Sabine coefficient (alpha) and NRC

Don Gato · Post by **Don Gato** » Sun Sep 07, 2003 10:44 am

Who can I calculate Sabine coefficient (alpha) from Noise Reduction Coefficient (NRC)?
Many manufacturers of acoustic materials provide absorption data as NRC, but the software I'm using (CARA) needs sabine coefficient data.
I found this in "Master handbook of acoustics" by Everest:
Sabine coef = -log e(1-X)
where X=energy absorption coeff (based on geometric mean reflection coeff.)
Everest does not refer explicity to X as NRC, but I'm assuming NRC = energy absorption coeff. (Am I in the right path???

)
With this assumption there's a problem when NRC >1 (very common in manufacturer data at certain frecuencies), as the log of a negative number is undefined. Your help is greatly appreciatted.
Thanx. Paco.

knightfly · Post by **knightfly** » Wed Sep 10, 2003 6:09 pm

Paco, if I'm not mistaken Everest also covers the phenomenon of coefficients greater than one. These numbers greater than one happen because the size of the samples that are tested for absorption, allow for extra absorption caused by the edges of the sample. Some companies who manufacture absorbent materials choose to "round down" all results so that the highest claimed absorption coefficient is equal to one, making the results more honest. Other companies leave the results un-factored, because it makes their products sound like they are more absorptive.

The simplest way I've found around this problem is to reduce all figures for a particular material by the amount that the highest number exceeds 1.0- this comes out close enough for calculations to work within the formula limitations.

It's been proven that the same square footage of absorbent cut into smaller pieces has more absorption, so there will be some variation from published figures depending on the size of absorbent patches used. I've not seen any concrete figures on how much you should compensate relative to patch size however - All I know is that figures are normally based on samples that are 8 feet by 9 feet, and still are sometimes as high as 1.20, maybe even more... Steve

Don Gato · Post by **Don Gato** » Wed Sep 17, 2003 3:16 am

Thanks Steve for your comments, sorry for the late reply. You're right, Everest mentions somewhere about absorption coeficients > 1, but I never though it would be something in the neighborhood of 1.3 (check published NRC by Auralex and such, done by Riverbank laboratories). Remainds me a Simpson episode, when this big time criminal is running for major of Springfield, the newsman announces the results: "Bob Patino (that's his name in Mexico): 100% of votes, Mister X (former major): 1% of votes" Continues the newsman: "this results may have an error of +/- 1%". So we have a more than perfect absorbers, thats what numbers say.
I found an interesting thread about test & measurement standards, they mention about this interesting measurement... freak?

. According to J. Szymanski (chief acoustic of Auralex), this may be due to the nature of the absorption test (ASTM C423) that measures differences in rate of decay (dB/second) and not directly % of absorbed energy.
You may find it in:
www.studiotips.com go to acoustic mailing list > list archives > thread 8760.

In the end, I think "normalizing" the NRC is OK, cause even with all the strict measurement standards, acoustics has a part that is still an art (I'm not neglecting science).

What you think?

knightfly · Post by **knightfly** » Wed Sep 17, 2003 1:22 pm

"In the end, I think "normalizing" the NRC is OK, cause even with all the strict measurement standards, acoustics has a part that is still an art (I'm not neglecting science). What you think?"

I agree - my own intent, once I actually get a "place of my own", is to concentrate on sound proofing first, only building in bass trapping and places for adjustable treatments. One thing that will take some thought is my hope to set up switchable acoustics between stereo and 5.1 mixing.

I'm about to order the ETF package which should allow a somewhat scientific approach to tuning - I think I can get close with calculations and design, but will probably find the last "tweaks" to be decided by testing and finalized by listening... Steve

Don Gato · Post by **Don Gato** » Sat Sep 20, 2003 7:56 am

Am I the only one experiencing problems to access to the site?

I tried one and a half days before succeeding … Hey, Steve, seems like we’re gonna be in the same boat, I’m just waiting for my paypal approval to buy the ETF stuff, it looks quite robust & also somewhat didactical oriented. I’m planning using it to design & tune the church’s recording studio. Little bit OT, but… what mic & preamp are you plan to use? I ordered a Beahringer ECM8000, the most cheapo measurement mic I can find, to use along my TASCAM US122 preamp/card. Nothing fancy at all, but surely I may start getting quantitative measures (as opposed to simulation data that is more directional or qualitative). It’s a good way to compare all assumptions coming out of simulation (like sabine coefficients for each material), vs. reality. I’m burning to get started comparing result. After all the numbers, the supreme test will come... Does it sound good to your ears? Time will tell...
Paco.

knightfly · Post by **knightfly** » Sat Nov 08, 2003 10:31 am

Hey, Paco - Somehow this got lost (by me) - if you're still around, sorry for the big delay.

I don't buy Behringer gear - not because I'm a snob, but because of other reasons -
They appear to me to be real close to thieves when it comes to design -
Their use of low cost components tells me there is no way I can get two identical channels without extreme luck -
I don't like having to buy the same thing twice, once to save money and the second time to actually get what I thought I'd already paid for.

As far as using things with ETF, it apparently isn't very critical - some parts of the audio chain can be "nulled out" with ETF's profiling, so it works fine with basic soundcards, etc - plus, a lot of room problems are so gross that a 6 dollar "freebie" mic that comes with the sound card would show them up.

My own ETF plans are still just that, plans - I've been so busy lately I've not had time to even check my Paypal account, much less the rest of it. When I do, I intend to get their entire package - mic, preamp, ETF and all the extensions. For portable testing, I'll also need to get some powered nearfields (I use only passives in my studio) so that will add another chunk of money to the cost.

For actual recording, I've gotten really good results with Audio Technica's mics - I have a 4033, two ATM 25's, a Pro-25 so far - looking at a pair of their 3035's and a pair of their SD condensers (forget the #) for overheads.

One preamp I want to audition when I get time is Art's Digital MPA - I've read a couple of quite good reviews, and it's certainly versatile from the look of all the controls, not to mention up to 192 kHz sampling rates if you want.

Oh well, enough "Gear lust" - there's plenty of boards to visit for that... Steve