A more scientific approach to a room which lost its 'soul'

[mk2]

Years ago my employer was vacating, so there was lots of acoustic treatment (from meeting rooms) which I repurposed for home use. Perhaps too much!

Now I have a measurement mic and a calibrated SPL meter, so I'm sharing results and looking for guidance.

The problem: the room has, in a way, lost its 'soul'.

At the time I wasn't very scientific, I just sort of threw panels up and it reduced annoying flutter echo (from parallel walls) significantly. Picture of the room is attached (24ft x 12ft, 9ft ceilings):

Music comes from the speakers (Dynaudio Audience 62) and literally sounds like that -- crisp, clear and detailed, and coming only from between a pair of speakers in front of you.

If I could articulate it, I'd say the sound just doesn't 'hug' the listener for their enjoyment.

So where do I go from here? Perhaps to 'fill' the room in a more controlled way?

Feels like I need to really define (beyond enjoyment of the music) what I'm actually trying to optimise, in terms of reflections etc. I read the 'Perceptual effects' post, though I'm a relative novice here. Music is electronic, funk/soul, often not 'audiophile' recordings.

So using my new tools, I started experimenting. I removed some of the panels behind the listening position and measured. The SPL curve looks the same, but REW's "Decay" seems to be the most appropriate graph for what I'm hearing; reverberations are around longer. Removing 2 out of 3 of the rear panels gives an extra 4.5dB at 160ms (I don't have a measurement from removing all 3)

The difference is audible, perhaps some soul has returned, but not to the extent I would hope. Now the flutter echo is back and similarly annoying just when using the room (not music). I'm limited a bit by the front wall use as a projector screen.

Your guidance appreciated, many thanks.

 

[JeffS7444]

You might try some variation of the Harman house curve which tilts downwards to the right:
https://www.audiosciencereview.com/forum/index.php?threads/harman-target-curve-calculator.9618/

 

[Blumlein 88]

Are you doing any eq? You have flat in room response. You'd like a slightly down tilt as frequency goes up. Dropping that low end peak would be good too.

 

[AnalogSteph]

Bottom left panel probably doesn't do much for playback at all, it'll just reduce flutter echo in the other part of the room.

From the sketch I suspect that proportions just may not be that good. You are generally better off with a listening position closer to the speakers than to the back wall, though back wall treatment will do a lot if that can't be avoided. (Otherwise a bit of dispersion at the back may well do.) What's the angle between speakers from the listening position? If substantially less than 60°, widen it up.

The dips in the lower midrange / upper bass are a bit odd and unlikely to be beneficial (this range is generally underrepresented so I'd expect rather thin sound, not good), and then there's the question of why in-room response in the treble is flat like that. (How's on axis looking?) What's the floor like? If it's hard / reflective, I'd throw a panel up on the ceiling in the middle between speakers and listening position and see what that does.

BTW, sound coming from in between the speakers is quite expected if the mix doesn't provide anything else (that's just what happens if sources are merely panned), but it should be sounding as if sound sources are floating in thin air, not coming from speakers. I am not a fan of unrealistic "romanticism" in stereo playback with fake spaciousness generated by reverb provided by the room - I want to hear it if it's on the recording. If you really want extra reverb you can always use a VST reverb plugin like @KSTR does, in which case I guess you may prefer a bit of a large room.

Recommended RT60 generally is around the 0.3 s mark or a bit lower and flat across the spectrum, btw.

 

[Dj7675]

I would agree with others.. If you have the ability to eq, do so below 200-300. Smooth the bass and have some slope from around 200 and lower. The difference can be pretty dramatic. Although the panels wouldn’t affect this area much.

 

[FeddyLost]

Recommended RT60 generally is around the 0.3 s mark or a bit lower and flat across the spectrum, btw.

It depends a lot on room size, tastes and familiar sound.
Sometimes it's OK for hi-fi room to have 500 msec with a lot of diffusion. When specular reflections absent totally, it might sound very spatial and dense, but requires a lot of knowledge and relatively expensive woodwork.

I'd say the sound just doesn't 'hug' the listener for their enjoyment

First of all, you shall remove all panels till the "soul" will return. Then measure thoroughly everything that you can, especially RT.
And when you'll put panels on walls, measure results again panel after panel.
Also, try to move speakers forward and sit farther from back wall. If your speakers and listening point are too close to wall, direct sound and walls' reflections will blend too much, so it can destroy scene completely. Just flat blobs of sound.

But when I look at your plan, I don't really imagine how can you get high RT without flutter echo.
Flutter echo arise between two parallel planes, so you have to get rid of those planes and still have side/back reflections.
All the physical solutions that I can offer barely will make you and your spouse happier.

If this room would be completely dedicated to stereo, I'd try to move speakers and rollable screen into bay window and most probably all could end at this.
Side walls can be decorated/diffused by sparse bookshelves with books, back half of room might stay as is, because it's far away and will not return sound back immediately. Big cabinet with books/CDs/vinyls on main axis in back half of room will get more diffused sound back for better spaciousness. Maybe some absorbent panels on ceiling and first reflections points to get imaging better.
But usually such big changes are not welcome.

 

[mk2]

Hi everyone, thanks for the assortment of suggestions. I'll try and answer everything that was raised, and then I'll show you some more results from experiments tonight.

EQ'ing: I don't think this is the the variable I'm looking for. I have the amp's bass+treble for now and whilst a bit of tone adjustment is nice to my own tastes it doesn't do anything other than, really, adjust the tone within the confinds of everything else.

The panels outside the listening area of the room: actually these have a surprisingly audible effect on the room. It shows up in the RT60 graph of untreated room, vs. the introduction of that first panel. And it goes on from there.

Adding reverb digitally: Haven't experimented with this. I do believe you can probably synthesise almost anything with DSP, I suspect I'm going to hear a reverby sound coming from right in front of me. One for next time...

The room: brick walls, floorboards, carpets, no treatment on the ceiling. Ceiling is plasterboard to the room above.

So on to my experiments.

I started with removing all panels except the 'bass' traps as I didn't want to touch the rockwool (I use quotes as they're not chunky enough to reasonably trap any proper bass.) Room gets crazy echoey and fluttery; this is a Victorian terrace and all these walls are brick.

I did first set of experiments in the untreated room. Moving the listening position I kind of ruled out. I tried closer, or further from the wall. It affects the SPL curves, but not much and it's no more/less 'wild', just different. Moving the speakers forward had more interesting effect; more on that in a moment.

I read up about RT60 and adding the first panel (bottom left, outside the listening area) drops the RT60 numbers notably. Both it and the second one (left) reduces a significant amount of flutter. Both continue with each panel added, until the final result shown. I don't think any are thick enough to significantly affect SPL levels, but RT60 reduces from about 300-400ms, to 200-250ms overall.

With the soft room, I went back to speaker position. Moving them forwards seemed to be the most significant way to improve something; interesting to see it makes the graphs a bit less 'wild', and feels more 'in' the music by being closer. Other things I tried (wider, narrower, touching the wall) were negative as graphs gained more extreme peaks and troughs.

But back to the science and me trying to work out what variable I'm actually seeking to improve.

Listening to music in the untreated room was very different. But did it have the 'soul' I hoped it had? Perhaps a little, but I was still seeking something else.

It feels like whatever I'm hoping to increase isn't EQ, and it's not room reverb. Intuitively it feels like it might be nice for my ears to have a bit more sound from behind (feels a bit anechoic, I can 'sense' it).

Swapping the panels to my left and right for diffusers is probably not very practical, but maybe that's what I would ideally be looking for? To allow a bit of energy to enter the listener's ears from the sides/behind, but not in a reverby way? Trouble is it would be expensive and time consuming to test.

Thanks

 

[pozz]

What's the distance from your head to the back wall?

How thick are the panels? Do you know the material?

There may be a simpler solution. I note that the 500hz area is more depressed, in a wide region, than the rest in the RT60 graph.

 

[mk2]

What's the distance from your head to the back wall?

It's 2ft, which is really the only practical position. In the measurements, moving forward didn't seem to have a significant difference in practice or in the measurements, though I may be missing something in the results.

How thick are the panels? Do you know the material?

They're 2.5" thick. A dense foam material, looks like the same used or chair padding, with a timber frame and cloth cover.

That's most of them except bottom left. That one (and the rest of the ones I acquired) are something more hi-tec; a rockwool-like material which has a paper surfacing and plastered edges. Very fragile. I'm told they were quite expensive and are really good panels. But that will mean really good for some specific purpose, of course.

There may be a simpler solution. I note that the 500hz area is more depressed, in a wide region, than the rest in the RT60 graph.

This sounds interesting. For my understanding can you elaborate on what is is that the measurements are more 'depressed' and what the significance is?

Thanks

 

[Hipper]

I would like you to start at the beginning, which is to deal with bass issues. Sort out the bass region, say 0-300Hz, and the rest of the frequencies will be 'released'.

Could you show us a frequency response graph for the 0-500Hz region with no smoothing?

As you can't deal with the bass region with further room treatment, your options are careful positioning, multiple subs, or DSP/EQ. Use REW frequency response graphs to find the best locations. If you can move your chair, even better. Most advice is based on symmetrical rectangular rooms so with your odd shaped room you may end up with strange positioning. Frankly I think this is where DSP comes into its own. I highly recommend you consider it. You would use this mostly in the bass region for exact correction but could also use it like slightly more flexible tone controls to improve the higher frequencies.

When you talk about soul I suspect you mean some sort of imaging. The higher frequencies (usually above around 240Hz) start behaving like light rays in a mirror. Looked at like that you can see that reflections from your right side, the bay window, will be different from the left side - the sofa and open space beyond. This different reflectivity will lead to an imbalance of the phantom image you hear. The way to balance this out is to use panels that absorb these frequencies. If you put a panel behind the sofa and another exactly opposite in the bay window that should help. If these panels are on feet you can move them out of the way when not listening.

Regarding the panels you have, if you know their make you might be able to locate the site that sells them and they may have some measurements of the frequencies they absorb.

 

[mk2]

Thanks, there's definitely scope to experiment here.

I doubt I can clearly identify the panels, as there's no markings. But there's also the possibility to put them between the mic and speaker and measure them. I have plenty of the 'high tech' ones to experiment to the sides and in the bay window (though I won't be able to easily get much height).

The bay window is currently occupied by a large plant, and slatted blinds; I imagine this will have a negligible effect, and then only some diffusion at the highest frequencies? But also, the glass will already be limited in its ability to actually reflect the bass back into the room?

Here's a chart of that lower end, comparing the room before and after the panels.

(Edit: swapped the image for the correct one)

 

[AnalogSteph]

Here's a chart of that lower end, comparing the room before and after the panels.

With looks substantially better-behaved than without, but the whole ~80 Hz to 400 Hz range could use about a 5 dB boost now (and the ~55 Hz fundamental mode some attenuation). Really odd. I find it hard to believe that the panels would only work well in this range (were they originally used just like this or with more higher-frequency absorption on top?). Again, it would be good to see some nearfield measurements to see what the speakers themselves are doing.

 

[ernestcarl]

According A. Grimani, he sometimes uses diffusors at the first reflection points:

Lots of acoustics discussion topics with him are posted at the AVPro channel, which I recommend if you have the time — his talks are all over 1 hour — well, 3 hours in one particular video, I believe.

 

[Hipper]

That 55Hz peak is around 20dB above the average. Your earlier spectral decay plots show that it is slow to decay. As it is a wide peak it will likely add boominess to the sound. Is that what you hear? You could download the Realtraps free Test Tone CD to help you:

https://realtraps.com/test-cd.htm

DSP/EQ could easily reduce this and by reducing the peak the decay times will also come down.

Generally it seems that wide peaks and dips are noticeable but very narrow ones are not - that's my experience too.

I'm not convinced about diffusion in a small room. I tried it on my side walls in a room 14' wide, so the diffusors were about 8' away from my ears yet I could hear their effects and it seemed to me that they messed up the details of the music. It seems they need a certain distance to perform properly. According to Ethan Winer (of Realtraps) books etc. do nothing for diffusion:

http://realtraps.com/art_basics.htm

Go about halfway down for a discussion on diffusion, particularly the video 'All About Diffusion'. However this site is full of useful info and I recommend you look at some of it. His views on EQ make a fair point but if all you can do is EQ then that is better then nothing.

Regarding the bay window, I suggest you try panels there as a temporary experiment to hear what happens. Then you have to decide what's best. You can buy decorative panels - even ones that can use your images. Plants are thought to do something for diffusion but I've no experience of that.

 

[RayDunzl]

The problem: the room has, in a way, lost its 'soul'.

Idea:

This can be used to instantly add a little Soul to any room where it is lacking:

 

[pozz]

@mk2 In your reverb time graphs several curves show a depression around 500hz. This may be characteristic of the products you have placed in your room, but maybe not. It's hard to tell exactly what's what. It is somewhat suspect that your room SPL is generally flat down to 1kHz but shows unevenness below that. Strong absorption would help down to about 200Hz with standing waves.

My general perspective about acoustic treatment and room sound is this:

• Thin absorptive panels (anything below 6"-8", or the equivalent combination of a panel and air gap) tend to unbalance sound by losing effectiveness in the midrange and below. Foam is like that.
• Real bass traps (membrane or Helmholtz type) are very expensive and a lot are necessary for good effects.
• Take the cheapest option first.

The cheapest option is room EQ and speaker positioning. If the latter is more or less set because of room decor, that's fine. But EQ is important. All the suggestions here haven't been about tone control, but precise adjustment for room effects in the bass region. Once you figure out room EQ some, but not all, of the soul is likely to return.

Question: how many panels have you used in total? What area in sqft do they account for? How much of the back wall do they cover and how much of it is left exposed? Same question for the far side wall with panels. Also, what are the room dimensions H×L×W? What is the distance between you and speakers? The speakers and the front wall, side wall and bay window? What is the bass trap made out of and is the shape actually a triangle? Is it floor to ceiling? Pics of the listening room and treated walls would help too, if you don't mind.

The reason I ask is because the sense of envelopment is entirely to do with the mix of reverberance vs. direct speaker sound.

Your goal is clear: remove flutter echo, keep the sense of envelopment and crisp imaging.

Provisionally, which is to say without having all the necessary info, the solution will be some adjustment to room treatment, like using fewer but thicker panels, room decor permitting.

 

[mk2]

Hi, thanks for the tips everyone. I spent yesterday evening tinkering. I picked on understanding the peak that is centered around 55Hz and if it's at the root of why I feel the room is lacking life.

It's not something that I'll say has ever been an irritation, but using a tone generator and you can hear it so let's investigate.

First I measured each speaker in isolation and from 8" distance to check everything is working correctly, which it is. See attached.

So I reversed the roles, and put a single speaker on the sofa at the listening position, and then walked around the room listening to a 55Hz tone to see where it's loud and where it's quiet (I'm not a professional in acoustics or anything so I'm just inventing here and learning as I go. Someone can tell me if I'm doing it wrong)

What I heard was an increase along the front and back wall of the listening area, and at any height in the room. So if I'm right that means it would be a function of the front-back dimension of the room, and happens from both speakers. Additional SPL measurements seem to confirm this.

But roughly I can't work out the maths for that, because that specific room dimension is 348cm (half a wavelength at 49Hz). Wavelength of 55Hz = 624cm. So there isn't some kind of clear relationship to me. No dimension of the room is that small. And actually, maybe you can see the 'other' peak more minor at 49Hz.

I did consider whether it could be a function of the distance to the ground, as I assume the carpet+floorboards won't absorb or reflect any of these frequencies, but the clay ground is about 1m below floor level.

It's also possible the speaker designers know something here. A sponge in the bass ports (front ported) removes about 5dB of our peak fairly cleanly and without affecting the rest of the spectrum.

So after many measurements, I found it was best minimised using a combination of both moving the speakers far forward (this is more practical than moving the listener) and a sponge in each bass port. Having the speakers here might not work in the long term, but it's fine while I experiment.

Does this give more 'soul' to the room. Unfortunately I don't think it's the variable I'm looking for. The experience is clearer, somewhat less full sound than I've become used to; probably more correct though. But it's still a sound occupying a fixed piece of space right in front of me. So I think it's an improvement on one axis, but not the axis I'm looking for.

If there has been a increased sense of "envelopment" though (good word, pozz), then that's just come from the extra width of sitting closer to the speakers. If we keep pushing this there'll be a point where maybe I should just get some headphones

I've written a lot here, so I'll try and be back with answers to the specific questions, and maybe some photographs if you're lucky Thanks for the interest so far.

 

[ernestcarl]

Hi, thanks for the tips everyone. I spent yesterday evening tinkering. I picked on understanding the peak that is centered around 55Hz and if it's at the root of why I feel the room is lacking life.

It's not something that I'll say has ever been an irritation, but using a tone generator and you can hear it so let's investigate.

First I measured each speaker in isolation and from 8" distance to check everything is working correctly, which it is. See attached.

So I reversed the roles, and put a single speaker on the sofa at the listening position, and then walked around the room listening to a 55Hz tone to see where it's loud and where it's quiet (I'm not a professional in acoustics or anything so I'm just inventing here and learning as I go. Someone can tell me if I'm doing it wrong)

What I heard was an increase along the front and back wall of the listening area, and at any height in the room. So if I'm right that means it would be a function of the front-back dimension of the room, and happens from both speakers. Additional SPL measurements seem to confirm this.

But roughly I can't work out the maths for that, because that specific room dimension is 348cm (half a wavelength at 49Hz). Wavelength of 55Hz = 624cm. So there isn't some kind of clear relationship to me. No dimension of the room is that small. And actually, maybe you can see the 'other' peak more minor at 49Hz.

I did consider whether it could be a function of the distance to the ground, as I assume the carpet+floorboards won't absorb or reflect any of these frequencies, but the clay ground is about 1m below floor level.

It's also possible the speaker designers know something here. A sponge in the bass ports (front ported) removes about 5dB of our peak fairly cleanly and without affecting the rest of the spectrum.

So after many measurements, I found it was best minimised using a combination of both moving the speakers far forward (this is more practical than moving the listener) and a sponge in each bass port. Having the speakers here might not work in the long term, but it's fine while I experiment.

Does this give more 'soul' to the room. Unfortunately I don't think it's the variable I'm looking for. The experience is clearer, somewhat less full sound than I've become used to; probably more correct though. But it's still a sound occupying a fixed piece of space right in front of me. So I think it's an improvement on one axis, but not the axis I'm looking for.

If there has been a increased sense of "envelopment" though (good word, pozz), then that's just come from the extra width of sitting closer to the speakers. If we keep pushing this there'll be a point where maybe I should just get some headphones

I've written a lot here, so I'll try and be back with answers to the specific questions, and maybe some photographs if you're lucky Thanks for the interest so far.

I'd rather just ditch the word 'soul' as I don't think it's helpful at all in conveying what you find lacking in your setup.

Do you have a reference of what it is that you consider to be especially good or desirable?

In most cases, it is your room and placement that's going to be the biggest issue. Not having a dedicated listening space will be the biggest compromise there is no matter the calibre of your gear -- which is precisely why a lot of folks opt for headphones.

You said moving your speakers closer (increasing direct vs reflected sound) and using port plugs (decreasing bass peak) improved things -- that's no big surprise.

There is another way of increasing the sense of envelopment other than by changing your speakers/room, and/or adding diffusors or DSP surround effects... and that is by upmixing to multi-channel -- I mean, your room layout and placement is still going to matter but it'll definitely increase the sound-field (making the experience more 'lively'), for sure.

 

[North_Sky]

Bad room dimensions: 24' x 12' x 8' (multiple of 4, 6 and 8...12).
* Yours is 24' by 12' by 9'

Fix the room's dimension and restore some scientific soul to it.
24' by 12' ... try to short the room's length by say 3 feet.
And, sitting near the back wall; not a good idea.

Cheers,

 

[mk2]

Just to follow up this and now a few weeks has passed and I got to actually listen to the music I enjoy. I keep converging on everything as I started but with a sock stuffed in the bass port of the speakers. That was the single biggest improvement. I suppose low frequencies may have had some masking effect on the higher ones; or its given some improvement in performance to the speakers that helps with 'envelopment'.

Moving the speakers or listening position tended to make the frequency curve more erratic and so felt to be counter productive in the end, despite lots of walking around the room with a spectrum analyser trying to find a better positioning. Surprisingly the flattest graphs come when both the speakers and listener are quite close to the front and back walls.

The peak at 55Hz is not really a serious problem, never mind one which can be practically addressed. It adds a bit of presence of the low end especially when listening volumes are not particularly loud here (it's a terraced house). I'm sure some DSP would be great, but that's a bit too much like my day job and I like the relative simplicity of this setup. No, I won't be changing the size of the room or moving house. Thanks for all suggestions and contributions.