Tracking my acoustic treatments...

[affeking]

I've recently undertaken the task of building some bass traps and broadband absorbers. So far, I've completed 1 of the 'studiotips superchunk' bass traps and was anxious to see what differences I would see. Well, it wasn't much. It tamed one dip, but not at a frequency I was expecting. I don't see anything going on below 80Hz where I was most hoping to see some improvement.

I still have 2 more of these to install in the corners, followed by some additional trapping in strategic spots. However, just wanted to get some opinions on what I'm seeing so far.

Here is the graph prior to the treatment:
http://picasaweb.google.com/jrames/R...67083608357154

Here is the graph with this first trap in place:
http://picasaweb.google.com/jrames/R...67079313389842

You can see a dip at about 110hz was tamed noticably, but nothing really happened at any lower frequencies. I can't seem to find the results I'd seen earlier, but I was certain these traps were effective at <100 frequencies.

I might just be jumping the gun - this is only the first very small portion of my treatment, but I was just surprised that the low end did not budge at all.

Thoughts?

I'll post some more graphs as I move forward. By the way - is there a way to show my before and after plot on the same graph? I couldn't find the option to do so..

Thanks,
Jeff

[srckkmack]

Yes you can overlay your plots. I know you can generate waterfalls, then overlay one onto the other. You may actually find that your bass trap has affected the decay times. Anyway you can reference the help file to find out more; look up the waterfall section.

-Steve

[mike c]

how do we hear the improvement of decay times? i mean in layman's terms ... what improves in SQ if the decay times improve?

[OvalNut]

mike c wrote:

how do we hear the improvement of decay times? i mean in layman's terms ... what improves in SQ if the decay times improve?

Instead of "Powwwwwowwwwowww", you hear "Poww!". Individual sounds stop on time; they don't loiter and bounce around the room. They don't then have the opportunity to interfere with subsequent sounds. That yields more detail able to be perceived, and less sounds smearing together.


Tim

[mike c]

thanks Tim

powwwww huh!?

[JohnM]

The "Measured" tab shows all of the measurements you have loaded on one graph. To have an effect at low frequencies you need a large treatment volume.

[Bob_99]

mike c wrote:

how do we hear the improvement of decay times? i mean in layman's terms ... what improves in SQ if the decay times improve?

Tim is correct and for my setup, when I finished with all the treatment, the soundstage just filled the front of the room. It is difficult to describe but simply great to experience. Movies with good soundtracks are much more enjoyable to listen to, even when the movie itself may not be so great.

Bob

[thewire]

I also tried a measurement during the building of my bass traps and it did not seem right. I didn't even save the result until I had at least filled the corners with Roxul. That gave me a large smooth bump from 20Hz up to my peak which was left unchanged in around the 50Hz area. If you are looking at reflections that are not in the corner use the impulse response window and set the Y axis to %FS. Then right click and hold control to measure from the largest peak (your starting point) to any other peaks that may be present.

Before I remembered to do this I knew I needed bass traps on my back wall but I couldn't seem to be able to measure anything indicating a problem in my front row. Then I found the peak at near 33ft indicating that my now largest concern is a reflection off of my back wall and my side wall combined.

[thewire]

Any update on how the traps are helping?

[affeking]

thewire wrote:

Any update on how the traps are helping?

Not yet. I did look at the waterfall graph, but it wasn't showing anything significant either with the one trap.

On Saturday I should be finishing up the other 2 superchunk traps and installing them. I'll provide the before / after graphs on those sometime this weekend. I'm also going to put some broadband absorbers on the 1st reflection points on the walls, tho I won't be measuring those quite yet. Need to get a broadband mic

Will let ya know..

Thanks
Jeff

[Lobotomy^]

You only seem to measure sub response, but corner traps might have big effects on clarity of main speakers.

I recommend making all the measurements with one speaker at the time. One sweep for L, R and sub before installing traps and then another three measurements with traps in place. It's really hard, or impossible, to track the effects of traps when there is three speakers playing same frequencys. You might get deepers dips just because you dump echoes and sound from speakers happen to be in opposite phase.

[affeking]

Lobotomy^ wrote:

I recommend making all the measurements with one speaker at the time.

That makes sense - I will try it. However, I did spend quite a bit of time testing the FR of the room with the sub phase control at a bunch of different positions. I settled on the smoothest curve, so I think I'm getting minimal phase issues.

I am going to have to redo my measurements. I didn't really get a great before / after because the mic was not in the exactly same position when I did my after measurements this weekend. I will go ahead and move the traps out of the way and try all of this again this week and post my results.

I'm not going to post the current before / afters (due to the test not being completely valid), but if it is valid I'm unimpressed. The curve looks almost identical from 15-150hz (my after measurements only went up that far). I need to redo it, but so far I'm a little worried my efforts were somewhat in vein, at least in the lower frequencies that I was trying to attack.

More on this later.
Jeff

[DrWho]

Absorption is really only good to about 1/4 wavelength....

90Hz is about 12feet which means you need an absorber panel on the order of 3ft thick in order for it to be "effective".

45Hz would need to be 6ft, and 20Hz about 12ft (yikes).

Another thing to watch out for is HF reflections....even though it's an "absorbent material", most of the foam type solutions are actually reflective at the higher frequencies. It's probably no worse than a bare wall, but the acoustical properties vary quite a bit with different frequencies.

The most effective solution for really low frequency acoustical treatment is actually a helmholtz resonator. The tricky part is that the tolerances are quite small in order for it to be effective...which is where having the ability to measure results becomes extremely handy.

[atledreier]

I was in the same boat. Put in some tratments, measured, put in some more, measured... didn't measure much difference. Then I gave up, and just left the treatments for a month. Decided to do another run of tests, and removed all the treatments. It was like walking into a church! The room sounded completely different, even in the lower region. So the treatments are doing SOMETHING, even if measurements don't show it that clearly.

[DrWho]

well if you really want to see the effect of treatments over the midband and high frequencies, then you really should be looking at the ETC curves...it should be readily apparent there.

[Bob_99]

DrWho wrote:

Absorption is really only good to about 1/4 wavelength....

90Hz is about 12feet which means you need an absorber panel on the order of 3ft thick in order for it to be "effective".

45Hz would need to be 6ft, and 20Hz about 12ft (yikes).

Another thing to watch out for is HF reflections....even though it's an "absorbent material", most of the foam type solutions are actually reflective at the higher frequencies. It's probably no worse than a bare wall, but the acoustical properties vary quite a bit with different frequencies.

The most effective solution for really low frequency acoustical treatment is actually a helmholtz resonator. The tricky part is that the tolerances are quite small in order for it to be effective...which is where having the ability to measure results becomes extremely handy.

Mike is correct but even thinner panels have some effect and if you put in enough panels, the sum of the smaller effects add up even at the lower frequencies. If you go to Ethan Winer's web site and look at his room and his FR curves, you'll get a good idea of what it takes to tame a room. Ethan also has some great articles on acoustics that are well worth reading. This amount of treatment isn't to everyone's taste but even if you do a bit, it helps a lot.

Bob

[affeking]

Bob_99 wrote:

even if you do a bit, it helps a lot.

This is exactly what I was basing my expectations on. I know treatments encounter a point of diminishing returns, and that your initial treatment (done correctly) can yield significant results.

Obviously, I'm not quite done yet. However, I have covered my 3 visible wall-wall corners from floor to ceiling with solid OC3. The panels aren't as deep as drwho mentions of course, but the face of them is about 18" from the corner (rough guess). I plan to add some more panels in key positions, using Ethan's pinknoise MP3 and my SPL meter to locate them.

Although I did test the phase of the sub extensively, I suspect some of this could be due to phase cancellation with the fronts. The reason I think this is that the sound is noticeably better when watching movies. My crossover is set to 80hz in my receiver, but when testing with REW I will be seeing overlap starting at about 45hz. I am going to have to test completely separately, as someone mentioned earlier.

So, I will get those better measurements in the coming days here and hopefully we'll see something. I know its not going to be earth-shattering below 100hz, but I certainly expected some improvement.

Thanks guys,
Jeff

[SRR]

If you are using the Roxul bendable wool stuff, you might try compressing it by a good amount, try 2/3 of the height that you have now. Roxuls Rockboard, is much better below 100Hz, and is twice as dense, and breaks before it bends to far kind of stiffness. I bought 12 sheets of 2" thick and 2' x 4' for under $150 shipped. I am thinking about getting some more and cutting some up into triangles and going to about 5' high in my front corners. I would go all the way up, but I have my main electrical panel in the front right corner, and per my electrician (Uncle ), I can't cover the panel up even if its a removable object. I guess that makes sense, don't want a breaker getting stuck and spew sparks on my bass traps :raped:.

[SRR]

Ooops I see you are using 703, guess I had Roxul on the brain. Does 703 compress? I know 705 is twice the density of 703 and thus better at bass trapping. So I wonder if you could compress it a little and see if that helps?

[affeking]

First off, yes SSR, you can compress the 703 to an extent. I did this a little bit, just to make sure everything was snug in the frame I built. I really should have gone with 705 for the corner traps, but it worked out better for me logistically to get 703.

On to the topic of those new graphs I promised. Got them, and I'm seeing more of the same. I'm a bit disapointed, but I am still positive the sound is better with the traps. First off - I measured with a radioshack meter as my mic, and went from 10 - 200hz. I think the mic can handle that range with decent accuracy, at least for the purposes of comparing before/afters.

I tested a few different ways. I wanted to isolate the sub and the mains, because ordinarily my crossover (set in the receiver) would be set at 80hz. For the tests with both hooked up here, I'm stuck without a crossover, as the sub does not have one built in and REW isn't sending digital sound out.
First, I left the fronts and the sub plugged in, and calibrated to sub volume.
Before: http://picasaweb.google.com/jrames/R...22202513198178
After: http://picasaweb.google.com/jrames/R...22202513198194
Also, here is a compare of the two on one graph - red is before and blue is after
http://picasaweb.google.com/jrames/R...24620579785890

Next, I tested the sub and fronts independant of each other, recalibrating with the SUB pinknoise each time (even for the fronts)
Sub only, no traps: http://picasaweb.google.com/jrames/R...22206808165506
Sub only, w/traps: http://picasaweb.google.com/jrames/R...22211103132818

Fronts only, no traps: http://picasaweb.google.com/jrames/R...26823898008770
Fronts only, w/ traps: http://picasaweb.google.com/jrames/R...26819603041458

Out of all of this, I see only two improvements. The dip at 10-15hz is reduced a bit, and the dip at 120 or so is also reduced. Not much else, at least at these frequencies. I'm also a bit confused, because the tests with the front / sub independant of each other showed differing changes before/after the traps at 120. For the sub graph, I see the dip was reduced. For the fronts only, I see a large dip was CREATED? at 120 when I added the traps? I did not touch the phase or alter the position of anything between tests, so I would have expected to see similar changes on the individual graphs as I did in the combined front/sub graphs. What's going on here? Maybe that after front graph was an oops.

So I'm a little discouraged. Obviously this is not a huge overhaul, but I did fill 3 corners floor to ceiling with OC703. I was hoping to notice at least some flattening of the FR sub 100hz. Not so much, at least not from this listening position (I only measured from my primary seat for these tests). On the other hand, things do SOUND better to me. My ears aren't tuned well enough to say exactly what frequencies I'm noticing improved clarity in, but my guess is its a lot higher stuff than what I've measured here.

I am going to try again with a better mic, but I'm not sure it'll be a huge change in this range. What do you all think?

Next step is to finish constructing my first reflection point absorbers and see how I'm doing at the higher frequencies.

Thanks,
Jeff

[atledreier]

I've noticed the same thing. No big changes to my graph, but noticeably drier and better sound, so I'm sticking more stuff in my room.

[brucek]

I am going to try again with a better mic, but I'm not sure it'll be a huge change in this range. What do you all think?

The effects of the treatment will be in the decay of the signal. Use waterfall graphs for signals below 200Hz. For higher frequencies the RT60 plot is used.
You are able to overlay before and after on the waterfall.

Also note the Measured tab in REW allows for overlay of response measures for comparisons.

brucek

[affeking]

brucek wrote:

Also note the Measured tab in REW allows for overlay of response measures for comparisons.

I did provide an overlaid graph in my note above...it might have been lost in the flurry of links I provided.
http://picasaweb.google.com/jrames/R...24620579785890

Another thought...I think I've already seen a lot of discussion about this, but I still don't understand the sentiment that a waterfall is going to show results that wouldn't also be present in a FR graph. I understand that a waterfall shows more directly what the bass trap is doing (reducing reflections), but the FR graph shows the result of THAT in terms of what we are really trying to accomplish.

The reason we want to reduce reflections is to keep them from canceling or amplifying waves. The reduction in reflections is seen in the waterfall, but the net result on the Frequency Response (what I ultimately care about) is seen on the FR. So why wouldn't we go straight to the FR to validate the success of treatments?

I am a beginner with acoustics, so please let me know where I've gone wrong in the above logic path.

Thanks,
Jeff

[atledreier]

The FR graph shows you what happens right after the sweep, a short window. The waterfall shows you that in the first slice, and the following slices show the FR in that instance of time (after20ms, 40ms, 60ms and so on, depending on settings).

[brucek]

but I still don't understand the sentiment that a waterfall is going to show results that wouldn't also be present in a FR graph

The 2D frequency response graph is fine to show the steady state response graph, but needs to be combined with the 3D waterfall to show the rooms decay of modal response.

In fact, look at the waterfall graph and set the vertical and horizontal axis of the graph to the same as your frequency response graph (and remember to use LOG and not LIN mode), and pull the slider from 30 slices to 1 slice. Look familiar? It's the 2D frequency response graph. Now move the slider to 2. This is what is happening 10ms later in time (30slices divided into 300ms time axis). Keep moving the slider and see what is happening out to 300ms (adjustable).

If there is a huge ring out of frequencies around 60Hz (for example), and there is still appreciable level of that signal at 300ms, this will be quite audible. That's one of the reasons bass can sound tight or like mush. You need this valuable information. 3D is very important (if not the most important) when evaluating room treatment...

brucek