Home Theater Forum and Systems banner

Measuring bass trap effectiveness

9K views 31 replies 9 participants last post by  irombeach 
#1 · (Edited)
I couldn't find any threads were people tried to measure bass trap effectiveness (no bass trap then add bass trap etc) so I did a simple experiment.

I measured the reponse with an untreated room, then added one floor to ceiling superduper chunk column in the right rear, then added another superduper chunk column in the left rear and finally added a regular super chunk nine feet long along the front floor/wall boundary. After each addition I remeasured. The superduper chunks were humongous at 34"x24"x24" and the regular super chunk was the normal 24"x17"x17". So I didn't skimp on the material!

Here's what I found:

Baseline:
Purple Violet Plot


Right rear column added:
Blue Tire Pattern Fence Metal


Left rear column added:
Green Leaf Grass Plant


Front floor chunk added:
Violet Purple Magenta Pattern


I note the following:

1) Below 40hz there doesn't seem to be much effect.

2) Surprisingly the 40-50Hz region improved with each superduper chunk. Presumably reaching this low is due to using the bigger chunk size.

I was planning to cut these in half to regular size but now wonder which is better...larger chunks but only two columns, or regular size chunks which gives two extra column worth of material to use elsewhere. Any advice?

3) Adding the second rear column changed things as did adding the front floor column but did it get better? I'm not sure. Certainly there isn't a steady march to filling in the peaks/valleys...each one clearly better than the last...! In fact the first column seemed to smooth things the most. What is happening? Pehaps reflections from walls are starting to dominate that aren't being addressed by the corner chunks. Once again, comments from the cognoscenti appreciated!

4) Changes in the 80hz peak is interesting. It improved the most with only the right rear corner. Then regressed although in all cases it is better than untreated.

5) Some pretty deep nulls seemed to be introduced! Complete cancellation at some points where previously there was some energy.


Overall, to be honest, I don't see the improvements I had hoped. Especially given the huge size of these things in the corners. Since this was a test, I'm not going to necessarily keep it this way. But, based on these plots, it isn't so obvious what would be the best deployment of all this stuff either... :eek:
 
See less See more
4
#2 ·
The decay time down low has definitely been effective.

There is some smoothing of response but not huge. This is a very good example of the fact that while corners are a very efficient place to do broadband bass control, it's not the only place and isn't a magic bullet. Some of the nulls you're getting that really didn't change much can be caused by a variety of things such as SBIR, null off the rear wall, main to sub integration/phase issue, etc.

The nulls that got deeper are interesting. What's happening is that the treatments are capturing a reflection that was previously cancelling a problem so to speak.

Looking at the baseline, I'd say there's a bit of work to do with speaker, sub, and seating position before treatment is applied. I'm sure we can get a better starting point and then begin to address what's left, what's causing it, etc.

Bryan
 
#4 · (Edited)
....now wonder which is better...larger chunks but only two columns, or regular size chunks which gives two extra column worth of material to use elsewhere...
Any comments regarding this trade-off?


...Looking at the baseline, I'd say there's a bit of work to do with speaker, sub, and seating position before treatment is applied. I'm sure we can get a better starting point and then begin to address what's left, what's causing it, etc....
It does appear to require an iterative process alright.

Do you always play with speaker/sitting position BEFORE any treatment? From this expirement it seems simple treatment can expose latent problems. So doing the basics (chunk corners at least since that is easy) might be a better starting spot?

In any case, I certainly agree. The baseline isn't the best starting point for this room!
 
#5 ·
I always recommend doing as much as you can with locations of seating/speaker/subs prior to doing any treatment.

If it were me, I'd do twice as many of the smaller ones. They're still large enough to be effective down pretty low.

Bryan
 
#8 ·
Hehe...there isn't much to see...since it was just a test all I did was stack the raw fibreglass chunks. I didn't actually build anything. Anyway, if my daughter ever returns my camera I'll snap a picture.

Same for the sound, I was testing if one could measure the addition of the bass traps and how they would manifest themselves in the plots. I wasn't too worried about the sound since it wasn't representative of a real listening setup. Having said that, the room sounds quieter and less "slap happy" with these pieces added.
 
#10 ·
Well...as mentioned it was primarily a measurement test not a listening test...but recognizing that, I'd say the audible difference was minor. Moving the sub or listening position makes a much bigger difference.

The test basically covered 25% of the dihedral corners and nothing else so perhaps one shouldn't expect much.
 
#11 ·
I went from 70% corner coverage to 100% and the difference was dramatic. Also alot of difference going from 0 to 70, but less 'in your face'. I sure noticed when I removed the traps, though! :) So easy to get used to that sound and tight bass, that you really don't notice until you no longer have it.
 
G
#12 · (Edited)
Just a comment that may diverge a bit from a few common assumptions. (in the hope that it may spur a bit of curiosity and further investigation.)

While porous (absorbent) bass traps can be effective to a small degree, they are not very efficient.
The reason is that 'particle velocity' of a sound wave (yes, I know it behaves as a compression wave at modal frequencies)near the walls and in corners is essentially zero for these long wavelengths. Additionally, their required depths to be effective at these wavelengths are prohibitive in a small acoustic space.

In fact, a porous/absorbent panel, assuming a 45 degree incidence would require a thickness of:

at 60 Hz 45 degree incidence: 3.33 feet 90 degree incidence: 4.71 feet

at 120 Hz 45 degree incidence: 1.60 feet 90 degree incidence: 2.26 feet

And that still assumes a high particle velocity!


A more efficient alternative absorber design, although a 'bit' more complex to calculate and design is a resonant membrane absorber (and its perforated variant). (For much deeper source detail into this topic, one might want to pursue the research of D.Y Maa)

The membrane converts pressure fluctuations into air motion. As the membrane sympathetically vibrates over a selective low frequency range, determined by its mass and the air spring compliance, it pushes air through an internal porous layer resulting in low frequency absorption. Basic relationships exist between the effective frequency and the membrane mass, stiffness and cavity depth.

Although beyond the scope of this entry, the design and construction of such an absorber is not overly complex assuming one has a bit of patience.

Overly simplified relationships are stated below, but their limitations should be noted, as for small areas the entire membrane may not vibrate freely due to edge mounting friction, and bending stiffness may tend to increase the resonant frequency.

f for normal (90 degree) incidence = 60/sqrt(md) for an empty cavity, and f= 50/sqrt(md) for a cavity with porous absorption.
f for oblique incidence = 60/{cos a sqrt(md)}

There is a bit more to account for real world variables before one run out and begins to model and build these absorbers! And in addition, one might find that porous membrane resonant panel absorbers are just a mite more predictable. But ala in all, the variations on the fundamental design are much more efficient at mitigating low frequency modal frequencies and they are also of benefit in that they do not absorb broadband specular reflections and thus avoid over-deadening the small acoustical space as is so common with the widespread use of absorption.

So...now you have an introduction to yet another more efficient variant of low frequency modal absorbers compared to the porous absorptive method that also exhibit the benefit of not being overly absorbent to the mid and high frequency energy that is comprised of focused specular reflections in the small acoustical space. And as such, they help avoid the all too common overly deadened room.
 
#13 ·
Agreed. Porous membranes can be extremely effective.

Just to be clear - the calculations shown for thickness of standard porous absprbers are for 'optimal' absorption. This should not be construed as not being effective at those frequencies and lower if not that thick.

A 6" thick panel straddling a corner or a 17x17x24" chunk style absorber filling a corner can be pretty effective well below 60Hz. A 6" thick absorber spaced 2" off of a wall surface can be pretty effective at and below 60Hz. It's just not as effective as it could be until it reaches the 1/4 wavelength of the bottom frequency to be considered.

Bryan
 
G
#15 ·
The behavior of porous absorption (eg, foam, rockwool, fiberglass,etc.) is very well understood.

When sound propagates in small spaces, such as the interconnected pores of a porous material, energy is lost. This is primarily due to viscous boundary layer effects. Air is a viscous fluid, and consequently sound energy is dissipated via friction with the pore walls. In addition to viscous effects, there will also be losses due to thermal conduction.

For the porous absorber to create significant absorption, it needs to be placed where the sound particle velocity is high. The particle velocity close to a boundary is usually small, so the portions of the absorber close to the boundary are not generating much absorption. It is the parts furthest from the backing surface which are most effective, which is why thicker layers of absorbent material are needed to absorb low frequencies.

For low frequencies, where the wavelength is large, one has to go a considerable distance from the wall to reach a point where the particle velocity is significant. (D'Antonio&Cox)

Hence this is why absorbent material can be placed away from the wall and still be effective, as it is being placed in the region of greater particle velocity. It is also why such companies as RPG have developed sinusoidally shaped panels (such as ProForm) which place material into this zone while providing attachment points (regions which are minimally effective).

A rough figure frequently quoted is that the absorbent material needs to be at least a tenth of the wavelength to cause significant absorption ( U. Ingard, Notes on Sound Absorbent Technology, 1994 ), and a quarter wavelength to absorb all of the incident sound energy, assuming the complete match in the acoustical acoustical impedance of the absorbent material.

Thus per the minimum quoted tenth of a wavelength (...and this assumes an ideal 100% absorptive acoustical impedance match):

at 60 Hz 90 degree normal incidence: 22.6 inches

at 120 Hz 90 degree normal incidence: 10.85 inches

...Still potentially prohibitive for many in an average small acoustic space.

The need for significant thickness compared to wavelength makes porous absorbers inefficient and not particularly useful at low frequencies. While they can be used if the associated trade offs are acceptable, it is why you see few of them in the product lines of the professional suppliers such as RPG.

Thus, this is part of the reason for not necessarily seeing huge changes in the CSD/waterfall plots by simply adding absorbent corner traps. They are relatively easy to build and to install, but one should be aware of their limitations.

And as such, much more effective and efficient absorbers are to be found in the resonant absorber design family, as they provide for much greater absorption for a given depth than porous absorbers.

This is not intended to scare anyone off from the use of absorbent bass traps.

But the behavior of such techniques are well understood.

The most significant variable involves the material itself, in the form of its acoustical impedance, which can vary widely. And few materials even come close to being 100% absorbent. In fact, the topic of acoustical impedance and its effects, just like in electrical impedance in the effective termination (absorption), reflection, transmission, diffraction, refraction, phase of reflection, etc. of acoustical waves,is fundamental and would be another worthwhile topic for a thread that should be considered prerequisite to the study of room treatment as it is completely predictive of the behavior of the material and technique. And I am already behind in another post!

Rather, having a bit more actual information on how they work and what determines their effectiveness allows one to make more informed choices. With a bit more information one can reasonably predict the effectiveness and determine the suitability of a particular technique in advance and hopefully choose the methodology that will most optimally compliment your needs, as well as your time, labor and pocketbook

...You pays your money and makes your choice...:bigsmile:
 
#16 · (Edited)
...the topic of acoustical impedance and its effects, just like in electrical impedance in the effective termination (absorption), reflection, transmission, diffraction, refraction, phase of reflection, etc. of acoustical waves,is fundamental and would be another worthwhile topic for a thread that should be considered prerequisite to the study of room treatment as it is completely predictive of the behavior of the material and technique. And I am already behind in another post!...
Sounds interesting when you have time to start the thread!


mas said:
Rather, having a bit more actual information on how they work and what determines their effectiveness allows one to make more informed choices. With a bit more information one can reasonably predict the effectiveness and determine the suitability of a particular technique in advance and hopefully choose the methodology that will most optimally compliment your needs, as well as your time, labor and pocketbook...
Since I have no experience in this area I've been spending some time measuring and playing with my acoustic panels to develop a feel for all this.

So far my impressions are that broadband porous absorbers such as my rockwool (either panels or chunks) are quite ineffective in absorbing reflections below 100hz (there is some action but it isn't enough to solve a problem if it exists at those frequencies). I'm also finding it marginally effective between 100-200hz (it does seem to make a difference...but not large...and one needs a lot of the stuff). Above say 500Hz it seems quite effective (a 4" panel blocking first reflections seems to reduce reflections in the ETC curve 10-12 dB without any trouble.)

There may be benefits of the absorbers that I'm not able to measure or am overlooking but my initial conclusions are that porous absorbers (ie rockwool) works well for that initial time delay gap and not so well for anything else.

On the other hand they are easy to build, relatively inexpensive, well documented for the hobbiest and completely within the range of a DIY'er. One can also use quantity to compensate for (lack of) effectiveness to some extent.

Resonant membrane absorbers sound interesting but a search of the various forums for proven DIY construction techniques produced nothing. And even here this is little point unless it can materially address room modes in the 40Hz to 120Hz range. Can one even cover a couple octaves effectively? Without practical implementations for the DIY'er they are relegated to interesting theory only.

After spending an afternoon with various woofer and listening positions, I have little hope for seeing a smooth low frequency curve for one seating position, let alone two or three. There is just too much uncontrolled stuff going on at those frequencies and most of it is bad... :thud:
 
#17 ·
Full, pure, non-porous membranes are much narrower in absorption range than the porous membranes or standard porous absorbers that were mentioned before. They can definitely have their place though.

As for the full porous absorbers, how effective they are is a function of impedance (gas flow resistivity), thickness, and distance from leading edge to boundary. The chunks you made are certainly thick enough to reach lower than 100Hz relatively effectively (not 100% though). You also have to remember that not all frequency related issues are going to be solved with corner placement.

If you're up for another experiment, try making 2 6" thick panels to hang centered on the rear wall behind the listening position - and maybe try them with and without a 2" gap behind them. To see which frequencies might be impacted, measure with them not installed at the current position, then again say 1' forward or backward so you can identify potential issues related mostly to the length dimension. Then install the panels and take a look.

Bryan
 
#20 ·
Thanks Bryan. I did the test as you suggested with a couple comments.

1) My panels are 4" thick so I used two each to double their thickness (ie 8" thick panels).

2) My rear wall (behind listening position) is actually a large patio sliding door to the outside. About 6' wide x 7' high in size so it is most of the height and approaching half the width of the rear wall.

Anyway, here are the results.
Text Plot Line Slope Wave

These are the three mic positions varied along the room's longitudinal axis. The brown middle trace is the listening position, the purple top trace is 0.5' rearward, and the bottom is 1' forward. All with no panels.

I then chose the 1' forward position (this meant I didn't have to move the mic) and added the 8" panels against the rear wall (ie patio door) with this result:
Text Blue White Line Pattern


There isn't much difference. :hissyfit:

Are we having fun yet? The real world is so cruel... LOL
 
#21 ·
Nope - not a lot. BUT, that position is MUCH better at removing a couple of the deeper nulls than the current position or the more forward position. We did pick up about 3db at 80 and about 8db at 150'ish though so it's progress still.

So, what does that tell us? It tells us what is impacted by seating position. Now, you need to play with speaker and sub position with those panels gone again and see what you can do to smooth things. Once you get it better, try a couple of those 4" panels directly behind the speakers.

You'll also need to work with subwoofer position and then work back and forth between subs and mains using phase, slope, and frequency to blend the best you can. My suspicion is that part of the big dip at 80Hz could easily be a phase anomoly between mains and sub if that's the xover point.

This is just an iterative process but well worth it in the long run. Also, did you try looking at the before and after waterfalls after adding these 2 8" thick panels? You should see some additional improvement in decay time.

Bryan
 
#22 ·
...My suspicion is that part of the big dip at 80Hz could easily be a phase anomoly between mains and sub if that's the xover point...
In this test I'm just running one subwoofer with xover bypassed. Nothing else. The subwoofer's quasi-anechoic response is smooth to past 200 hz so all the pertubations are due to room effects.


Also, did you try looking at the before and after waterfalls after adding these 2 8" thick panels? You should see some additional improvement in decay time.
Yes, I looked at these also. There was no difference.

I started this thread to examine objectively whether broadband absorption panels does something worthwhile at the lower frequencies (say below 100Hz) and at this point I have to conclude they don't. If someone has good evidence otherwise or can suggest better tests then let me know.
 
#23 ·
Well, that's your choice. I've seen way too many times where it absolutely positively does make a difference. Gotcha on the sub only - that's just a very common thing to see in real setups where someone just guesses on phase.

I have a hard time believing that there was NO difference adding 8 sq ft of 8" thick material to the room - sorry. I just went back through and looked - how big is this room you're experimenting in? Is it closed off or open to another area? I'll agree that broadband isn't the do-all, be-all, fix-all solution to all room problems - it isn't - but it has it's place in pretty much every room as a portion of the solution.

You yourself saw the changes when you started adding things to the room initially as you posted in your graphs. No, it might not be huge, but again, just like the moving things around, if we can address a bit of the problem, then we're making progress.

If broadband absorption isn't going to fix your FR problems, then targeted absorption isn't going to either. As you said though, this is just an experiment and there has been no listening going on. I suspect that if you actually listened to familiar source material before and after, you'd hear a difference. I could be wrong though.

Bryan
 
#24 · (Edited)
Thinking about your comments Bryan, I realized my last test had a flaw in the test methodology. The baseline "no panel" curves were contaminated by the fact the panels were in the room with me (just lying against the side wall) when really they should have been completely out of the room for a proper "before and after" test.

I redid the test accordingly. I also did some minor experimenting with placing the panels against the rear wall. The changes in the 70-80Hz are more noticeable this time. I then extended the test by adding two more panels in the front left corner (blocking two windows that are right in the corner). These two extra panels noticeably improved the dip in the 100-200Hz region (had no effect on the 70-80Hz region). A plot showing the final results is attached:
Green Text White Blue Line


Question: Why don't people put FSK facing on the corner chunks? I understand FSK facing helps the low frequency absorption as well has protects from excessive HF absorption so it seems like a good thing to do...

My room is about 2000 cu feet (14x17x8). It is closed, doesn't open to any other part of the house.

I think I'm about done with the tests for now. Special thanks to you and Mas for your insights and contributions. Very much appreciated! My overall summary in post #16 still seems about right to me. But I take your point that they are only part of the overall solution and every bit helps.

In any case I really need to stop getting side-tracked and actually start setting up my room!
 
#25 ·
Yeah - I'd say. That's a 10db improvement at 80Hz - certainly nothing to sneeze at. Another 10db at 150'ish. There's still a very broad dip 50-85'ish. Hard to say with just a sub and not having played a lot with sub positioning - but that wasn't the point of this experiment. It does, however, show how important placement is. You can treat and EQ the room/system till you're blue in the face but if seating, speaker, and sub placement are bad, you're still going to have problems.

The FSK adds a membrane effect. The 'increased bass absorbtion' is in a narrower band - not across the entire bass spectrum and is defined by a variety of variables. In a home theater, we WANT the front wall to be absorbant at higher frequencies. Now, in the rear corners, if they're required there, then FSK may be appropriate.

Bryan
 
#26 ·
This is a great THREAD.
pre/post pictures of spl , allowed me to understand something very important.
THANKS :clap:

Now, i've some simple questions because i don't understand some english words (sorry)

1) which are the differences from chunks and panels ?
2) how should we position the absorber? in front of the corner ?
3) what about panels? would they go on the wall?
4) Tube traps are equals tu chanks?

5) How can I test ceiling panels? can a person support it with hands? no problems with measurements?


thanks a lot

ale
 
#27 ·
1. A panel is typically a 2'x4' piece that is mounted straddling a corner to form a triangular cavity behind. A chunk is the same type of material cut into triangles and stacked in a corner to form a solid absorber with no cavity behind it. Better performance and takes less space in the room.

2. Many different places depending on what the room requires.

3. Some of them for reflections and to deal with the front wall - yes.

4. Tubes tend to need to be VERY large diameter to reach into the deep bass. Think 20" diameter or bigger.

5. What are you looking to find out about them?

Bryan
 
#28 ·
Hi Bryan, i cannot understand what do you mean in your question number five.

do you mean :
1) what i expect from ceiling panels?
2) where i buy them?


i'm talking about tube traps because i'm making 5 tubes . height 40'' diameter 16'' and i hope they will do something, because they are the simplest obejcts you can put and remove from your room.


other questions :

have chunks a wood frame inside?

how can i understand from spl graphs, which problems are due to ceiling or floor or from the walls?(from the time passed?)


many many thanks
ale
 
#29 ·
Sorry - misread it. Tacks or stick pins might work to temporarily hold them up.

You can do tubes that size and they can work just don't expect miracles at 40hz. fill them with insulation to improve things. What are you making the tubes from? Pipe insulation is usually the best bet - nothing hard or solid.

Chunks can be done in different ways depending on how permanent you want them.

To understand the graphs, you need to understand which modes are caused by which dimensions, which problems are not necessarily modal related, etc.

Bryan
 
#30 ·
ok Bryan,
the only thing to do is trying a lot.......

in this weekend i'll try moving mic horizontally as you suggested me.
and i'll look the results.

please tell me if this is a good approach :

looking my spl and ir graphs, it's seems that a lot of problems are "late" (20 , 30 , 40 ms) and so,
talking about axial reflections, they should be due to far walls.


i did measurements on the right channel....and the farest (very distant) wall is the big glass on the left.

So , moving mic to the left, i should see reflection earleir....

am i right?:foottap:
 
#31 ·
Maybe but don't get too hung up on that just yet. Some of those reflections are not ones which are going to impact bass frequency response. Let's address one thing at a time.

Bryan
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top