thismanysounds - Basement Studio in London UK

[thismanysounds]

Okay, one more question.

It is looking like the height of the inner leaf will be around 2.8-2.9m. Following the advice of 1.4m tweeter height leaves the speaker basically in the middle of the room, which isn't idea for modes. What is the best way of dealing with this? Is it even an issue with an absorptive ceiling?

Colm

[Soundman2020]

Following the advice of 1.4m tweeter height

Not sure where that came from, but it isn't correct. The correct height for any speaker is where the ACOUSTIC AXIS of the speaker is at the same height as your ears when seated. For most people in most studios, that is 1.2m above the floor. The height of the tweeter is irrelevant: what matters is the height of the acoustic axis. With some speakers, that might happen to be very close to the tweeter, with others it might be further away... It depends on the speaker.

And if the speaker happens to be tilted, then the correct height is whatever gets the acoustic axis aimed at your ears, when seated in the normal listening position.

- Stuart -

[thismanysounds]

Oops, slow response from me. That figure was just pulled from my (bad) memory, so I did a bit more research. Stuart you are very right about tweeter height vs acoustic axis.

Newell's book says

Site the mid-range units approximately 145-148 cm above the floor. This is a mid position between the ear height of a small person sitting down and a tall person standing up.

. My natural position with a console is standing, whereas this 'hybrid setup' will be more inclined to sitting. I will need to think about this more...

Stuart, I know you are run ragged at the moment, but can I bother you with these questions again:

Walls
- Between the rooms there is 2x19mm PB - >250mm air (including studs) - 2x19mm PB. Is it worthwhile putting extra plasterboard in one of the rooms (perhaps the 2nd control room, which will be use independently) and/or increasing the gap, so that it is more independent?
- Similarly, should I increase the air gap between the CR front wall and the studio?

Windows
As with a lot of studio build stuff, it is often more cost effective to go with less specialised products. This got me to thinking about other places where thick glass is used. I came up with glass flooring, for example (http://www.diometonline.co.uk/product/2 ... loor-glass). I called up and they said it was 12-12-1.5 construction. They also have a 15mm which would be of similar density to 2x19mm plasterboard (assuming the actual glass is 14mm). Would this work? Also in terms of size, is there a good ratio (square looks odd) and or a minimum size to go for? Does it make sense to use a larger pane in the live room for a bigger viewing angle (possibly not worth the extra complexity)?

Ceiling
I plan on have an independent joist ceiling built on the stud walls, with the same 19mm-GG-19mm sandwich. This will require something like a 220mm deep joist for the span. If I used a resilient hanger instead, then I could use the space for duct runs. Would that be a weak link in my design?

Colm

[Soundman2020]

- Between the rooms there is 2x19mm PB - >250mm air (including studs) - 2x19mm PB. Is it worthwhile putting extra plasterboard in one of the rooms (perhaps the 2nd control room, which will be use independently) and/or increasing the gap, so that it is more independent?
- Similarly, should I increase the air gap between the CR front wall and the studio?

That's a very decent amount of mass, and a very decent sized airspace, so you are probably fine with that, unless you want extreme isolation. I don't have a lot of time right now to go back and check the rest of your design, but as long as your rooms are fully decoupled (nothing touching anywhere), and you don't need major extreme isolation, then you should by more than fine with what you have already.

Windows
As with a lot of studio build stuff, it is often more cost effective to go with less specialised products. This got me to thinking about other places where thick glass is used. I came up with glass flooring, for example (http://www.diometonline.co.uk/product/2 ... loor-glass). I called up and they said it was 12-12-1.5 construction. They also have a 15mm which would be of similar density to 2x19mm plasterboard (assuming the actual glass is 14mm). Would this work? Also in terms of size, is there a good ratio (square looks odd) and or a minimum size to go for? Does it make sense to use a larger pane in the live room for a bigger viewing angle (possibly not worth the extra complexity)?

Here too, as long as your glass density roughly matches your wall density, and the air gap sizes are similar, then you should be fine. Glass is about 3 to 4 times more dense than drywall, so as long as your glass is at least one third the thickness of your drywall, you're good. 38mm of drywall works out to about 13mm of glass, very roughly. The actual density of both your glass and your drywall might be different from the "standard" values I'm using, but you are certainly in the ball park.

Ceiling
I plan on have an independent joist ceiling built on the stud walls, with the same 19mm-GG-19mm sandwich. This will require something like a 220mm deep joist for the span. If I used a resilient hanger instead, then I could use the space for duct runs. Would that be a weak link in my design?

Have you considered building your ceiling "inside out", with the joists underneath and the drywall on top? That would leave the space above the actual ceiling surface free for ducts, and also gives you plenty of space between the joists, which are now inside the room, for ceiling treatment...

Another possibility might be to bring your joists closer together, and/or use better quality wood for your joists, so you can use smaller dimensions (less than 220mm).

Resilient channel is good stuff, but it won't get you the same level of isolation as fully decoupled. Also, I'm not sure it is able to safely carry 2 layers of 19mm drywall... you should check with the manufacturer to see what the maximum load is for his product.

Hope that helps!


- Stuart -

[thismanysounds]

This project now has the go ahead, so I have been scrambling to finish the design as much as possible before we start.

The main changes:
- Control A has been turned 90 degrees, and now loosely RFZ. The window is now on the side and part of redirecting early reflections to the back of the room.
- A small machine cupboard has been added.
- The framing reflects known obstructions (such as beams & columns)

Floorplan:

Floorplan 20140121.png

3D render:

3D 20140121.jpg

[thismanysounds]

EDIT: I have marked all the questions I found answers for in RED

Okay, now the first of my 'finalising plan question posts'.... HVAC

Cooling
I was looking into mini-split systems, and had the realisation that the noise specifications on slim ducted units were much lower for the same capacity. Take this unit for example:
http://www.orionairsales.co.uk/fujitsu- ... 3309-p.asp

I am considering two options:
A - Locating the AHU outside the room, connected with silencers.
Pros & Cons:
+ Potentially quieter
- Need silencers
- Potentially takes more space in live room.
- More risk of reducing isolation

B - Locating everything in the room, with short runs to vents.
Pros & Cons:
+ No penetrations
+ Cheaper & faster
+ Low static pressure
- Potentially noisier
- Worse access if hidden behind acoustic treatment

So questions:
1) Is there a reason why I haven't seen slimline ducted systems used? Can they handle the static pressure created by the silencers?
2) Would it be possible to integrate ventilation into design A? Is it better to have them working independently?


Ventilation:
I plan on using using inline fans and silencer boxes to give the required 10l/s per person. I am also interested in the possibilities of free cooling, and "smarter" HRV systems.

3) As a HRV in 'free cooling mode' will not offer any more cooling power than using inline fans, are they a waste of money? The main advantage I see is the ability to maintain a constant temperature.
4) If I were to use free cooling alone would I run into issues with humidity?
5) To use free cooling effectively would I need to increase the number of air exchanges (and therefore ducts, silencers etc) significantly?
6) How do I size silencer boxes? Would a full height (2.8m) box without any turns be effective?
http://www.gearslutz.com/board/9630612-post9.html 3-4 is usually sufficient
7) How do I calculate the static pressure of DIY silencers?
http://www.pdhonline.org/courses/m199/m199content.pdf

[thismanysounds]

Next up... Isolation!

An acoustician friend of mine was kind enough to conduct impact noise tests using a B&K tapping machine (very cool to see in action) as a favour. We tested the impact isolation between the office above and where the studio is (results attached).

Here is the current floor construction:
Office --> Carpet tile --> Raised access floor (~300mm high, not resiliently mounted) --> Concrete slab (believed to be 150mm) --> Place where studio will be.

Here is the proposed construction:
Office --> Carpet tile --> Raised access floor (~300mm high, not resiliently mounted) --> Concrete slab (believed to be 150mm) --> 250mm air gap --> 2x19mm PB (hung off 200x50mm joists sitting on wooden studs) --> Inside of rooms.

I have looked for information to calculate the effect of adding the studio inner leaf. Here is what I have come up with, please correct me if make incorrect assumptions:

National Research Council Canada - http://www.nrc-cnrc.gc.ca/ctu-sc/files/ ... 35_eng.pdf

Gypsum board ceilings suspended resiliently (independently) from a concrete slab offer another approach for increasing the impact sound attenuation. Unfortunately, there is little information available about the attenuation values of such floor systems. However, it is known that increasing the mass of the gypsum board, the depth of the cavity, or the amount of sound-absorbing material all increase the IIC relative to that of the bare slab. The increase can range from four or five points to more than 30. These three factors are the same ones that improve the control of airborne sound transmission.

Mass Law - Impact noise manifests in the receiver room by the floor re-radiating the impact energy into the air. Once this has happened can the airborne noise reduction of the new leaf be added to the impact noise measurements? (as long as the new leaf is structurally decoupled)

Kinetics test information - http://www.kineticsnoise.com/arch/tests ... 001049.pdf http://www.kineticsnoise.com/arch/tests ... 001084.pdf

Chart

Comparison of NISPL.jpg

Analysis
Comparing our slab to their test measurements, the raised access floor seems to improve impact noise especially at high frequencies. The main exception is 125Hz, the possible resonant frequency of the system.

The IIC of the ceiling increases by 23dB by adding a resiliently hung ceiling. This is in the range suggested by NRCC. The results obtained using Kinetics hangers are significantly lower than predicted by mass law. Possible reasons:
- Lower range hanger so may not provide 100% decoupling
- Air coupling between the concrete and air. This would imply that mass law is inappropriate.

Questions
A) The tapping machine used is calibrated to international specs. Are the impact levels higher than I should expect from office footfall?

The tapping machine is not designed to simulate any one type of impact, such as male or female footsteps or to simulate the weight of a human walker. Also, measurements described in this method and ratings based on the results are restricted to a specific frequency range. Thus the subjectively annoying creak or boom generated by human footfalls on a limber floor-ceiling assembly may not be adequately evaluated by this test method. ATSM E1007

B) I would like to achieve NR20. How can I predict the modifications to my ceiling will achieve this?

[thismanysounds]

Sorry to bump the thread. I have found a few of the answers (changed to red), but I am still at a loss for interpreting impact noise measurements. If someone has some insight or can point me in the right direction it would be fab.


Thanks in advance,

Colm

[thismanysounds]

Construction has started!!

Day1.jpg

[Soundman2020]

Here is the current floor construction:
Here is the proposed construction:
Office --> Carpet tile --> Raised access floor (~300mm high, not resiliently mounted) --> Concrete slab (believed to be 150mm) --> 250mm air gap --> 2x19mm PB (hung off 200x50mm joists sitting on wooden studs) --> Inside of rooms.

Ummm... That sounds like a double-floated floor to me! Gulp! Either that, or I'm not understanding what you are saying. Maybe you are using the term "floor" ambiguously? YOUR floor, in YOUR room is what you stand on. Whatever is above your head is your ceiling, even though it might be someone else's floor.

To confuse matters even more, your latest picture shows you pulling out a raised technical floor, and I'm assuming that this is YOUR room, not the room above or below?

So to clarify, what is YOUR floor going to be made of (from the ground up), and what is YOUR ceiling going to be made of (from the inside of your room, going up).

Here is what I have come up with, please correct me if make incorrect assumptions:

It seems to me that your concern here is with impact noise coming down through your ceiling from above. Right? If that's the case, then your very best option is to eliminate or damp that at the source: once it gets into the building structure itself, it is very difficult to deal with. Something as simple as laying a good thick carpet on a good thick impact absorbent underlay. properly laid on the floor above your room, can very likely eliminate all of that impact noise, assuming it is just typical house/office impact noise (footsteps).

If you cannot carpet that floor, could you at least install resilient pads under the raise floor somehow, to break the flanking path?

The final (and least effective) option is to hang your ceiling resiliently, or better still fully-decoupled, and with plenty of mass on it.

Mass Law - Impact noise manifests in the receiver room by the floor re-radiating the impact energy into the air. Once this has happened can the airborne noise reduction of the new leaf be added to the impact noise measurements? (as long as the new leaf is structurally decoupled)

Mass law is only applicable if you do not decouple your ceiling. It applies to single-leaf barriers for air-borne noise. In reality, that floor above you is already more of a coupled two-leaf system, since it already has mass - air gap - mass, in addition to the flanking paths, and we are talking about impact noise in the building structure, not airborne noise . So mass law is not really governing here: it is more likely to be a combination of simple mechanical transmission and MSM law.

The results obtained using Kinetics hangers are significantly lower than predicted by mass law. Possible reasons:
- Lower range hanger so may not provide 100% decoupling
- Air coupling between the concrete and air. This would imply that mass law is inappropriate

Your conclusion is right, but for the wrong reason! You are correct that mass law is not applicable, but not because of air coupling. It's because adding a second leaf to the system (the raised floor above) moves you out of simple "mass law" territory and into "Mass-Spring-Mass law" (or MSM) territory, which is an entirely different ball game.

Mass law for transmission loss calculations goes like this:

TL = 20 log (F * M) - 47 dB
Where: F is the frequency (Hz), and M is the mass per unit area (kg/m²)

MSM law is a bit different. Like mass law, it is frequency dependent, so first the frequency part:

F0=c[(m1 + m2)]^.5 / [(m1 x m2 x d)]^.5

Where:
C is a constant, and is roughly 43 for imperial measurements or 1897 for metric (but it can change, depending on things like temperature, humidity, and the amount of insulation inside the wall).
m1 and m2 are the surface density of leaf 1 and leaf 2, respectively (in kg/m²), and "d" is the distance between those two leaves, sometimes referred to as the "air gap".

That tells you what the resonant frequency is for the system. The isolation then works like this: at resonance, the wall does not isolate at all, and can indeed even amplify sound. At 1.414 times the resonant frequency, the wall starts to isolate, and from there on up the amount of isolation is given by mass law but using an increase of 18 dB per octave, instead of 6 dB per octave for pure mass law alone. That carries on up the spectrum, up to the critical frequency, or "coincidence dip", where there is a sudden drastic drop in isolation, then it increases again at the rate given by mass law above coincidence.

Then to calculate the actual transmission loss for the wall, you need the equations below. There are three basic equations here, since there are three areas of the spectrum governed by different things:

TL = 20log( f (m1 + m2 )) - 47 ...... ( for f < f )
TL = TL1 + TL2 + 20log(fd) -29 ...... ( for f0 < f < f1 )
TL = TL1 + TL2 + 6 ...... ( for f > f1 )

Where:
m1 and m2 are the surface masses of each panel and d is the depth of the airspace between,
f0 is the mass-air-mass resonance frequency and f1 is equal to (55/d) Hz.
TL1 and TL2 are the transmission losses for the individual panels calculated from Equation #2.

and Equation #2 is:
TL = 20log(m * f) -10log(2 * h * w /pi * w * c) - 47

Which you can also simplify back to JUST PLAIN MASS LAW: TL = 20log(m * f)

So for a two-leaf system, things are a little more complicated, due to the resonance of the system.

However, what you are proposing is actually a THREE leaf system, since you are adding another leaf down below: your ceiling, which is drywall.

For any give 3-leaf system, the low frequency isolation will be WORSE than for the equivalent 2-leaf system. This is because there are now TWO air gaps, with two different resonant things going on, plus the overall resonance, plus some other fun stuff.

For a 3-leaf system, the equations for the two fundamental resonant frequencies are:

TRIPLE-LEAF-EQUATION.jpg

Where:
ρ = air density
c = speed of sound in air
m1, m2, m3 are the surface densities of the three leaves
d1, d2 are the distances leaf1 to leaf 2, and leaf2 to leaf3, respectively.

So the isolation of your proposed system will be less than that of a two-leaf system with the same total mass and air gap, and there will be several resonant dips along the way.

B) I would like to achieve NR20. How can I predict the modifications to my ceiling will achieve this?

NR20 is a curve that shows the final sound level inside the room, and depends on a number of factors beyond simple isolation. For example, HVAC noise is a big part of NR curves: the hiss of air coming out the registers, and the rumble of turbulence in the ducts, and the whine of HVAC fans, compressors and other machinery all factor in to the final NR curve. So does the noise made by your own audio gear (fans, etc.) and the speakers too with no signal (hiss, rumble), and even the lights. It's not just about isolation.

But assuming that you build a perfect HVAC system that makes no noise at all, and that your gear and speakers are silent, then in order to decide on what isolation you need in order to hit NR20, you first need to know what the noise is like on the other side. So you'd need to graph a spectrum of the type of noise you are trying to block. Eg, if you have helicopters flying over, then you'd need to record that and look at the spectrum of the noise it makes. Ditto for phones ringing, thunder, rain, radios, people talking, etc. You'd need to look at the spectrum of all of the noises that you are trying to block, average them out, and compare that against the NR20 curve. The difference in level for each frequency band then defines the TL graph that you are looking for. With that in hand, you can then look through the literature to find out which types of walls, floors and ceilings provide the TL curve that you need, then you build it!

Construction has started!!

That's the old raised floor coming out of your room, right? Not going in, I hope! ?

How are you planning to deal with those beams and cable trays up in your ceiling area? The cables obviously come from somewhere else and go to somewhere else, so there are wall penetrations there: that has to be fixed! You CANNOT have wall penetrations if you hope to isolate your room. The outer leaf must be a single, monolithic, perfectly sealed shell, with consistent mass density, and the inner-leaf must be a totally separate single, monolithic, perfectly sealed shell, with consistent mass density. The two must be fully disconnected from each other, mechanically, and there can be no penetrations through either leaf.

Also, if construction has started, then how about if you post your final detailed plans, for us to check, before you go any further?


- Stuart -

[thismanysounds]

Thanks for the response Stuart.

You are right, please excuse my lax terminology. I am indeed worried about impact noise from the office above.

Floor construction:
Soil --> Concrete slab --> Underlay --> Engineered wood floor --> Inside of my room

Ceiling construction: Inside of my room -->19mm plasterboard --> Green glue --> 19mm plasterboard -->200x50mm joists (resting on inner leaf stud walls) -->50mm air --> 150mm concrete --> Raised access floor (~300mm high, not resiliently mounted) --> Carpet tile (glued) --> Inside of office

The photo shows me removing the raised access floor in my room. The same system is used in the office above. We are unable (or at least very unlikely) to alter the raised access floor in the office above. As a result we are stuck with:

The final (and least effective) option is to hang your ceiling resiliently, or better still fully-decoupled, and with plenty of mass on it.

The inner leaf of our ceiling (2x19mm plasterboard with green glue) is held on independent ceiling joists and therefore decoupled.

NR20 is a curve that shows the final sound level inside the room, and depends on a number of factors beyond simple isolation

I understand this, and agree that HVAC noise is likely to be what defines the NR rating in my room. As impact noise is impulsive and therefore likely more objectionable it probably needs to be >6dB below the HVAC to be masked.

Are the MSM and MSMSM formula's you have given applicable for impact noise or are you concerned about airborne transmission? I will measure the weight of the raised access floor panels asap.

Also, if construction has started, then how about if you post your final detailed plans, for us to check, before you go any further?

As always, some details are going to be added/revised/changed, but I have attached the current plan to this post.

[Soundman2020]

Are the MSM and MSMSM formula's you have given applicable for impact noise or are you concerned about airborne transmission?

The predict the resonant frequency for the wall system. Those are the frequencies that the entire wall itself will resonate at, and isolate at, regardless of the source of the sound. That said, impact noise is a special case since the impact itself basically causes the relevant leaf to resonate at it's own resonant frequency. Tap any object, and that object will "ring" at it's own natural frequency. In this case, it is the resonant frequency of the leaf that gets tapped, which is not necessarily the same as the resonant frequency of the entire wall system (hopefully!). The equation for the resonant frequencies of a panel can be calculated as follows.

F(r) = 0.45*vl*t*[(r/w)^2+(r/h)^2]

Where:
vl = the longitudinal velocity of sound in the partition (m/s),
t = the panel thickness (meters),
w and h = panel's width and height (meters)
r = harmonic number (1 for the fundamental frequency, 2,3,4 = harmonics)

That can get a little complex, since you need to know hard-to-estimate numbers to do the calculation. Fortunately, you can approximate it by considering that the wall is acting as a panel trap, and using the equation for that:

f=60 / ( sqrt (m*d) )
m=panel mass (kg/m2)
d=distance between leaves (meters)

That's not very accurate (because your wall is not actually a panel trap!), but it gives you a very rough ball-park idea of what the fundamental resonant frequency will be. If you compare that with the frequencies of the MSM and MSMSM calculations, you'll see roughly where your impact noise lies with respect to the isolation of the wall as a whole, and thus estimate how well the wall will isolate for that impact.

- Stuart -

[thismanysounds]

Hi Stuart,

Sorry your last message flew under my radar. I have tried doing these calculations, but the numbers are a bit strange. I have guessed that M is the total mass per square meter.

Here is my code:

Code: Select all

P=1.225
C=340.3
m1=13.88888889
m2=356
m3=29
d1=0.3
d2=0.25
M=m1+m2+m3
 
 
A = (1/2*m2) * ( ((m1+m2)/(m1*d1)) + ((m2+m3)/(m3*d2)) )
 
B = M/(m1*m2*m3*d1*d2)

 
Falpha = (1/2*pi) * sqrt(3.6*P*C*C) * sqrt ( A + sqrt((A^2)-B))
 
Fbeta = (1/2*pi) * sqrt(3.6*P*C*C) * sqrt ( A - sqrt((A^2)-B))

This gives me:

Code: Select all

A =   2.5254e+04

B =    0.0371

Falpha =   2.5228e+05

Fbeta =    0.9620

Any pointers?

[thismanysounds]

Okay, I have found the source of the equations (Long - Architectural Acoustics).

Fixed some sloppy programming, and am now getting more reasonable numbers. This:

Code: Select all

P=1.225
C=340.3
m1=13.88888889
m2=356
m3=29
d1=0.3
d2=0.25
M=m1+m2+m3

A=(1/(2*m2))*(((m1+m2)/(m1*d1))+((m2+m3)/(m3*d2)))

B = M/(m1*m2*m3*d1*d2)

 
Falpha = (1/(2*pi)) * sqrt(3.6*P*C*C) * sqrt ( A + sqrt((A^2)-B))
 
Fbeta = (1/(2*pi)) * sqrt(3.6*P*C*C) * sqrt ( A - sqrt((A^2)-B))

Gives me:

Code: Select all

Falpha =

   56.9142


Fbeta =

   43.7743

I still have some doubts that this is correct. Calculating each pair of leaf's gives me 21.4Hz and 16.6Hz respectively. This thread (http://www.johnlsayers.com/phpBB2/viewt ... A&start=15) has lead me to believe that a triple leaf wall should have one resonant frequency below and one above this pair. Also, the formula states that Fbeta >Falpha, which is not happing for me.

HELP!!

[thismanysounds]

I was using the constant form insulated walls (43), which I don't think it is included in the 3 leaf formula. This means the individual MAM frequencies were too low. They should be 30.0 and 23.2Hz. Still, my triple leaf frequencies of 56.9 and 43.8Hz are high, and in the wrong order. Any ideas?

Colm