goes away and so does the combing.
We see how bad the glass box room is, all hard parallel surfaces meaning many, many reflections . The treated and sloped glass room is better, but there is still some reverb there as seen starting at 40ms.
These impulses were used to make the sound files in the next post.
 
Last edited:
Administrator
Joined 2004
Paid Member
Here are links to some simulated samples. There are two source files.
  1. Male voice speaking English.
  2. Drum kit recorded dry.
Each of these has had the simulated sound added.
  1. Room with no walls
  2. Glass box room
  3. Room with sloped glass walls and heavy ceiling treatment.
The room with no walls does sound dull, but let it be your reference to judge the more reverberent rooms.

Voice:
  1. https://on.soundcloud.com/3FmU8m8uMA1dNU15A
  2. https://on.soundcloud.com/PP2G3TU9z93dkY539
  3. https://on.soundcloud.com/CNp4vDzLjhRA4LweA

Drums:
  1. https://on.soundcloud.com/2TPkHV5CLE7Roper8
  2. https://on.soundcloud.com/GVUAJctZt5xtb7iA6
  3. https://on.soundcloud.com/apaABMpiDUeW8iXN8