The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Calculating room modes for an irregular shape
Old 16th December 2010
  #61
Gear Guru
 
Ethan Winer's Avatar
 

Lightbulb

Quote:
Originally Posted by bpape View Post
IF (big if) the nulls you're seeing are truly modal in nature.
Modal and non-modal peaks / nulls are the same thing. The only difference between a modal frequency and a non-modal frequency is that the acoustic wave for a modal frequency happens to fit exactly within a pair of opposing boundaries. We all know there's a deep "modal" null halfway back in a room. But if you remove the front wall of the room and leave everything else as is, you get the same "modal" null at the same frequency even though there are no opposing walls!

--Ethan

________________
The Acoustic Treatment Experts
Old 16th December 2010
  #62
JWL
Lives for gear
 
JWL's Avatar
 

Quote:
Originally Posted by recordinghopkins View Post
What I am gathering is that even computers have a hard time calculating modal resonances of a room with more than 6 dimensions, yes? Much less actually doing it with a handheld calculator and pencil.

I suppose I was hoping to learn some new formula or matrix that allows the user to input the results of F=s/W for each dimension of the room.
Bingo. That's what I was trying to convey earier in the thread. There is no simple linear equation for this stuff, Ethan's post (a few up from this) digs deeper on this point.
Old 16th December 2010
  #63
Lives for gear
 
recordinghopkins's Avatar
Quote:
Originally Posted by jwl View Post
Bingo. That's what I was trying to convey earier in the thread. There is no simple linear equation for this stuff, Ethan's post (a few up from this) digs deeper on this point.
I saw those posts, but took it to mean that it is just really difficult and time consuming, not nearly impossible. I get it now why there was such a huge initial response to this thread topic! hahaha yea, I am 62 posts behind.
Old 16th December 2010
  #64
Gear Maniac
 

Quote:
Originally Posted by recordinghopkins View Post
Here's what I was thinking: nulls and summation points result as a wave is reflected off of a boundary and combines with the source wholly or partially out of phase. My thought is, if you can trap the wave near the boundary, then it will minimize the effect. There are no modes if there are effectively no boundaries, right? Now clearly, it would take an impractical amount of absorption to trap 60Hz, which is why I was thinking resonators might do the job more effectively. If that is wrong, then a clear simple explanation will go a lot farther than talking over my head.
That's not too far off. The acoustic behaviour in rooms with simple, regular geometries can be accurately modelled using ray tracing techniques, or the image source method, which is in essence the same thing. In the image source method one source represents the direct signal and additional sources are added in locations that are outside the boundaries of the room to correspond to the reflections from surfaces on the paths to the listening position, the distances of the reflected images provide the correct time delay such that superposition of the original direct signal and the signals from each image source (removing the surfaces) produces the combined signal that would be measured at the listening position. The amplitude and (in more sophisticated models) spectral shaping of the image sources corresponds to the absorption characteristics of the boundaries on the associated path.

To get a perfect cancellation from a single surface requires a reflection equal in amplitude to the original signal, which doesn't happen in practice as any surface has some absorption. At frequencies where the path length difference gives odd multiples of 180 degrees of phase shift the nulls occur. Increasing the absorption of the surface reduces the depth of the null by reducing the amplitude of the reflection, a perfectly absorbent surface (or no surface at all, to use your example of removing the wall) means no reflection and no nulls.

The same modelling technique reproduces the modal resonance behaviour of the room, those correspond to closed paths in which each leg of the path has an integer number of half wavelengths at the modal frequency. The absorption of the surfaces on the path (at the mode's frequency) and the path length determine the rate of decay of the mode and the heights of the peaks/depths of the nulls.

Getting accurate results from image source models requires quite high orders of path tracing (and hence many, many, many image sources) which is computationally very intensive. Numerical solution of the differential equations governing the behaviour, as used in finite element analysis, is more practical. If your aim is to identify resonant features in a space that already exists, however (modelling is most useful before spaces have been constructed) then you are best making measurements and analysing those measurements to identify the problem areas.
Old 16th December 2010
  #65
Gear Guru
 
Ethan Winer's Avatar
 

Lightbulb

Sorry I missed this earlier:

Quote:
Originally Posted by recordinghopkins View Post
it would take an impractical amount of absorption to trap 60Hz, which is why I was thinking resonators might do the job more effectively.
Porous absorbers can be effective down to lower than 60 Hz if they're thick enough and you have enough of them. The Before / After graphs below from my Hearing is Believing video show a real improvement all the way down to the lowest peak at 40 Hz. Now, this is with a lot of bass traps. But the traps are "only" six inches thick, and they're not even straddling corners.

--Ethan

________________
The Acoustic Treatment Experts



Old 16th December 2010
  #66
Lives for gear
 
johndykstra's Avatar
 

^^^^^

It sure would appear that a lot of the nulls have been minimized as well.
Old 16th December 2010
  #67
Lives for gear
 
recordinghopkins's Avatar
you guys are a wealth of knowledge

*soaking it in*
Old 17th December 2010
  #68
Gear Guru
 
Glenn Kuras's Avatar
Quote:
Originally Posted by Ethan Winer View Post
Sorry I missed this earlier:



Porous absorbers can be effective down to lower than 60 Hz if they're thick enough and you have enough of them. The Before / After graphs below from my Hearing is Believing video show a real improvement all the way down to the lowest peak at 40 Hz. Now, this is with a lot of bass traps. But the traps are "only" six inches thick, and they're not even straddling corners.
s[/url][/b]



Out of all the articles you have done I score this one WAY up at the top. thumbsupthumbsup
Old 18th December 2010
  #69
Gear Guru
 
Ethan Winer's Avatar
 

Lightbulb

Quote:
Originally Posted by Ethan Winer View Post
Modal and non-modal peaks / nulls are the same thing. The only difference between a modal frequency and a non-modal frequency is that the acoustic wave for a modal frequency happens to fit exactly within a pair of opposing boundaries. We all know there's a deep "modal" null halfway back in a room. But if you remove the front wall of the room and leave everything else as is, you get the same "modal" null at the same frequency even though there are no opposing walls!
I expected heads to explode after I wrote this, especially the last sentence. Yet all I hear is crickets.



Nobody has even one question? heh

--Ethan

________________
The Acoustic Treatment Experts
Old 18th December 2010
  #70
SAC
Registered User
 

Why? heh (...I know, I know!...just kidding...)
I am rather surprised as well...the silence was deafening...

The effective resonate 'pipe' still determines the tuning.

What changes is that at the open end you now have a node(null), rather than the anti-node (peak) that occurs at a 'closed' termination.

This is a nice 'bonus' that can be exploited in open or partially open quasi-greatroom spaces in a home or club.

{As far as "modal and non-modal peaks being the same thing" requires a bit of qualification and more than a fair amount of reductive over-simplification. In the sense that they are both a result of acoustic energy and the summation of reflected energy, they are 'the same'. But in terms of how this reflected energy sums (superposes), they are not. A modal standing wave is a result of constructive reinforcement in the form of an in-phase reflection. Comb filtering of specular waves results in destructive interference as a result of the superposition of energy that is not in-phase. And such destructive interference can result in both nulls and peaks - neither of which reinforce the original incident energy magnitude - and in this respect, they are not at all the 'same' in their resultant behavior, despite their all beginning 'the same' as energy that is reflected....

Unfortunately, in terms of the practical ramifications, simply reducing the phenomena to their most base level and stating that they are both a result of reflected energy is a rather nebulous criterion, which is anything but determinant. We might as well take this one step further and say that all acoustical behavior is the same as it all is a result of the action/interaction of energy within a specified portion of the spectrum. And since its all the same...hey, why bother to learn that acoustics stuff. Its all the 'same'. Should we extend this model to include the complete EM spectrum as well?

I am just not sure how reducing the affect of standing waves and specular waves to being"the same thing" is any more beneficial than simply stating that 'sound is sound'; as neither provide much practical insight into how either behavioral ultimately manifest themselves, nor in how one ultimately evaluates and addresses issues pertaining to such behavior. }



But that all can quickly change if the opening is into another 'significant' coupled space which can effectively change the length of the 'tuned pipe' and results not only in a modified resonant frequency, but also adds all sorts of new and 'fascinating' resonances of varying levels and distributions.

Coupled spaces can be a great deal of "fun". And on a practical note, you don't try to calculate this behavior unless you have the luxury of something like EASE in the design stages - and that is still more in an attempt to simply avoid catastrophic trending indications than to actually determine real behavior! And if you are dealing with an already existing space, you don't even waste time messing with 'guesstimate' calculators...you measure and deal with the complex reality in all its glorious complexity.

There, I managed to describe the practical issues surrounding the 'wonderful' behavior of complex coupled spaces without using any cuss words! ...Quite an accomplishment! heh

Edit: Hmmm...I'm not sure how some folks are interpreting the later comment regarding coupled spaces. If you are not sure, look up "litote".
Old 18th December 2010
  #71
Gear Addict
 

Quote:
Originally Posted by SAC View Post

Coupled spaces can be a great deal of "fun". And on a practical note, you don't try to calculate this behavior unless you have the luxury of something like EASE in the design stages - and that is still more in an attempt to simply avoid catastrophic trending indications than to actually determine real behavior! And if you are dealing with an already existing space, you don't even waste time messing with 'guesstimate' calculators...you measure and deal with the complex reality in all its glorious complexity.

There, I managed to describe the practical issues surrounding the 'wonderful' behavior of complex coupled spaces without using any cuss words! ...Quite an accomplishment! heh
Glorious complexity?? hmmm, I'll take cuss words for $500 Alex.

A 30db null that doesn't budge is totally fun in a coupled space tutt
Old 18th December 2010
  #72
Lives for gear
 
johndykstra's Avatar
 

Quote:
Originally Posted by johndykstra View Post
^^^^^

It sure would appear that a lot of the nulls have been minimized as well.
likewise I expected to hear how bass traps could be responsible for reducing nulls as shown in Ethan's graphs and video, even though there's been information provided that would suggest this isn't possible.

I found your statement interesting Ethan, but I didn't want to fall into an "idiot trap" and have all you big cats pointing and laughing when I took the bait... that or I thought you had taken up smoking salvia. It makes sense though, as brass and woodwinds would be pretty hard to explain if this didn't occur.
Old 19th December 2010
  #73
Gear Guru
 
Glenn Kuras's Avatar
Quote:
Originally Posted by Ethan Winer View Post
I expected heads to explode after I wrote this, especially the last sentence. Yet all I hear is crickets.



Nobody has even one question? heh

--Ethan

[/url][/b]
I found funny, actually witty was

Quote:
Modal and non-modal peaks / nulls are the same thing.
They are??? Oh wait

Quote:
The only difference between a modal frequency and a non-modal frequency is that the acoustic wave for a modal frequency happens to fit exactly within a pair of opposing boundaries.
Ethan you rock!!heh
Old 19th December 2010
  #74
Gear Guru
 
Ethan Winer's Avatar
 

Lightbulb

Quote:
Originally Posted by johndykstra View Post
I didn't want to fall into an "idiot trap" and have all you big cats pointing and laughing when I took the bait.
Never happen, my friend. I do point fingers and laugh sometimes, but never when someone asks a legitimate question.

Quote:
Originally Posted by Glenn Kuras View Post
They are??? Oh wait
Ethan you rock!!heh
Yep, they sure are the same.

Well okay then, I guess my job is done here.

--Ethan

________________
The Acoustic Treatment Experts
Old 20th December 2010
  #75
Gear Nut
 

Quote:
Originally Posted by SAC View Post
Covering some ground twice, and a few areas for the first time, here are a few aspects of the overall behavior. Each has in some form or fashion been mentioned or at lest inadvertently touched upon...

This is simply an attempt to clarify a few points distinguishing some of the wave behavior confusion discussed earlier.

As mentioned above, do not become obsessed with trying to distinguish modal from specular reflections. Oh, they are distinctly different in nature and behavior. But if the appropriate tools are employed to evaluate them, the distinction becomes rather academic and simple. And you can obsess over the logical descriptions and distinctions at you leisure.


As has been stated on multiple occasions, a modal standing wave behaves differently then a focused specular reflection.

The primary distinguishing factor is their size.

Specular reflections are reflected based upon their wavelength being equal to or smaller than the dimensions of an incident surface.

Modal standing waves have wavelengths larger than the dimensions of a boundary surface.
This’ break point’ is referred to as the Schroeder ‘critical frequency’. But please do not go trying to calculate or determine this! It is neither necessary no clear cut, as you will generally have a region of overlapping modal and specular behavior as you will have various boundary dimensions resulting in some energy being reflected specularly while some energy is reflected in a modal manner. Again – don’t worry, as the measurement methods will sort all of this out very elegantly as mentioned above.

With reference to the earlier reference to Ethan’s example, the only way to remove the standing waves in a small acoustic space is to effectively remove the boundary - thus effectively requiring a 100% efficient absorber covering all of the LF frequencies! Now, as fascinating as this imagined concept is, does anyone here posit that absorbing 100% of the LF is a realistic viable option - especially considering that most would kill for a reliable LF absorber capable of decreasing the gain at an anti-node (peak) by just 6 dB SPL??

Thus, in a real world room, you will NOT remove ALL of the returned energy!

The following erroneous notion has been mentioned several times, and it seems a source of at least part the confusion.





In that statement we are trying to differentiate a standing wave from modal behavior and associate a standing wave with specular reflections. This has gotten the notions confused. If you want to associate labels, associate the standing wave with modal behavior.


Thus we are still confusing specular behavior with modal behavior. And modal behavior is not the result of two specular reflections superposing (summing). Modal behavior is the result of the same incident energy that is reflected and returned that reinforces (itself) and both the magnitude and the null of the resultant wave.

Not to get too far afield on a tangent with this element, but one that will be mentioned as it is fundamental to the behavior describes - One will also note that boundaries featuring greater acoustical impedance than the transmission medium return the reflected incident wave energy in phase, thus resulting in the constructive reinforcement of the resultant (incident + reflected) waveform. And I know of no practical effective room boundary featuring a lesser complex acoustical impedance capable of delaying or inverting the phase of the reflected energy propagation sufficiently to disrupt said reinforcement. Perhaps when we begin manufacturing walls of closed cell aerogel or of sparse matrix carbon nanotube Buckyballs…. So in a practical bounded enclosure, we are ‘stuck’ with this condition…

Jumping over to specular behavior for a moment to present a contrast, the energy content of mid and high frequency wavelengths is a fraction of that of low frequency wavelengths. There is good reason why low frequency waves, such as one might experience with thunder or ELF - extremely low frequency communications which can be perceived far away from the actual source while relatively short low energy content specular energy has already been dissipated by friction.

So, with specular reflections, we are able to control the resultant polar lobing that appears as comb filtering in a frequency response because we are more able to intercept and absorb the more limited (focused) spatial distribution of the relatively low energy content of the constituent mid-high frequency shorter wavelength energy of which specular reflections are comprised.


With low frequencies the difficulty of doing this is increased substantially.


And I know folks don’t want to hear about this, but our conceptualization of this phenomena is also confused by our wave models. We so often imagine wave behavior in the form of a transverse wave (you know, what appears to be a sine or cosine wave). But a bounded modal space the air column more effectively exhibits a behavior more consistent with a longitudinal wave featuring periodic regions of low and high pressure.


Oh, and other aspect that many forget. Sound has size. Oh, we talk of wavelengths all the time, but how often do folks actually think of the relative size of the wavelengths in the real world? Specular energy is specular precisely because the wavelengths are smaller than the incident boundaries. They are focused, meaning they do not occupy the entire room, instead being limited to smaller ray like distributions. And as a result, not only are the effects of superposition position related, but due to their relatively small size, they are able to be effective controlled in the room by the application of absorption, diffusion, or redirection/controlled reflection.


The modal standing waves, on the other hand, are not small localized waves. They are large relative to the room boundaries and to the room itself. They fill the room. Thus you have areas of relative high and low pressure distribution, but you do not in a small space have regions of modal behavior and regions lacking it. It dominates the room.

And as far as absorbing a null… how do you absorb that which is effectively 'absent', meaning the particle velocity is effectively 'zero' and the pressure is low - rarified. Technically, neither porous velocity traps nor tuned resonant absorbers are effective here! It is akin to trying to absorb a shadow. The null is a function dependent upon the relation of the dimensions of the room and the wavelength of the energy that is returned. And at the risk of creating more misconceptions rather than increased understanding (hey, can it get worse?? ;-), modal behavior is a LF behavior akin to mid-high frequency flutter echo that occurs between parallel surfaces, only modes have big honkin' wavelengths (and where small surface angle irregularities are not critical due precisely to the relative size of the wavelengths involved).

Oh, and let's jump over for a moment and consider the compounding influence of coupled spaces in the suggested room posited in the thread...At its simplest, the room functions, as you will, as a large tuned pipe With closed ends terminated in anti-nodes and open ends terminated in nodes. It is this perspective by which the myriad traditional modal calculators function. And due to ideal geometric and simplistic bounded acoustical impedance assumptions, they are estimates at best.

In the worst case, you have a series of irregularly bounded spaces that are coupled in a complex manner. And this complex topology effectively functions as multiple enclosures of the various component sizes as well as in all of the various combinations and permutations of the various component and summed space. Such calculations are complex (to say the least) in the simplest of configurations where the boundaries are assumed to be 'regular' and of constant impedance. But the difficulty scales exponentially with the complexity of the assemblage of bounded spaces - compounded still further by the real world acoustical impedances of the boundary surfaces which function as frequency and time dependent absorbers and reflectors varying with the angle of incidence. And not only are the reinforced frequencies more difficult to calculate, but the energy density distribution becomes much more complex.

Suffice it to say that reliable prediction of such complex behavior is beyond any practical desktop solution currently available. We can obtain a rough idea that tends to eliminate inadvertently creating really bad spaces, but such modeling is not capable of producing a model sufficient to replace real world proof of performance measurement and verification. (In other words, we can make them look really good for the purpose of making sales and marketing proposals! ;-)


Both resonant tuned absorbers and velocity based porous absorbers function best in regions high pressure and high velocity, respectively. And neither condition exists at the null. Thus, unlike baseball where you “hit ‘em where they ain’t”, in the treatment of modes, you “hit ‘em where they are” based upon how absorbers work.


We have a bounded space. And we have modal behavior. Our tools are relatively ineffective in the nulls. But we can mitigate the ‘peakiness’ by absorbing the peaks and minimizing the reinforced resonant energy. And if we are designing a space, we can chose dimensions that most effectively distribute the modes - spreading them out as much as possible. But as far as the nulls in an existing space, the simplest and most effective solution is to literally avoid them. Move. Assuming no other adjustments, adjust your seating arrangement forward or backward (assuming a left right symmetry) such that you are in between the null and the peak.

Then treat the room modal behavior as suggested above.

And then you can begin obsessing over the specular energy and obtain practice making and interpreting and discovering just how treatment moderates specular reflections. And that is quickly summarized above.

Thanks, SAC
John
Old 24th December 2010
  #76
Gear Addict
 

Hmm, I came upon an interesting realization/coicindence of treating nulls- against what is said here. i.e. the adjacent peaks are not showing much change.
I have measurements that IMO possibly conclude that you can treat a null..or 2. Big caveat is that I have a coupled space that could be the main cause , but an interesting find. I need to move a 8' panel out of the live room into the CR-but damn its on the ceiling/ wall corner. That , or I can make a new panel -no biggy ,as I have plans for a few more panels to make.

Will post measurements when I can soon.
Old 10th December 2018
  #77
Here for the gear
 

I am constructing a new room to be used as a mastering studio. I started with an ideal bolt area room. (In this case 9'x12'). From there I chamfered the corners of the room. So now with these chamfered corners, I am trying to figure out how that affects my bolt area. To be more specific, i'm trying to figure out how far apart to place the side walls. (it was 9' when it was only a rectangle). I can't exactly measure the room and then move the wall and measure again...I know that calculating room modes for an irregular room shape is complex, but I am prepared for the complexity...Can someone point me in the right direction of where to learn about this complex area of acoustics?
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
fitzj / Studio building / acoustics
16
Puncher / Studio building / acoustics
18
monitorgain / Studio building / acoustics
18
UkuDouge / Studio building / acoustics
6

Forum Jump
Forum Jump