R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: Where does reflection start?  (Read 2826 times)

Tom C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 377
Where does reflection start?
« on: September 11, 2006, 04:46:30 am »

Hello everybody,
I have a very simple question, but can't find any accurate information here or elsewhere, maybe I just need the right keyword to investigate further.

simple example:
  • (normal) walls reflect the lower frequencies, wheras thosefrequencies easily pass (acoustically) thinner material like a desk
  • higher frequencies are reflected by a desk


These 2 facts imply that there is some kind of function that depends on the material (mass, structure, surface, ...) and the frequency to decide when reflection stops and absorption and/or passing through of sound starts.

Any pointers to further reading?
Thanks in advance.
Logged
Tom

.signature failure

jimmyjazz

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1885
Re: Where does reflection start?
« Reply #1 on: September 11, 2006, 11:01:19 am »

You are correct.  For any wall, there is a frequency-dependent parameter known as its "transmission loss".  Transmission loss can be thought of as a measure of how well the wall blocks the transmission of sound.  In other words, a wall with high transmission loss is a good acoustic isolator, because most sound is reflected back into the room and very little escapes.

At extremely low frequencies approaching DC, the transmission loss is governed by the stiffness of the wall.  Starting at DC, the transmission loss drops with increasing frequency until it reaches a minimum value at resonance.  The rate at which transmission loss drops is roughly 6 dB/octave.  Near resonance, the transmission loss is governed almost completely by damping.  Above the resonant frequency, the wall quickly becomes "mass controlled".  In this region, the transmission loss increases by rougly 6 dB/octave.

I should add that this discussion only holds strictly true for a "normally incident" sound wave; i.e., one that hits the wall perpendicular to the wall plane.  At oblique angles of incidence, the mass law starts to break down at certain frequencies, and the wall will exhibit a striking amount of acoustic transmission over a very narrow frequency range.  This is called "coincidence".  It's a very exasperating real-world phenomenon . . . for instance, I recently encounted a situation in which a steel cabinent housing a rotating machine was transmitting much more noise than seemed reasonable using mass law calculations.  As it turns out, the thickness of the sheet metal was just right (wrong!) and there was a large amount of coincident sound transmission through the panels.  The solution?  THINNER sheet metal!  Strange, but true.  (This is why ordinary window glass is as thin as it is -- thicker glass actually transmits more everyday sound.)
Logged

Tom C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 377
Re: Where does reflection start?
« Reply #2 on: September 11, 2006, 02:26:19 pm »

Great, "transmission loss" was the keyword I needed.

Thanks a lot!
Logged
Tom

.signature failure

Ethan Winer

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 571
Re: Where does reflection start?
« Reply #3 on: September 11, 2006, 05:52:27 pm »

Jimmy,

> The solution? THINNER sheet metal! <

Indeed. Construction that improves isolation between rooms makes the bass response inside each room worse. Pro studios need good isolation and good bass response. But for home studios people are better off not adding extra layers of drywall etc unless the additional isolation is truly needed.

--Ethan

franman

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 580
Re: Where does reflection start?
« Reply #4 on: September 11, 2006, 07:45:43 pm »

I will agree with Jimmy and Ethan here.. but I will point out that increasing the mass and stiffness of typical home wall construction (by doubling up the drywall) is usually an improvement in TL and has the added benefit of making the walls more significant boudaries to LF... For us, in rectangular rooms, this makes it easier to predict model response which is generally good, unless of course your rectangle sucks... then you might wish you ripped down that drywall and started by moving the walls.... This is why #1 on our evaluation process is a series of modal simulations... this is the hardest thing to fix with treatments...

HF reflections start at almost any boundary, but LF reflections may not start at the obvious walls, as they may not be massive enough. Takes some mass and stiffness to really be a boundary to LF..
Logged
Francis Manzella - President, FM Design Ltd.
                 - Managing Director, Griffin Audio
fmdesign.com
griffinaudiousa.com

jimmyjazz

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1885
Re: Where does reflection start?
« Reply #5 on: September 11, 2006, 09:11:20 pm »

Fran, do you typically address discrete low-frequency modes with tuned resonators or with broadband bass trapping?

Surely you aren't saying you double the sheetrock as a matter of course in order to accurately predict room modes!  That would be far more costly than just measuring the room response.  Can you elaborate on your thought process here?

Ethan's comment is similar to those I've seen proffered on rec.audio.pro and elsewhere . . . basically, the idea being that well-constructed, heavy, rigid rooms aren't necessarily the right way to go for the best in-room low frequency response.  No doubt the fast majority of installations demand high transmission loss, so it's almost a moot point, but if you were allowed to design a room in which transmission loss wasn't a requirement, would you?

I suppose we're approaching "outdoors", in the limit, which is a ridiculous asymtote, but you get the point . . .
Logged

franman

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 580
Re: Where does reflection start?
« Reply #6 on: September 12, 2006, 06:59:38 pm »

Yes, we do typically design "heavy" inner shells. It really is part of the typical requirement for isolation performance in the majority of the projects we are involved in. Our inner wall systems are typically three-four layer composites using a variety of standard materials. Outer walls vary depending on STC requirements and other issues such as proximity, etc.

If I had a blank slate where STC was not an issue at all, I'm note sure to tell you the truth. We've become very reliant on the ability to predict rectangular control room shells as a basic tenant of "good" listening room design. This is not to say that we don't do any projects with non-rectangular geometry. We certainly do. This is usually dictated by other design concerns such as sight lines or making everything fit!! (a pretty important design concern!).

On the bass trapping issue, with good room modal design we rely heavily on lots of broadband trapping. We use tuned resonators when there are known (predicted) problems or as final tuning devices during in-progress testing. I'm sure there are other ways to go with more membranous wall systems (light weight) acting as LF absorbers and not limiting LF boudaries, but we don't get to build many (any) rooms like this in the real world... In the real world EVERYBODY has space concerns and limitations.

We are always exploring ways to do "free field" tests on our Griffin Monitors, but the outdoors control room hasn't really come up yet with any of my clients...

I wonder if Ethans comment on "thinner sheetmetal" is directed towards a membrane type device that is an efficient LF absorber but this still would reflect high frequency, no??

anybody else have any input on the "free field control room"??
Logged
Francis Manzella - President, FM Design Ltd.
                 - Managing Director, Griffin Audio
fmdesign.com
griffinaudiousa.com

jfrigo

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1029
Re: Where does reflection start?
« Reply #7 on: September 13, 2006, 12:48:29 am »

Quote:

(normal) walls reflect the lower frequencies, wheras those frequencies easily pass (acoustically) thinner material like a desk
*   higher frequencies are reflected by a desk


Actually, low frequencies will diffract around an object like a desk, not pass through. You can't reflect a wavelength larger than the dimension of the object, which is why using a 4x8 gobo (baffle) in the studio won't isolate the bass amp from the sax player's mic on the adjacent side. Transmission Loss and difffraction are distinct, though each accounts for the bass getting through in different circumstances.

Quote:

the mass law starts to break down at certain frequencies, and the wall will exhibit a striking amount of acoustic transmission over a very narrow frequency range. This is called "coincidence".


The coincidence dip will vary with the mass and rigidity of the wall, which is of course what was mentioned with the thinner sheet performing better at a certain frequency than the thicker. As the wall gets thicker and more rigid, the frequency of the coincidence dip lowers. For speech-like frequencies, a boundry of a certain type may exhibit a marked dip, and making it thinner may increase the TL for that frequency, even while making the TL across the rest of the range a little worse. If you are only interested in that frequency, then thinner will work. If you need it to perform across a wide spectrum, thinner may not be the best solution for the system as a whole.

Working in concert with the coincidence dip is the mass-air resonance. This is why three panes of glass is worse than two panes for a control room window. Separating the span into two air-spaces instead of one gives worse performance for the same amount of material, and indeed even if you use more material (mass) in the version with the extra divider. Same with walls. Put your mass on the outside, and don't divide the air space.

There are ways to damp resonances, and also ways to influence them lower or higher, and considering the whole system and the overall goals, you can make decisions that minimize weaknesses and feature strengths. More mass is better than less mass for TL in general, but mass law reaches a point of diminishing returns in required materials, space, weight, and cost of construction. This is why studio designs combine mass and split construction for isolation.

I know this has turned into a long post, but one more important point to make on the subject.

Quote:

Surely you aren't saying you double the sheetrock as a matter of course in order to accurately predict room modes! That would be far more costly than just measuring the room response. Can you elaborate on your thought process here?



As you make a more massive and rigid shell, more low frequency energy is contained in the room. Gypsum board (aka drywall or sheetrock) on studs is a flexural absorber. A typical wall not only lets some low frequency energy through, but also absorbs some. A wall in and of itself is a bass absorber to some extent. As you add mass and rigidity, less is transmitted, and less is absorbed. Therefore, while the predictions inside the room are easier, you also need more bass trapping, or at least better planned bass trapping to create an even response within the room. In the end, I think you'll have a better room overall, but you do need to pay attention to trapping all the more in a room with a "heavy" inner shell.

Whew...

Logged

Ethan Winer

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 571
Re: Where does reflection start?
« Reply #8 on: September 13, 2006, 05:09:00 pm »

Fran,

> I wonder if Ethans comment on "thinner sheetmetal" is directed towards a membrane type device that is an efficient LF absorber but this still would reflect high frequency, no?? <

I said "sheet metal" only when quoting Jimmy. He said it first. Very Happy

From the perspective of in-room acoustics, the best material to make a wall out of is probably heavy cardboard. Shocked

--Ethan

L_Tofastrud

  • Newbie
  • *
  • Offline Offline
  • Posts: 29
Re: Where does reflection start?
« Reply #9 on: September 13, 2006, 05:22:08 pm »

When designing hi-fi rooms back in Norway I sometimes would tune the walls' panel resonanse to that of the lowest mode that would develop on that surface.  It definetly lowers the Q of the mode and evens out the response.  As several people has pointed out it's not exactly good for sound isolation though...

Lars T
Logged
Pages: [1]   Go Up