General construction points

Hi, My first post. Apologies if this in the wrong place. I've been lurking for days reading here, and it's all fascinating. I'm itching to get back into recording (after too many years of distractions). I have a copy of the Rod Gervais bible and noticed that on page 63 of the 2010 revised edition, that double walled 2x4 stud walls with 2x 1/2" plasterboard insulated with a 1" air gap can give a STC of 63dB (approximately). This got me thinking (along with all the 'room within a room' theory). Why not extend the same technique to the ceiling? This has probably been done/debated/discounted before, but I ask the question anyway. Attached (subject to the image not violating forum rules) is a sketch of what I mean. The idea (in principle - ignore the angles/caulking and other issues) is that what works for walls works for ceilings. You have a 'room within a room within a room' - the structures are not mechanically linked. Further to that point, I have two areas of open land in the back yard of sizes 4 x 6m and 4.6 x 5.6m. They are separated by a communal access driveway (couple of metres). My intention is to build a control room on one patch, and a live room on the other. They will be linked by windows facing each other (three panes, one angled), and probably a steel cable at high level holding many wires. Since I'm outdoors and the UK climate is hardly that of a Syrian desert, what in your opinions, would be the effect acoustically of adding an exterior ply sheet to the plasterboard + some roofing felt (all walls) for weatherproofing? Adding mass is good I understand? Thanks for reading, and I look forward to your comments. Cheers, Dave

welcome! yes the same techniques can be used in the ceiling though often times a shed build will have an overhead space for a slanted roof and air flow for moisture control. what you've shown in your design could be considered an inside-out ceiling which if you have a larger space open overhead for the roof would be a good start along with the necessary moisture proofing.

The ceiling area as all areas that are involved in what is the potential air cavity, are always involved in this assembly...nothing new under the sun.

Thanks guys. That answered a lot of my questions! I will do some more sketches and think about a roof and concrete base. I think I will also build a ~1m^3 box and put a small bookshelf speaker inside it to test for soundproofing using a sound meter and some test tones. A comparison with open air readings and enclosed readings using test tones ought to be useful. Thanks again, Dave

What is it you are trying to build, maybe we can flatten the learning curve... John's manual @ http://johnlsayers.com/Recmanual/index.htm is required reading around here as well as the other books you may have.

xSpace wrote:

What is it you are trying to build, maybe we can flatten the learning curve... John's manual @ http://johnlsayers.com/Recmanual/index.htm is required reading around here as well as the other books you may have.

Brien, You can be sure that I have read all those pages many times over! I appreciate your input. John's STC page does not agree well with Rod's book data so I will experiment without having to build anything expensive at this point. I have on order a cheap +/- 1.5dB accuracy SPL meter. I'm going to make an airtight box with 2 x 4s and plasterboard as per my first post sketch to enclose a small bookshelf speaker (a JBL f3 = 30hz to 20k). The purpose of this experiment is to prove the sound proofing technique described is something around STC 63. Assuming it is something near to that, I'll start pinging control room design sketches onto here. Then maybe a live room. To summarise: 1. Put bookshelf speaker on concrete floor (probably on a foam pad). 2. Place SPL meter at 1m from speaker. 3. Play test tones of 20Hz, 40, 80, 160... n# octaves...16kHz 4. Take dB readings at tones in (3) 5. Graph response readings. 6. Build small box with all walls of design in first post but without a base (all joints caulked). 7. Place box over speaker and caulk to floor. 8. Repeat steps (1) to (5) with meter in different positions to obtain a mean reading (at 1m from speaker) 9. Generate STC data by subtracting graph (8) from graph (5). 10. Line exterior of box with ply and roofing felt to simulate an outdoor environment. 11. Repeat (8) and (9) to find impact of (10). This will just serve as a test rig for now so I can evaluate sound-proofing. The box will be designed with the golden ratio technique and be large enough to negate infinite baffle resonances. I'm well versed in all that stuff as I design loudspeaker enclosures. When I'm done with that experiment, I'll post the results. Sound like a plan? Cheers, Dave

That work has been done...and what you are trying to do will not be near as accurate: http://epaper.buildthisroom.com/wall-ma ... trock.html and: http://epaper.buildthisroom.com/wall-ma ... -mass.html This is the data that is real and should save you a lot of heartache...this is basically a bookmark page of mine.

EDIT: I misunderstood the tests facility set up that prefaces the prototypes. The rest of this message is drivel. Brien, Thanks once again. You have just saved me from buying £50 of timber (2x4 is £1.40 per metre here) for a pointless experiment. I must learn to do thorough literature searches before embarking on projects... One thing I can't understand/believe is something that I modelled. Perhaps you can help out. My current music room is on the second floor (one up from ground - in the UK its known as the first floor...). It's small (11'x7'x8'). One wall is brick (single course) two are solid stone (around 400mm) and the other standard plasterboard/2x4 stud. The floor is carpet/underlay/tongue and groove floorboards/6x2 joists. The ceiling is 2x4 + plaster. The speakers are customised T/L and go flat from 25Hz to 20kHz - not in that room though(!). So referring to the STC in Wyle WR 73-5R.pdf prototype H wall (pages 137 et seq in the pdf, original pages -124- et seq) I do the following: 1) Open Sergeant Pepper Beatles track up in Adobe Audition, play and crank up the volume so it nearly deafens me. Lots of bass and a fair spread of everything else. 2) Open up a 30 band equaliser and mimic the measured curve on page -125- 3) Play the track again - can hardly hear it. Leave the door open and go downstairs. All I can hear is the bass. Is this a valid experiment? For an STC of 43 there seems a massive loss (good!) The control room (if I build it) will be on a concrete base some 10m from any dwelling. If the experiment is correct, I will have ample sound insulation - I think.... Cheers, Dave

"Is this a valid experiment? For an STC of 43 there seems a massive loss (good!)" No sir. What you may be experiencing is the reduction in sound purely from the distance you are moving away from the speakers and the hard boundaries that sound encounters when trying to get back to your ears. The stc scale is used for the frequency of human voice so that does not take lower frequency into account...and it does not take higher than human speech frequency into account either. The testing methods used are done in a lab by scientists...which I have not one of the former or am not one of the later:) Transmission loss is the name of the game in developing wall assemblies that isolate sound well since the construction of these considers the low frequency first and taking care of the low frequency will always take care of the high frequency, in terms of isolation.