Boomy bass -- is wooden floor the problem?

[Diapason]

Hi folks,

I have a dedicated room for 2-channel that's giving me no end of annoyance. To start, the dimensions are awkward: 21' x 9'4" x 8' with the speakers firing down the length of the room. I sit roughly 38% from the rear wall, speakers are about 5-6 feet from the front wall. Of course, they have to go quite near the side walls to get any kind of stereo image. Construction is dry-line over concrete on all 4 sides, with a fairly sizeable window in the middle of the front wall. The door into the room is behind me on the left long wall.

Initially, the room was crazily reflective and very woolly, with slap and flutter echoes to beat the band! An assortment of panels from GIK has worked wonders in this regard, the room now sounding much tighter across the whole frequency range. To my ears at least, bass peaks and nulls have been much improved, and most things sound pretty satisfying.

However, I'm left with one bass region that I can't seem to tame. Most of the other frequencies I can adjust by speaker/seat/panel positioning, but one fairly narrow bass frequency range is just wild. Now, I can't tell you in Hz what it is, but in a track with "walking bass", one particular note will just jump out at you, while the notes either side will be relatively okay. It gets to the stage that the ear is drawn so much to that one note that you hear it over everything else. Very frustrating.

Now, what I'm wondering is, could this be the floor, or is it more likely to just be a node I haven't yet addressed. The floor is a laminate wood floor, probably over concrete, and possibly raised on battens with some kind of underlay. It can be "springy" in places, and when music is playing you can feel the floor's vibration fairly seriously through your feet. I have a big thick rug on it, but it obviously does nothing for the vibration. My speakers (PMC OB-1's) stand on Nordost Pulsar points.

Could this boom be at the resonant frequency of the floor? Can you suggest any tests that I could do before I go ripping up the floorboards in a fit of madness? While we're on the subject, if I put carpet directly onto the concrete, would that cause any other problems?

All suggestions gratefully received. If it helps, I can try to isolate the problem frequency later today. I think I have a disc that increases in 1Hz increments somewhere.

Thanks,

Si

 

[Ethan Winer]

> Could this boom be at the resonant frequency of the floor? <

Maybe, but not likely. A flexing floor often gives some amount of bass absorption. You probably just need more bass trapping.

--Ethan

 

[Glenn Kuras]

Hey Si,

I have to agree with Ethan on this. If you would like give me a call and lets try to track down what the heck is going on.

Thanks,
Glenn

 

[j_garcia]

Sounds like a major peak. You should really do a response curve of your room to find out exactly where it is.

 

[goodman]

Have you tried removing the Nordost points?

 

[Diapason]

So I did a bit of further examination last night using a Cara CD that someone loaned me. The problem seems to be centred squarely on 80Hz. At that frequency, everything in the room seems to vibrate, floors, door frames, my skull! Walking the length of the room exhibits severe peaks and dips, with the peak quite near my listening position. Keeping my ears at tweeter height, the problem is even more pronounced and the characteristic peaks and dips don't seem so apparent when moving up and down the room at this height. Basically it's awful pretty much everywhere!

So, it would seem to be a room (rather than floorboards) issue. I'm just not sure I can put enough bass trapping in to sort this out and still leave room for a seat. I have a sound meter somewhere, I'll try to dig it out and detail the extent of the problem.

Regarding the pulsar points, I haven't tried removing them recently. I just found them a stable way to put the speakers on the floor and they tightened up the sound a little. I may try removing them just to see what happens, but I'm not really expecting much!

Glenn, I'll definitely give you a call. I must admit, the room is sounding much much better now that the GIKs are in there. It's really just this one problem frequency remaining. By the way, I thought you were on vacation on some paradise island??

Si

 

[sploo]

Just a thought, if you've got a particularly bad peak at a single frequency (something like ETF may be useful in getting a good reading of your room) then perhaps a suitable Helmholtz resonator would do the job?

http://www.audioholics.com/techtips/roomacoustics/Helmholtzresonatorabsorber.php

 

[j_garcia]

What x-over point are you using? Maybe try shifting it lower?

 

[bpape]

Or try shifting your listening position about 6" to one side of center in the width dimension. People pay attention to the length and forget about the width.

If it turns out to be height dependent, a 4" panel directly over the listening position can also help.

Bryan

 

[Savant]

Or try shifting your listening position about 6" to one side of center in the width dimension. People pay attention to the length and forget about the width.

If it turns out to be height dependent, a 4" panel directly over the listening position can also help.

Bryan

That might work...but...

The {3,0,0} mode (axial along the length) is 80.7 Hz. This would be my guess as to the main cause of the problem. When the room starts resonating at this frequency, everything else will start vibrating right along with it. This sort of thing is not uncommon. There is also a tangential mode {2,1,0} at 81 Hz, which is probably being excited at the same time, exacerbating the problem.

Note: Using CARA, I roughly modeled your room. I have attached a good graphical representation of the spatial sound pressure level response - at ear height - for 80 Hz based on the information you've provided. The "balloons" show the rough locations of your loudspeakers (blue) and the listener (yellow) in the room. The x- and y-scales are the room dimensions in meters and the z-scale is a relative SPL scale for ear height in the room. As should be obvious, the listening location corresponds to a significant 80 Hz SPL peak.

I would concur with sploo - a good tuned absorber would help take care of the problem. I would suggest one of the perforated variety. This might help take the edge off that frequency. If the resonance is that bad, you will probably need quite a number of small-to-medium sized absorbers, or several large absorbers. I usually suggest building a few small ones, placing, evaluating, building some more, etc. - an iterative process.

You might also consider:

- Moving the loudspeakers closer to the wall in 6" increments. According to my calculations, you should be able to find a spot somewhere around 3.5' from the front wall that won't excite the 80 Hz mode at all. Of course, don't bank on this - it is theoretical. But the same calcs showed a strong spike at 80 Hz, right where you've indicated there is a problem. And you may discover that 54 Hz suddenly gives you trouble...

- Spikes will enhance any resonance occurring between the loudspeaker and the structure. If this sort of structural resonance coincides with a room mode, thing will get worse. IMO, lose the spikes. (Please note: I'm not a fan of spikes.)

- Experimenting with your GIK panels could also improve things. Changing the airspace depth behind a panel (up to the thickness of the panel) could improve their performance in the 80 Hz range.

- You could also modify some of the GIK panels - or purchase more of them - by (re)facing them with the right type of perforated board (as alluded to above). This way, they can be tuned to the desired 80 Hz range.

If you're literate with Excel, there's a good spreadsheet here ("Porous Absorber Calculator") that can help you calculate the right design of perforated board. (And keep in mind that there is more than one design possible.)

 

[Diapason]

Gentlemen,

First of all, sorry for my tardy response -- it was a holiday weekend here and I'm still catching up!

I appreciate the time and effort put in by everyone here to come up with suggestions. Jeff, that's over and above the call of duty, and I'm especially grateful. You just might stop me from going mad altogether. I must admit I'm beginning to think that maybe I should just sell the system and buy a pool table. Not really of course, but...

Now, regarding the suggestion to change the crossover point, that's a no-can-do I'm afraid. This is purely a 2-channel system with no sub.

The lateral shift of the listening position does little to help in this case. I have suffered from problems in the past that this seemed to improve, but it doesn't seem to work here.

I should point out of course that my listening position and speaker positions are 100% flexible so I can try anything. Moving the loudspeaker closer to the wall did indeed give a different signature, and the 80Hz peak was somewhat ameliorated. The overall effect was unpleasant, however, and peaks seemed to appear elsewhere, so I've started to move them back out again. I'm still experimenting here, but I just can't seem to find the right spot for them.

I started to measure the room in 1Hz intervals from 20Hz up. My approach was pretty haphazard, as I just balanced the radioshack SPL meter on the back of my listening chair close to where my head would usually be. Nonetheless, I was pretty gobsmacked by what I found. I started by setting the volume so the pink noise track was around 75dB. Without looking up the specifics (which I have cleverly left at home) I basically measured a response curve varying between 65dB and 95dB in the range from 20Hz to 120Hz, with a couple of peaks and dips along the way. I was disturbed before I finished, and I conducted this test in a position I wasn't happy with anyway, with current sound treatments incorrectly placed, so I intend to do it again. Even so, that seems like a pretty serious peak to dip ratio to me. I'm starting to worry that this room is unsalvageable!

My concern at this point is that further treatment is going to be
a) Really huge
b) Really expensive
c) Ineffective anyway

Do helmholtz resonators work well? Can I really tame such massive peaks? Can I put enough broadband absorption in here to sort the issues out? Will it impact on the mid and high frequencies (which are already starting to sound bright and thin)?

I'll post the results of a full frequency sweep later this week if I can. Thanks again for all the help.

Si

 

[Glenn Kuras]

If you are going to shoot the room you really should look at a program like ETF. Doing the way you are going about it will work, but really you need to several tests (think more like 10) and ave them. The nice part about ETF is you can shoot the room in about 5 seconds.
Like I said at the start though, give us call and lets figure out how to maybe better place the panels. It may come down to getting more bass traps but would like to talk about it first.

Glenn

 

[Savant]

I started to measure the room in 1Hz intervals from 20Hz up. My approach was pretty haphazard, as I just balanced the radioshack SPL meter on the back of my listening chair close to where my head would usually be. Nonetheless, I was pretty gobsmacked by what I found. I started by setting the volume so the pink noise track was around 75dB. Without looking up the specifics (which I have cleverly left at home) I basically measured a response curve varying between 65dB and 95dB in the range from 20Hz to 120Hz, with a couple of peaks and dips along the way. I was disturbed before I finished, and I conducted this test in a position I wasn't happy with anyway, with current sound treatments incorrectly placed, so I intend to do it again. Even so, that seems like a pretty serious peak to dip ratio to me. I'm starting to worry that this room is unsalvageable!

Just one question/comment: What weighting network was selected on the RS SLM; A or C? If it was A, then the numbers are - unfortunately - quite difficult to interpret. It would seem it's more likely that C was used based on the numbers you've given. If so, you should note that, while this is the "correct" setting to use for this application, there is some LF roll-off below 60 Hz with C-weighting that needs to be accounted for. This should also be taken into consideration if the RS meter is being used as the microphone for a system like ETF. See this page for more specifics.

As for measuring room acoustics in general, ETF provides, shall we say, more thoroughness. Did you also mention that you're using CARA? If so, isn't there an analysis feature with it? Could be another option if you don't have ETF. (Though I've never used CARA for room analysis - only design.)

 

[Diapason]

Hi Jeff,

No I don't have Cara, a friend loaned me the disc of test tones to play around with. I've since burned a disc of 1Hz intervals from 10-299Hz that I found online somewhere. I was using the "C" weighting, "slow" on the meter, and when (if?) I do it again I'll take into account the roll-off. Thanks for that.

I suppose I'm trying to get good sound without the hassle of having to measure the room properly etc. etc. Lazy I know! The truth is, and you may be amazed by this, I don't have a PC at home. I sit in front of one all day, and I have always resisted getting one for home use. I'm gradually being worn down, though. Of course, I could borrow a laptop if I wanted to do the ETF tests.

Having spoken to Glenn yesterday, I have a few more things to try. Glenn was his usual helpful self, and was confident we can sort things out. I'll report back.

One thing that's been "interesting" with all this is the lack of support from my hifi buddies over here. It's made me realise just how misunderstood is the room's influence on sound. As it stands, they think I'm nuts. I can't quite explain to them that the problems aren't easily solved. Comments like:
a) Why don't you try new speakers?
b) just put up some paintings and that will sort out the sound.
c) your CD player is too old
d) it's not the room, you've just got the wrong system
are reasonably common. I'm starting to understand the frustration you guys must feel.

Thanks again for the help. I'll keep you posted on developments.

Si

 

[bpape]

Well, it was worth a shot. I was hoping that shifting the position sideways would get you out of the (2,1,0) intersection - sometimes it works, sometimes it doesn't. But hey, it's free, might as well try it.

Whether to use additional broadband bass treatments or specifically tuned ones is a decision to make. One of the things you can do to help you make that decision is to calculate or measure the decay time in the room from 20-300Hz. If it's still broadly high in terms of the target for your room size and usage, then broadband may be a better solution. If it's in the right range but just this one narrow band still has issues, then a tuned absorber may be a better way to go.

Bryan

 

[Ethan Winer]

Si,

a) Why don't you try new speakers?
b) just put up some paintings and that will sort out the sound.
c) your CD player is too old
d) it's not the room, you've just got the wrong system
are reasonably common. I'm starting to understand the frustration you guys must feel.

Tell me about it. And it's not just audiophiles either. I once called the editor of a leading magazine aimed at professional recording engineers, to ask about getting a product review. He said they have a policy of not reviewing any acoustic products because they all behave differently in different rooms.

So the loudspeakers and microphones they review endlessly month after month don't?

Clueless - utterly and hopelessly clueless.

--Ethan

 

[Glenn Kuras]

Si,



Tell me about it. And it's not just audiophiles either. I once called the editor of a leading magazine aimed at professional recording engineers, to ask about getting a product review. He said they have a policy of not reviewing any acoustic products because they all behave differently in different rooms.

So the loudspeakers and microphones they review endlessly month after month don't?

Clueless - utterly and hopelessly clueless.

--Ethan

You have to be kidding me!

Glenn

 

[Savant]

I suppose I'm trying to get good sound without the hassle of having to measure the room properly etc. etc. Lazy I know! The truth is, and you may be amazed by this, I don't have a PC at home. I sit in front of one all day, and I have always resisted getting one for home use. I'm gradually being worn down, though. Of course, I could borrow a laptop if I wanted to do the ETF tests.

There are other, non-PC, options. The first one that comes to mind is recording a tone sweep played back through your loudspeakers. If you can (a) record using an omni mic and (b) convert (or have a friend convert) the recorded file to a WAV, I might be able to run the analysis on my end for you. Let me know if you have that capability - and feel free to PM me so as not to burden others with our discussion of the details.

Another option is to purchase or borrow a stand-alone analyzer. Something like an Audio Toolbox. A PC is needed in the case of upgrading firmware or downloading saved data, but it's not essential. (You could always perform those tasks at work! )

A third option would be to hire someone to run the analysis for you. There are plenty of acoustical engineers, acousticians, and consultants out there that probably wouldn't charge you much money to simply run some tests in the room.

 

[Ethan Winer]

Glenn,

> You have to be kidding me! <

I wish I was kidding.

--Ethan