now imagine a decoupled wall, let's just take a double wood stud wall for one example
the same thing that happens to a single panel happens here - sound hits the drywall on one side of the wall, and vibrates it. the resistance to vibration is JUST LIKE ABOVE, mass -minus- resonance
however, because the two sides of the wall aren't physicall connected, the vibration can't immediately transfer to the other side. pretty straightforward, that.
The thing about decoupling that many people don't grasp, is that the air in the cavity (and the force of the resilient channel in an RC wall or the force of an RSIC clip in an RSIC wall) behaves like a spring.
And with this spring force, and the mass on either side of the wall, you get a new resonance, called a mass-spring resonance. 3 frequency ranges are thusly created:
1. At this resonance point, the performance of the wall won't be so good, it will be WORSE than if you hadn't decoupled the wall
2. well below this resonance point, the performance of the wall will be the same as if no decoupling were present. At very low frequencies, performance will be THE SAME as if you hadn't decoupled the wall
3. well above this resonance point, performance will improve DRAMATICALLY. well above this resonance point, no other method is as effective at improving the performance of a wall as decoupling.
so, this mass-spring interaction results in this situation:
1. sound shakes the inner mass
2. around the resonance, the spring behavior amplifies this vibration, and MORE vibration occurs on the other side
3. well above the resonance, the vibration cannot transfer and performance soars
The attached pic shows this:
1. at very low frequencies the air + resilient channel behaves as a "stiff" mass
2. at the resonance point performance suffers considerable
3. well above the resonance point performance soars
so it is obvious from this assessment that the most important thing we can possibly do for a decoupled wall is lower the resonance frequency. This lowers the weak-spot to a frequency that's less offensive (it's harder to hear 30hz than 60hz), and it moves the frequency where the wall starts to get really, really good down.
1. use as much mass as possible on both sides of the wall
2. use as deep an air cavity as possible
3. use insulation
4. don't use resilient channel
will always help you steer the resonance point down in frequency. resilient channel has too strong a spring force, and keeps the resonance much higher in frequency than other methods, making it a worse performer below the STC frequency range (STC only goes to 125hz)
The graph below outlines this perfectly. The enormous gains at mid/high frequencies caused by the combination of decoupling + absorption are so immense, that you would have to utilize literally 40-60+ layers of drywall to match them.
But, at low frequencies, the resonance makes it WORSE than no decoupling.
I spliced some NRC data into this graph as they have a higher-limit-facility and to use the AUdio Alloy data would short-change the actual gains wrough by decoupling. I used Audio Alloy/Orfield low freq data as all that data is same lab/same time, and makes for a better comparison.
The overall gains due to decoupling sort of flatten out at mid/high frequencies for a couple of reasons
1. sound transferring through the channel, or through other slight structural connections
2. sound making it through the air / limitations of absorption in the cavity
So that's the awesomely potent, but sometimes risky, second principle of sound isolation - decoupling. awesome in potency well above it's resonance, risky because if you plop that resonance at a problematic frequency, you'll find yourself scratching your head and wondering why your uber-high STC wall still lets hoards of bass through.
And that's as good of a summary of how decouling works as you'll find anywhere, provided i explained it clear enough to make sense.