Clarity C-80 and Impulse response - Home Theater Forum and Systems - HomeTheaterShack.com

Old 01-04-13, 03:43 PM Thread Starter
Shackster

Join Date: Nov 2012
Location: Grenoble
Posts: 9
Clarity C-80 and Impulse response

Hi to you all,

I've used REW for a few weeks now regarding an acoustic analysis of an amphitheater in my university. I arrive to some good result but I have some hard time to extract information from my graphs. My goal is to compute the clarity C-80 which is ten time the log(10) of the integration of energy on the 80 first milliseconds divided by the integration of energy on the rest of the time [80ms to infinity].

C-80 = 10 log( ∫[0,80] E(t) dt / ∫[80,∞] E(t) dt)

My guess was to take the impulse response of my amphitheater, then as my schroeder integral was quite linear I approximate my decreasing of intensity by

Lp(t) = - at,

with Lp in dB intensity in function of time. Now I remember my formula

Lp = 20 log (P(t)/P0)

with P0 = 2*10⁵, then I extract P(t)²

P(t)² = 10^( -at /10) - P0²

which is the energy E(t), so finally I integrate (I use wolfram Alpha) (b = a/10)

∫ [0,t]P(t)² dt = [-t^(-bt) / (b log 10) + P0²t] [0,t]

With a = 28,0 dB/S my final value for C-80 is 13,5dB which, first, is huge and then I have lots of difficulties to find the meaning of the values of C-80. I look up lots of stuff but it was really confusing and unclear. I know that it should be between +/- 3dB but I couldn't find what does this number tells us and does this value apply for a room made for speech. Finally I did the same with another amphitheater which is really good acoustically speaking and my value for C-80 was - 13dB. So I wonder...
One of my worried is that my approximation made the result to far away from reality and I don't now what to do otherwise.

I upload the two impulses responses that allow me those computation. The red one is the main amphitheater with a C-80 of 13,5 dB the green one correspond to the C-80 of -13 dB

Well thanks for any of your thought.

Thaumas
Attached Thumbnails

thaumas is offline

Old 01-05-13, 12:26 PM
Shackster
jamin

Join Date: Jun 2006
Location: moot point
Posts: 6
Re: Clarity C-80 and Impulse response

I’ll go out on a limb and take a crack at this. Remember that the coffee may not be working and my limbs often crack!

For 20Log(P(t)/Po)=C*t

I get (P(t))^2 = (Po)^2*10 ^ (C*t /10)
And letting
K=(Po)^2
E(t)=(P(t))^2
D=C/10

Yields
E(t)=K*10^(D*t)

With the integral of E(t) dt showing as
K*(10^(D*t)) / (D*Ln(10))

And since C_80 is the ratio of integrals the constants
K
And
D*Ln(10)
Both drop out leaving only

[10^(D*(.08)) - 10^(D*(0))] / [10^(D*(infinity)) - 10^(D*(.08))]
to be evaluated.

Realizing that D = C/10
And C would be (–a) in your example.

This may yield a better result.
Or, I could have missed completely !!

PS. FWIW Po is equal to 2E(-5) not 2E(5)

plug in to play
jamin is offline
Old 01-06-13, 06:45 AM Thread Starter
Shackster

Join Date: Nov 2012
Location: Grenoble
Posts: 9
Re: Clarity C-80 and Impulse response

Thanks Jamin! I see that your limbs didn't crack! I finallly get the mistake now. And I find a much better result around C-80 = -1,7 for the red graph (the main study) even though the other one is still a mystery C-80 = 20.
Now that the computation are fixed I would need advice about the meaning of the result if you anybody know something about it'll be great by the way I still wonder about the aproximation :/

Thaumas
thaumas is offline

 Bookmarks

 Tags c80 , clarity , impulse , response

Message:
Options

## Register Now

Random Question
Random Question #2

User Name:
OR

## Log-in

Human Verification

In order to verify that you are a human and not a spam bot, please enter the answer into the following box below based on the instructions contained in the graphic.