absorption coefficients

[MatthewDougherty]

Has anyone used REW to calculate sound coefficients for walls?

Generally, acoustical theory puts sound into three categories when striking a surface: reflection, absorption and transmission.

When analyzing an empty room, REW will tell you a lot about the reflectivity of the room, but nothing directly about absorption and transmission..

If free standing acoustical panels in the room are used, transmission is still energy in the room, absorption is the only mechanism removing sound energy from the room.

In the case of walls, absorption and transmission are energies leaving the room. I am ignoring ambient sound outside the room entering the room through the walls.

It seems like one could devise a test using REW similar to this youtube video, to get some reasonable coefficients by placing a reference mic outside the room in proximity to a speaker focused closely to the wall, as in the video.

 

[Igor Kurilovich]

Це дуже цікава тема! Для акустичного оформлення приміщень часто не вистачає певних параметрів доступних акустичних матеріалів. найчастіше мені не бракує питомого опору потоку (Па.с/м2). Якщо це стане можливим з REW, я буду щасливіший))

 

[MatthewDougherty]

I was thinking using a 1ft diameter HVAC duct pipe, a studio monitor with a known frequency graph, and two reference mics (one in front of the material, one behind).

Using several materials with known absorption coefficients, get their profiles using REW. Then use materials with unknown absorption coefficient profiles. Guesstimate the unknown coefficients using the known materials.

My next step is to talk with the acoustic engineers at treble.tech on their opinion. Trying to do acoustic room simulations with treble software for dome room; need the coefficients to close the gap between theory & empirical. Recent simulation tests varying the coefficients dramatically changes the RT60, impulse and clarity of room.

 

[Igor Kurilovich]

Я готовий понести витрати, я можу виготовити щось на кшталт "трубки Кундта", я готовий купити ліцензію на багатоканальні вимірювання... Чи є якась література, яку я повинен прочитати, перш ніж почати? ))

 

[MatthewDougherty]

the google translate did not completely make sense.

I'm willing to spend, I can make something like a "Kundt tube", I'm willing to buy a license for multi-channel measurements... Is there any literature I should read before I start? ))

I think you are asking the same question I have: will this produce useful results?

Unfortunately I am not an acoustical scientist or engineer. What I suggested is probably uncharted, and the accuracy is a guesstimate. The key is empirical testing with known materials with known absorption coefficients. If these do not roughly match, it is a failure. It is a fishing expedition, will it sink the bank account or waste my time? After I talk to treble, who are acoustical scientists and engineers, I will report back. Was hoping someone who has done this (success or failure) might comment.

regarding literature. There are international standards for this, and some youtube videos describing how this is measured.

 

[MatthewDougherty]

quite a stack of speakers. How loud do they get?

In the 1980's I was in bldg 49 at JSC/NASA, the vibration & acoustic test facility, originally built for Saturn 5 rocket tests. The many horns were 30-50ft long. It could get to 200+ db. When rocket size doors were open birds would get in. If they did not get out when the doors closed, it was fatal, and birds had to be swept up after.

 

[MatthewDougherty]

1) Research thesis using REW with an impedance tube

https://opus4.kobv.de/opus4-filmuniversitaet/files/267/FaciusBA.pdf

2) A related REW post
ttps://www.avnirvana.com/threads/using-rew-to-measure-sound-absorption-in-an-impedance-tube.3569/


3) A previous attempt indicating a problem using REW

"Measurement of the Absorption Coefficient for Destructive
Interference Absorbers Produced by Additive
Manufacturing" report on researchgate

Initially, the plan was to collect the frequency response of the microphones with the freeware acoustics software Room EQ Wizard (REW) and then to calculate the absorption coefficient in MatLab or Excel following the guidelines of the EN ISO 10534-2 (BSI, 2001) integrated with other online resources (Wikibooks, 2017). However, a preliminary test – conducted on the 10th of May in the same studio – revealed problems that had not been considered. Since REW is not designed to measure transmission, reflection or absorption of materials, it can take only one input signal at a time, while this kind of measurements requires two input signals.

To try to fix the problem, two separate measurements were taken, one with channel 1 and the other with channel 2 as input. Unfortunately, the random misalignment between one measurement and the other affected too much the transfer function calculated between the microphones, generating an error that grew with the calculations and, in the end, caused a totally corrupted outcome.

 

[John Mulcahy]

A previous attempt indicating a problem using REW

Shouldn't be a problem using multi-input capture.

 

[MatthewDougherty]

That is how I read their paper, they should have upgraded to multi-input capture. Why they did not, and went with a more expensive method, is a mystery. If they had used REW rigorously, comparing it to Holmarc results, the paper would have had a wider useful impact and provided broader testing design insights. As I pick apart the paper I do expect to get some useful testing ideas.

I am surprised no one has published a paper on material absorption coefficient testing using REW. Compared to the YouTube video above, REW should deliver far superior precision and more credible results. On the practical side, being able to go to a construction site and measure actual coefficients, I would think it might have commercial value.


Since my Dec 31 post, I have decided to go with three mics, two on one side of the wall as in a typical impedance tube, and one mic on the other side of the wall to get transmission data. The testing parts have begun arriving, and assembly shall begin in a few days. An aside, I plan to collect temperature data, over the next ten days Houston is expected to have a 60 degree Fahrenheit swing; will that noticeably affect real acoustical results?

As I start simulating the dome acoustics with Treble, I am realizing getting the coefficients correct is critical to converging REW and Treble results. My first simulations are indicating there are a lot possibilities in resolving acoustical objections noted by others in my previous REW thread, as well as addressing a wider negative consensus at other home theater websites as noted by architects and engineers. The simulations indicate turning the dome into an anechoic chamber, if desired. Considering the uniformity of a dome shape, I wonder why such rooms are boxes instead.

I am hypothesizing the hemispherical longitudinal wooden frame supporting the gypsum/plaster, varies the coefficient values along a polar axis as a function of height. A coefficient variation similar to going from gypsum/plaster mounted rectangularly on 16" studs at the lower dome areas, to gypsum/plaster on 2" plywood at the top where the wood frame converges.

 

[MatthewDougherty]

a useful reference

Deshpande, S. and Rao, M. (2014) Development of a low-cost impedance tube to measure acoustic absorption and transmission loss of materials. In: 121st ASEE Annual Conference & Exposition, June 15-18, 2014, Indianapolis USA. Paper ID#8776

The diameter of the tube seems to be a major factor in matching the DIY to the PRO tube results.

"Generally, the tube cross-section (diameter) must decrease if the maximum frequency of interest fu increases."

 

[MatthewDougherty]

This contains matlab code for computing the absorption coefficients using two microphones.

https://ideaexchange.uakron.edu/cgi/viewcontent.cgi?article=1707&context=honors_research_projects

IT_absorption/absorption.m at master · nathanpalella/IT_absorption

Preliminary Matlab function to obtain the normal absorption coefficient in an impedance tube - nathanpalella/IT_absorption

github.com