Natural Science Forum / Physics / Acoustics / December 2004

The 8 dB rule is still there. In my opinion it is vital whenever the 100
Hz band is not considered (as ASTM E413 still has it). We in ASTM E33
have argued this now for decades. But please be aware that mostly all
labs worldwide now procure the 100 Hz, and often 80 Hz, TL values in
their fundamental test. All you need do is request that from the
manufacturer or the laboratory, with manufacturer permission. So you can
calculate Rw yourself.

The main reason this remains so is that we acousticians in countries
having wood stud (and now metal stud) interior construction know well
that the residual partition weakness is the transmission of sound below
100 Hz.

For that issue also, be aware that ASTM has had E1332 since 1990 the
"Outdoor-Indoor Transmission class (OITC)", which is, in place of the
dog-leg rating curve, the A-Weighted result of imposing the spectrum of
a composite transportation noise spectrum onto the specimen TL values
over 80-4,000Hz. This MUCH better represents windows and even walls in
music-related noise control. The ISO committees are looking at this
concept for a "better" rating than Rw, especially for building facade
materials.

"As-Is" works for STC vs Rw to about a plus or minus two point 'accuracy'.

HOWEVER, when considering floor/ceiling impact sound ratings, be aware
that the North American "IIC", though using the same test data and
frequency range, is the inverse of the ISO "L" values. (But also, The
IIC method applies the same old 8-dB rule to help overcome the same low
frequency weakness of this rating method).
That is, an "IIC" of 45 is an ISO "L" value of 65 (obtained by
subtracting IIC from 110).

Angelo campanella

there is another correction in Rw, but not one that i'm familiar
with...  i should buy the new standard.

the 8dB rule is a very nice thing about the STC rating, if you sit down
and do an analysis against another means of assessing a wall (dBA or
ISO equal loudness or whatever), that 8dB rule often demonstrates it's
value in improving the accuracy of the STC system.

now, STC is fairly useless as a rating for many applications as it only
works  from 125-4000hz, but it's a remarkably (to me, remarkably) good
system for matching the dBA log-calculated single number rating for a
flat noise curve in the 125-4000hz band.  i presume (based only upon
this observation, i've never seen a formal reference to this effect)
that this was the intent of those who designed the STC standard.
considering the "by-hand" nature of the system, they did good work.

but today...  today it's easier to calculate things logarithmically ala
the OITC standard than to utilize the STC system.

Brian