John Sayers' Recording Studio Design Forum Archive

Hello, I am preparing to build about a dozen broadband absorbers and I need advice to choose between two insulation options. OC 703 or 705 is not available in my location, but I have found two products at the bazaar which will suffice:
- One is compressed mineral wool which is approximately the same density as OC 703 = about 50kg per cubic meter.
- The second option is compressed fiberglass which is much denser, about 170kg per cubic meter.
I was hoping to find something in between these two densities (like OC 705), but there isn't anything available. Also, unfortunately, I don't have any flow rates ratings for these (I determined the density with a bathroom scale and a package of each.)

I have built a new mix room, and now I'm putting in the absorbers. The ceiling is stuffed full of lightweight fiberglass, and I am wanting to tame the walls. Which of these two options would you recommend for the standard 4" or 6"-thick wall panels and for floor-corner superchunks? Using Bob Gold's numbers, the 703-style mineral wool seems it would be a bit lacking for bass absorption, but he doesn't list coefficients for anything as dense as 170kg/m3.


I am also curious at what point do you decide to do a panel membrane bass trap (or even a Helmholtz) rather than an absorber? Do you have to try it with absorption and run an REW test and see if you need to fix a particular frequency and then rip out a superchunk so that you can put a panel in its place? Or is that something I should install from the outset based on room dimension calculations?

- One is compressed mineral wool which is approximately the same density as OC 703 = about 50kg per cubic meter.

OC-703 is not made of mineral wool: it is made of fiberglass, so the material you have is not comparable. Even though the desnity is similar (50kg/m3 vs. 45 kg/m3), the properties of the fibers themselves are very different, so you can't extrapolate the performance of one product to the other.

- The second option is compressed fiberglass which is much denser, about 170kg per cubic meter.

That's way too dense. About SIX TIMES too much. Not much use at all for broadband! It might be OK if you wanted something to treat only mid-highs without too much effect on the lows, but not much good for a broadband panel

I was hoping to find something in between these two densities (like OC 705),

When you say "broadband panel", are you intending that they work down to low frequencies? OC-705 is not much good for low frequencies, compared to OC-703, for example.

I have built a new mix room, and now I'm putting in the absorbers.

Have you done any tests on the empty room, using REW? Do you already have your speakers and mix position set up in the optimal locations? If so, then please post the test files (mdat) some place where we can download them.

The ceiling is stuffed full of lightweight fiberglass,

What density? How thick? "Light weight" can mean different thongs to different people...

Which of these two options would you recommend for the standard 4" or 6"-thick wall panels and for floor-corner superchunks?

If those are the only two potions you have, then go with the "compressed mineral wool which is ... about 50kg per cubic meter.". That's the only one that makes sense. However, I'm confused here, because you say you only have two options, but then you say that you also have "lightweight fiberglass" insulation in the ceiling! That's clearly a third option... why are you not considering that one for your bass traps and broadband panels?

Using Bob Gold's numbers, the 703-style mineral wool seems it would be a bit lacking for bass absorption,

I'm not sure why you would say that! OC-703 is widely recognized, and widely used, in bass traps. It works very well. 4" thick =C-703 on ASTM A-mounting has a coefficient of absorption of 0.84 at 125 Hz, and 1.24 at 250 Hz. Why do you say that this is not good?

but he doesn't list coefficients for anything as dense as 170kg/m3.

Probably because that would have terrible low-frequency absorption characteristics. It is way too dense. Higher density in general is worse for low frequencies.

the 703-style mineral wool ...

Once again, 703 is fiberglass, not mineral wool, so it is not comparable. You cannot extrapolate the performance of one to the other, as they are different products. You cannot extrapolate the performance of a gasoline powered car engine to a diesel powered truck engine: they are two entirely different things. You can only compare things that are similar.

I am also curious at what point do you decide to do a panel membrane bass trap (or even a Helmholtz) rather than an absorber?

Personally, I practically never do that. There are very, very few occasions where I have needed to use a membrane trap or Helmholtz trap, and I would be very, very hesitant to try to deal with modal issues using a Helmholtz resonator. It's just too complicated to get the tuning exactly right, and they take up too much space in the room... even assuming it is feasible to get it in the right location. Almost all of the rooms I design, I use porous absorption for the majority of the low frequency treatment. Very occasionally I might go for a panel trap.

? Do you have to try it with absorption and run an REW test and see if you need to fix a particular frequency and then rip out a superchunk so that you can put a panel in its place?

I have NEVER had to do that. Superchunks are very effective down to low frequencies, and even if I did still have a problematic mode after the superchunks are in, I would never take them out to install a membrane trap. To start with, putting a membrane trap straddling the corner would not be useful at all, as it would not be in the best location, and the Q would be way too broad, and secondly, I would not destroy a useful, working device that is dealing with multiple issues all at once, and replace it with a device that only deals with one single issue. If I did have to build a membrane trap of some type, I would put it in the best location... assuming that there is enough room to do that.

Or is that something I should install from the outset based on room dimension calculations?

Room prediction using equations will only take you so far. You will still need to measure the room with REW after it is built, to see how it is really performing. Prediction assumes that you use perfect materials, perfect techniques, cut everything perfectly, build with perfect accuracy, and that the air temperature, pressure and humidity in your room are perfectly consistent everywhere, and perfectly match standard conditions. In the real world, none of that happens, so the room will not perform the way you predicted.

So just build the room, then set up the speakers and mix position at the correct places, then measure it, then decide what treatment you will need where. Then put in the first round of treatment, and measure again after each device goes in, to see how well it worked, and what else still needs to be done.


- Stuart -

Wow Stuart, Thank you for chiming in on this! It's nice to get advice from the passionate.

Yes, I am looking to get down to low frequencies. My room is unfortunately almost square (5.2m) and 2.5m tall, so I need all the low-end help I can get.

If those are the only two options you have, then go with the "compressed mineral wool which is ... about 50kg per cubic meter.". That's the only one that makes sense. However, I'm confused here, because you say you only have two options, but then you say that you also have "lightweight fiberglass" insulation in the ceiling! That's clearly a third option... why are you not considering that one for your bass traps and broadband panels?

Yes, I do have a third option of fiberglass batt rolls. I have put about 30cm of that in the ceiling, held in place by a net, so the ceiling is soft. Most of the "diy panels" online use compressed fiberglass, so I didn't even mention the rolls. How thick would such a panel have to be? It looks like it would have to be at least 30cm, and that would be a lot of room space to give up...
I do not have any spec data for the batt rolls, but I will see what I can find.

Have you done any tests on the empty room, using REW? Do you already have your speakers and mix position set up in the optimal locations? If so, then please post the test files (mdat) some place where we can download them.

I did run REW on the room after the ceiling insuulation was in, but I am not sure if I did it right because my graphs don't look like I expected. :0 I'm using a Behringer ECM8000 and an Aphex USB500 interface. I've put the mdat file here:
http://s000.tinyupload.com/?file_id=020 ... 2062023437

So just build the room, then set up the speakers and mix position at the correct places, then measure it, then decide what treatment you will need where. Then put in the first round of treatment, and measure again after each device goes in, to see how well it worked, and what else still needs to be done.
- Stuart -

Thank you again.

Most of the "diy panels" online use compressed fiberglass, so I didn't even mention the rolls. How thick would such a panel have to be? It looks like it would have to be at least 30cm, and that would be a lot of room space to give up...

How thick for what purpose? Thickness will not change, regardless of the materiel that you use. If your panels would need to be 30cm thick for fiberglass, then they would also need to be 30cm thick for medium density mineral wool, or high density mineral wool. Thew thickness si more related to the frequency range that you need to deal with, than it is to the type of material. Thicker panels have better low-frequency extension.

It looks like it would have to be at least 30cm,

Why? Why do you think it wold need to be 30cm thick? How did you calculate that?

I do not have any spec data for the batt rolls, but I will see what I can find.

Density is the most important, or gas flow resistivity would be even better, but you probably won't be able to find that... not many manufacturers bother measuring it.

I did run REW on the room after the ceiling insuulation was in, but I am not sure if I did it right because my graphs don't look like I expected

They look a bit off but not too bad. But it looks like you did not calibrate REW before you started! Calibrate REW with each speaker producing 80 dBC y itself (with the other one turned off) using REW's calibration pink noise, then you should automatically get 86 dBC with both speakers on.

Then to three tests: one with just the left speaker, one with just the right, and one with both.


- Stuart -

Stuart, you continue to be a gold mine for me. I have been reading books and threads for a year and somehow I got it into my head that the denser products were more effective at lower frequencies. Thank you for the corrective - I've gone back and looked at the charts more carefully and realized that now.

Soundman2020 wrote:

How thick would such a panel have to be? It looks like it would have to be at least 30cm, and that would be a lot of room space to give up...

How thick for what purpose? Thickness will not change, regardless of the materiel that you use. If your panels would need to be 30cm thick for fiberglass, then they would also need to be 30cm thick for medium density mineral wool, or high density mineral wool. Thew thickness is more related to the frequency range that you need to deal with, than it is to the type of material. Thicker panels have better low-frequency extension. Why do you think it would need to be 30cm thick? How did you calculate that?

I mention 30cm because I want to get down to the two ugly nodes REW showed - 52hz and 79hz - and I was just unscientifically extrapolating down from the "fluffy pink" listed on bobgolds. I have a feeling you're about to tell me some brilliant way of solving this problem

Density is the most important, or gas flow resistivity would be even better, but you probably won't be able to find that... not many manufacturers bother measuring it.

I went back to the market today and was able to get a lable for the batt rolls. I was surprised to find that they are Knauf! Also, though I can't be 100% sure, I am pretty sure that they are this product: Knauf Earthwool Loft Roll 44 ( http://www.knaufinsulation.co.uk/produc ... ft-roll-44 ). So, it seems they're not exactly fiberglass either. They're described as "a flexible low density glass mineral wool quilt." The spec sheet lists the density as "from 9 to 35 kg/m3". I'm assuming that when it's opened and fluffed up it'll be 9kg/m3.
I can find other options. For example, I also saw some faced batt rolls in the market today, but all they could tell me was that they were made in Russia (lol! So much for a spec sheet...)

it looks like you did not calibrate REW before you started! Calibrate REW with each speaker producing 80 dBC y itself (with the other one turned off) using REW's calibration pink noise, then you should automatically get 86 dBC with both speakers on.
Then do three tests: one with just the left speaker, one with just the right, and one with both.
- Stuart -

I will get right on this.

At this point, do you have enough information to recommend me a particular thickness or style of panel?
I really appreciate your taking the time to speak to these questions, so if you need more information I'll do my best to get you it.

Stuart, you continue to be a gold mine for me.

Thanks!

I have been reading books and threads for a year and somehow I got it into my head that the denser products were more effective at lower frequencies. Thank you for the corrective - I've gone back and looked at the charts more carefully and realized that now.

It's a common misconception, so don't feel bad! It just sort of "sounds" right that a "heavier" bass problem would need a "heavier" material to absorb it... but it ain0t true! A lot of acoustics is not intuitive.

because I want to get down to the two ugly nodes REW showed - 52hz and 79hz -

Ummmm.... I thought we were talking about broadband panels, not bass traps? Broadband panels hung along the side walls are not going to do a lot for 52Gz and 79 Hz modes (not "nodes"!).

In any event, I'm not seeing any modal issues at 52 and 79. This is what I see:

52Hz: 79Hz: You have intensity peaks there, yes, but they do not seem to be ringing, so they are probably not modal, and therefore even 30cm panels won't have much effect on those.

On the other hand, you do seem to have some issues in the mids, at around 220 Hz to 1 kHz, such as this biggie at 344 Hz:
Those will likely be tamed rather nicely by your panels... but they don't need to be 30cm thick. 15cm would be fine for those, or maybe even 10cm.

In order to hit the low frequencies, you'll need proper bass traps. Something much more substantial than broadband panels. I'd suggest superchunk style traps, diagonally accross the vertical room corners. I'd suggest that you make those extend about 90cm out along the walls, to get them deep enough. And they will need to run from floor to ceiling, to be effective for all modes.

I have a feeling you're about to tell me some brilliant way of solving this problem

Another option would be to treat your entire rear wall with hangers: Make that about 50cm deep, and angle the hangers a bit so they can be longer than 50 cm. The ones in the MIDDLE should be 50cm deep, but the ones in the corners need to be bigger: I'd go for 90cm again here, at least.

So, it seems they're not exactly fiberglass either. They're described as "a flexible low density glass mineral wool quilt."

It's probably close enough to fiberglass that you could use the basic assumptions for fiberglass insulation.

The spec sheet lists the density as "from 9 to 35 kg/m3"

That's not a very useful spec, now is it? That wuold be like MacDnalds advertising that you get "between 9 and 35 french fries in your happy meal". Or Great Wall advertising that there newest engine will get "between 9 and 36 km per liter". That's a rather broad range.... not so useful. Maybe you can get one pack, and weight it, then see how much volume it fills when fluffed out, to figure out the density, aprox? It wont be wasted: You can still use that stuff at some point in your studio.

but all they could tell me was that they were made in Russia (lol! So much for a spec sheet...)

That's a very useful spec, actually! It tells me that you probably don't want to use it in your studio!

At this point, do you have enough information to recommend me a particular thickness or style of panel?

I'd need to know the size of your room( accurate dimensions), the make and model of your speakers, the location of the speakers and mix position in the room, whether or not the speakers are soffit mounted (and if not, WHY not????), the materials that the room is built from, the correct REW data, and any other info on the room.

I did respond to your other thread a couple of days ago.... not sure if you saw that... Some of that info is in your other thread, but given my comments there, I'm not sure if it is valid at all....


- Stuart -

I looked over at the other thread. It's been a couple months, and due to factors out of my hands, I've had to forge ahead without the kind of specific guidance you're giving now.
NB: Helping people with music is an after-hours side-gig for me; my day job is as an academic. I really like this non-profit I'm helping, but they don't have the realestate to have a single-purpose room. It will have to be functional for a few things, so they've insisted on a few features - such as closets. The guy they hired to do the framing made the walls take up more room than I wanted, etc. So building this room has been a process of dialogue rather than dictatorship.

So here is the mix room I ended up with (for better or worse). 1)HVAC was already present, and it is sufficient. Heat is provided from in-floor pipes. Ventilation comes via 3 x 12cm tubes which snake around to the outside wall and have fans there. There is plenty of fresh air without the fans, and they are controlled by a simple switch.

2)The walls are 15cm studs on 60cm centers with the same batt insulation in between. Behind the studs is a concrete basement wall (to which they are not connected). The wall is faced with a layer of drywall/gypsusm on top of a layer of OSB

3)The ceiling is soft. By that I mean the very top layer of the ceiling is an 8" concrete slab (the floor upstairs). Below that we put a wooden frame (sitting on the walls, not connected to the concrete ceiling) and stuffed 30cm of batt insulation in it. The insulation is held in place with a cloth held up by a net (similar to a football goal net)

4)I have not done any soffits because the door is on the front wall. This may have been a massive folly, but my concern has been about the stereo imaging. I oriented the room this way in order to have a window. Do you have advice on soffits here? So far the speakers are just on stands near the wall, but I will try to figure out a way to mount them.

Ummmm.... I thought we were talking about broadband panels, not bass traps? Broadband panels hung along the side walls are not going to do a lot for 52Gz and 79 Hz modes (not "nodes"!)...
In any event, I'm not seeing any modal issues at 52 and 79. This is what I see... You have intensity peaks there, yes, but they do not seem to be ringing, so they are probably not modal, and therefore even 30cm panels won't have much effect on those.
On the other hand, you do seem to have some issues in the mids, at around 220 Hz to 1 kHz, such as this biggie at 344 Hz

Ahh. Interesting. Now I see what you're looking at in the waterfall graph.

Those will likely be tamed rather nicely by your panels... but they don't need to be 30cm thick. 15cm would be fine for those, or maybe even 10cm.

And, just to be clear, you would recommend these panels to be made from the fluffy batt insulation rather than the compressed mineral wool (50kg/m3)?
Would you run floor-to-ceilng as well?
What about the floor-wall corner?

In order to hit the low frequencies, you'll need proper bass traps. Something much more substantial than broadband panels. I'd suggest superchunk style traps, diagonally accross the vertical room corners. I'd suggest that you make those extend about 90cm out along the walls, to get them deep enough. And they will need to run from floor to ceiling, to be effective for all modes.

So, just trying to clear out the muddle in my head, is there anything substantially different between a broadband panel and a bass trap other than 1)the thickness and 2)the location? ie, are they both just the same type of insulation stacked inside? (of course, I'm not meaning panel traps or resonators)

Another option would be to treat your entire rear wall with hangers: Make that about 50cm deep, and angle the hangers a bit so they can be longer than 50 cm. The ones in the MIDDLE should be 50cm deep, but the ones in the corners need to be bigger: I'd go for 90cm again here, at least.

Ahh, this would be nice. Unfortunatly that's where I'll run into the external constraints. They want to be able to use this room for more than mixing (small meetings, language lessons, etc), so I don't think I can take that much space. I will see what I can fight for

The spec sheet lists the density as "from 9 to 35 kg/m3"

Maybe you can get one pack, and weight it, then see how much volume it fills when fluffed out, to figure out the density, aprox?

My rather-questionable-bathroom-scale tells me that it's about 10kg/m3, so I think the spec sheet is listing the difference between packed-and-unpacked.

but all they could tell me was that they were made in Russia (lol! So much for a spec sheet...)

That's a very useful spec, actually! It tells me that you probably don't want to use it in your studio! [/quote]
lol!

At this point, do you have enough information to recommend me a particular thickness or style of panel?

I'd need to know the size of your room( accurate dimensions), the make and model of your speakers, the location of the speakers and mix position in the room, whether or not the speakers are soffit mounted (and if not, WHY not????), the materials that the room is built from, the correct REW data, and any other info on the room.
I did respond to your other thread a couple of days ago.... not sure if you saw that... Some of that info is in your other thread, but given my comments there, I'm not sure if it is valid at all....

I think I got your check list answered except the speakers and the REW data. I will get you that as soon as I go back to the studio.
Thanks again!

Just a note to say I find this thread fascinating, since my current understanding of room acoustics is probably on a par Lathqe's - so with every erudite post from Soundman I probably learn as much as him.

Just as a minor example, I had the same idea as Lathqe that denser fibres would absorb more bass energy, but with him now know it's panel thickness that is the prime determinant, and less dense fibres are more effective (I assume up to a point, because very sparse fibres may not absorb much wave energy). Anyway, I'll keep following with interest, while also proceeding to build my first, modest broadband panels.

mark77 wrote:Just a note to say I find this thread fascinating, since my current understanding of room acoustics is probably on a par Lathqe's - so with every erudite post from Soundman I probably learn as much as him.

Just as a minor example, I had the same idea as Lathqe that denser fibres would absorb more bass energy, but with him now know it's panel thickness that is the prime determinant, and less dense fibres are more effective (I assume up to a point, because very sparse fibres may not absorb much wave energy). Anyway, I'll keep following with interest, while also proceeding to build my first, modest broadband panels.

Hi Marc, and welcome to the forum!

Yep, as with so may physical and statistical properties, there's a roughly bell-shaped curve that related density to absorption. or rather, it relates gas flow resistivity to impedance! And that, in turn, is related to density. It isn't actually the density as such that matters, but rather the gas flow resistivity of the insulation. As the name implies that's basically a measure of how the insulation resists the flow of gas going through it, and since sound waves are basically "moving gas" (vibrating air), that GFR is a good measure of how the insulation reacts to sound waves. GFR is related to impedance, and is measured in the unlikely units of MKS rayls. But you wont see that marked on the majority of insulation packages at your local Home Depot! Most manufactures don't ever both measuring it, because it is not relevant for the primary purpose of their product: thermal insulation. It only matters to us nut cases that want to use the same product for something that it was never specifically design for: acoustic isolation in our studios. But most manufacturers do measure the density of their products, and do publish that, or you can measure it yourself with a simple bathroom scale and a tape measure, and it turns out that there is a rough relationship between density and GFR, for each type of insulation. The relationship is not very linear, but it's close enough for our purposes. As long as you realize that it is different for each type of insulation (one relationship for fiberglass, another for mineral wool, another for polyester, another for denim, another for cellulose, etc), then you are fine, and can use the density to figure out the acoustic use.

Regarding the common misconception that "bass needs heavy insulation", that's one of the many, many things about acoustics that sounds right, but is wrong. Much of acoustics is not intuitive, as in this case. It just seems logical that "heavy" waves need "heavy" treatment, but once you get into the the theory of how sound actually works, rather than the way we assume it works, you can see why this is wrong. If you want a different mental picture that "makes sense" (but is also wrong... ) then you can think of it this way: "Big waves need big spaces between the fibers to get thorough, otherwise they bounce off and go back". That's not correct, but it does give you the right answer! (Albeit for the wrong reasons). In reality, it is all about impedance and gas flow resistivity, and equations with both real and imaginary numbers in them, which is harder to explain in simple mental pictures....

- Stuart -

Thanks for the welcome and the fibre info, Stuart - putting a scientific angle on a topic that you often see anecdotal stuff on. I had to read it twice, and certainly get the gist.

I may quickly air a gripe - some major manufacturers of acoustic fibres, like CSR Bradford in Australia, do not have GFR or even rudimentary acoustic data for some of their products. The CSR fibre "Acoustigard", for example, is aimed purely at sound applications but when I called them this past week, they only had frequency vs absorption figures for 11kg/m3 and 14kg/m3, and nothing for densities from 18 through 32kg/m3). So how they expect industrial customers to make product selection decisions is unclear.

However, I don't want to hijack Lathqe's thread which displays such a nice progression, because I've been learning much from his endeavours and your replies. Since I'm also just starting to tentatively run REW at my place, I hope he'll soon come back with his promised REW measurements so the thread can roll on. In the mean time I'll butt out to give him clear air.

Hello Mark,
As one newbie to another, welcome to the forum. Feel free to keep chiming in!
I apologize for the radio silence - I was unexpectedly called in to teach a modular course during the times when I was planning to work on the studio. I never have really enjoyed reading forum posts from 5 years prior which say "I'll get back to you on Monday," but now I understand why they're useful: this is a conversation and it's a bit rude to just disappear. If all goes well, I'll be back to the studio a couple times this week

I do, however, have one question about calibrating REW: Using a usb interface and an ECM8000, where does REW reference its volume? I mean, is it possible to calibrate REW without an independent SPL meter? I don't have an SPL meter. When I made my original mdat, I had adjusted the preamp input and speaker outputs to read the same levels in the program (about -12dbc). I have read the manual and watched GIK's video and other posts, but I still can't seem to figure out how it would be possible to calibrate without an independent meter...

but I still can't seem to figure out how it would be possible to calibrate without an independent meter...

Short answer: YOu can't!

You are absolutely correct: Unless you have an external sound level meter, then you cannot calibrate REW to the real sound level in teh room.

Some people try to use apps on cell phones, and while that might be better than nothing, it still is not accurate. Cell phone mics are not acoustic measurement mics: The aren't even omni mics! You can't compare a US$ 0.50 mic against a US$ 100 or US$ 500 calibrated acoustic measurement mic. Cell phone mics are designed to capture the human voice at a distance of a an inch or two, and to reject ambient sound. That's sort of the opposite of what you want!

If you have nothing else, then in an emergency you could use a cell phone app, but you'd need to get a real hand-held sound level meter at some point to re-calibrate it correctly.

- Stuart -

Lathqe, I ran into the same issue.

I own a ubiquitous analogue Radio Shack SPL meter and thought I could use that until I remembered that, years ago, I poked a screwdriver into the calibration screw hole just to "check it out". Sadly, I couldn't remember whether I turned it or not, so I had to get another SPL meter since it was quite expensive to rent a calibrated meter to compare mine with.

Thought I'd get Class-I meter, but once I saw the prices I went for a Class-II instead (at 20Hz, the tolerances are +/-2.5dB for Class 1 and +/- 3.5dB for Class 2, and I thought the latter could be ok).

I may throw in a question about the new one, a model Landtek "SL5868P" sold to me as a Class-II for around usd140. I see they are also sold on Amazon. The spec for this says "Built in Calibration Signal: 94 dB at 1k Hz (sinusoidal)".

What does that mean - can you quickly re-calibrate this SPL meter by pressing a button without using an "external" sound-generating calibrator? Unfortunately, the manaul is not clear on this.

Soundman2020 wrote:

but I still can't seem to figure out how it would be possible to calibrate without an independent meter...

Short answer: YOu can't!
You are absolutely correct: Unless you have an external sound level meter, then you cannot calibrate REW to the real sound level in teh room.

whew. My universe is restored to balance.