Room modes, calculated versus measured

Posted: Tue Mar 15, 2005 6:22 pm

eric_desart wrote:

Rod,

I do like and respect you and confirm this publically.
And you stated before that this is your home, and also I do respect that.
And I'm convinced of your very good intents, always.

And when I use words they come from my hart and my being, ALWAYS.

Well, in your home live many children.

Warm regards
Eric

Eric,

I do like and respect you as well............ I too have stated this publicly and will continue to do so. Nothing and no-one will ever change this.

You are correct - in this house lives very many children........

This is the same as in my real home - where I raised 5 children - and here (in the forum home) as in my real home - sometimes I have to step in between those children and try to make a peace.

Sometimes it works - sometimes it doesn't - and sometimes everyone gets angry with me - but still must I try.

You are correct - my intentions are good -

Sometimes I cringe at the things I hear - but let them go because maybe the under=riding theme means more than any one comment -

But every once in a while - especially when I see an olive branch - I feel the need to stop things from escalating and give everyone a chance to take a deep breath.

I hope to call you friend for a long time to come Eric, but I can never give up who I am in the process.

I hope you can see my point in this.

Sincerely,

Rod

Posted: Wed Mar 16, 2005 7:12 am

edit: i'll just retract my (i thought it was funny) lighthearted commentary and offer this

if this feud is one of your larger concerns in life, truly you are a fortunate soul, cuz this ain't that serious in the grand scheme

Posted: Wed Mar 16, 2005 3:52 pm

Paul,

> You don't trust me becuase I belong to the posse that's always going around causing you trouble eh? <

I see no point continuing in this vein, so I won't. But a mutual friend of ours suggested I ask you to discuss some of these issues further with me by email. In the past you have refused, but in the interest of "olive branches" and all, hopefully you'll now reconsider? You know where to find my email address, yes?

Thanks.

--Ethan

Posted: Wed Mar 16, 2005 5:56 pm

Ethan Winer wrote:

Paul,

> You don't trust me becuase I belong to the posse that's always going around causing you trouble eh? <

I see no point continuing in this vein, so I won't. But a mutual friend of ours suggested I ask you to discuss some of these issues further with me by email. In the past you have refused, but in the interest of "olive branches" and all, hopefully you'll now reconsider? You know where to find my email address, yes?

Thanks.

--Ethan

Ok

I would have preferred an 'open' talk, as the issue of 'open' and 'openness' is one of the root issues IMO

However I do appreciate that maybe an audience isn't helping right now, so here ya go...

paulwoodlock@creepers.fsnet.co.uk

but... ONLY... if you're prepared to ditch the preconception that anything I say that's personal is automatically an attack and talk on a level that isn't emotionally charged by this war.

Paul

Posted: Thu Mar 17, 2005 2:06 am

Ethan,

You start about modes based on a nonsense argument to prove how wrong my related posts are.

You got 2 examples PLUS your own room.
Jeff's AND Bob's AND Ethan's.
Plus you deleted a message from your own EQ Forum, where a poster told that you were wrong by telling the original questioner, that the dominantant resonance he reported couldn't be a mode, since you calculator told so. You even explicit stated that you calculated it and that it must been something else since the closests modes were far away.
When this poster stated that this was related to 2 very close coinciding tangential modes, you choose, rather than investigating for the benifit of the original questioner, just to delete the reply of that poster. It's even possible the questioner never got the chance to see this reply.
The deletion of this message was indeed a very scientific approach. Just let contradicting info dissapear.
So that's 4 examples already.

Are you going to retreat your Room mode Calculator now?
Or minimum have the responsibility towards others to inform them (FAQ and Site visitors + text and Help File of room calculator) about this limitation as I really gently asked/suggested maybe 1.5 to 2 years ago already, but complete ignored by you (as well as the several other times this is discussed).
You could also simply refer to other calculators (there are many good ones on the net) but I know that sins against who you are.

Or does this discussion restarts once more and/or somewhere else as if this one never existed?

I still stand behind any related letter I wrote.

It's handy to evade the core of this thread now, after trying to ridicule my related posts you so generously linked to.

WE

And this is the complete quote from the SUBSTANCIATED article you referred to:
http://forum.studiotips.com/viewtopic.php?p=5570 (original formatting quote preserved)

A strong advice:

NEVER use a Room Modes Calculator which DOES NOT take the Tangential and Oblique modes into account

One just doesn't know!

This is a comment CLEARLY written in my Room Mode Calculator itself:

Asymmetric absorption in function of spatial distribution and frequency can/will cause Room Modes to shift.
As such one should be cautious in interpreting the results from a room modes calculator, which can be a very helpful tool, but is no substitute for acoustic knowledge.

My Room mode calculator is NOT way off. It's the known math.
But NO software tool is substitute for knowledge AND made within the assumption of stylized specific boundaries.
It's an ADDITIONAL tool, not more and NOT less. But VERY useful.
So NOT my Room modes calculator is WAY OFF, but the one making such a statement resorting to Headliner (pseudo) Acoustics.
And I don't mind that People use other calculators. In the by you referred article of mine I link to several room mode calculators from different people.

And this are the hardly described measurements of Ethan combined in one graph.
Such measurements should be cause to investigate and measure further to analyze. Hardly to prove whatever.
Also this assumes one knows what to look for.
One must understand the phase relationship of the speakers with the phase pattern of the modes, the coupling (strength) of the source with the modes in function of position, and the effect of the exact microphone position.
And one must know how the modal patterns of the different order modes look like inclusive the NON-axial patterns.

And this is NO breach of Copyright but Fair Use of data, only meant for others to make the offered data in this thread readable.
Which shows that none of them really relate and two which are meant to relate (assuming exact similar conditions) only poorly do.

THIS PICTURE IS EXCLUSIVELY MEANT AND MADE FOR USE IN THIS VERY THREAD WHERE IT MUST STAY.
And more for others to put the current created confusion in context.

And being not able to explain something IS NEVER BEEN PROOF of whatever, only cause to further investigation, from which this data is HARDLY an example.
This data originates from an Author who writes in numerous articles that room acoustics isn't that difficult and dips and peaks are (easy?) predictable.

People can decide how important or not the additional data of the NON- axials is in this picture.
But however poor, using this measurement data as proof that your calculator covers it, tells more about the one making the statement.

Posted: Thu Mar 17, 2005 12:55 pm

Eric,

> You start about modes based on a nonsense argument to prove how wrong my related posts are. <

I may claim you're wrong on some acoustics issues, but I usually avoid words like "nonsense argument." A little reciprocal respect will only help you make your case. Insults never do anyone any good, and especially not you.

> Plus you deleted a message from your own EQ Forum <

A lot has been made of that deletion, but you have it all wrong. I deleted Scott's post because it was insulting and inflammatory. Not because it contradicted me! I have no problem with people saying I'm wrong if it's done politely. I gladly acknowledge that sometimes I am wrong! If only everyone else had that same attitude.

> Are you going to retreat your Room mode Calculator now? <

There's nothing wrong with my mode calculator. It gives a useful indication about the most important modes. As I explained to someone else this morning in an email:

Quote:

Many people are stymied trying to interpret my program, so what are they to make of the output from Eric's? More to the point, my program shows a list of the acceptable ratios, whether they count on non-axials as Eric said or not. I even explain in the instructions that people should see how their proposed room compares with the listed ratios. So in that sense my program omits nothing. Heck, if I removed the display at the bottom and just calculated the current ratio and displayed the "good" ratios, it would be just as useful, no?

> It's handy to evade the core of this thread <

Again, you'll get farther if you drop the insults and accusations, and stick to the science.

> My Room mode calculator is NOT way off. It's the known math. <

Then please explain why most of your calculator's predicted axial modes are so far off from what I measured. Please show me the math that accounts for this 20 percent disparity. I've looked through the formulas in your Excel worksheet for some explanation of this and found nothing.

> It's an ADDITIONAL tool <

Yep, just like mine. Eric, you have to remember that most of the people who read my FAQ and download my ModeCalc program are beginners. They have no idea how to interpret the data your calculator outputs. If you want to think of my program as "dumbed down" a little for the masses, I have no problem with that. But to continue insulting me and my effort to help people over such a silly thing is, well, silly.

If it will make you happy I'll gladly add a note to my ModeCalc explanation page that other modes exist which are less significant than axials, but they do exist. I would even add a link to your program as an example of a more complete program. But I certainly will not link to your site or to StudioTips as long as both are filled with insults toward me.

> Such measurements should be cause to investigate and measure further to analyze. <

Great, tell me what to try next and I'll gladly do it and report back here. I already told you I have another series of ETF plots made at different locations in the same room. Shall I post those?

> not able to explain something IS NEVER BEEN PROOF of whatever <

Agreed, and perhaps I overstated that position. However, as a born skeptic I need proof before I accept anything that is not intuitively obvious. And the importance of non-axials is certainly not obvious, at least not to me.

> But however poor, using this measurement data as proof that your calculator covers it, tells more about the one making the statement. <

Again, stop with the personal insults and maybe we'll get somewhere, okay? Elsewhere, Paul suggested we could all use more Embarassed

and less

and I think that applies to you as much as to me.

--Ethan

Posted: Thu Mar 17, 2005 1:23 pm

eric, you make a mean picture, really, they are nice lookin.

i meant to enter an analogy about room measurements way back at the beginning of this thread, before it imploded.

about the dips and peaks not lining up with the dots on a spreadsheet...

you observe the same thing in mechanical systems under impact or steady state simulation. take an accelerometer, and a panel of drywall or plywood or whatever you like, move the accelerometer around on the panel, taking measurements over and over while striking in the same place...

you will observe wildly varying effects, and you can really only observe any given mode or antiresonance by finding a spot where it is maximally expressed. And even then, you will get confusing FFT if some other activity is present in neighboring frequencies.

all those peaks and nulls and modes and antimodes ... add up to make a reasonably chaotic mess.

fortunately, when testing mechanical panels, the density of the modes is alot less than in a room, they tend to be nicely spaced.

The astm standard for measuring damping uses a beam of specified dimensions, well, more suggested ratio's of dimensions, that is clamped rigidly on one end, sort of forcing predictable behaviors.

to make things worse, reflections can cause trouble as well. put a "shaker" on a panel, and you will find edge-reflection effects marring the situation even more.

i had intended to take some measurements of a panel, showing what seems to be modal behavior changing in frequency on the same panel, but i did not get around to it quickly enough, and frankly i wonder if input into these "discussions" is really, what's the word? sought?

so i guess, intuitively, i wouldn't worry too much about any specific dip or peak and how it lined up on a graph. with so many phenomenon interacting - reflections, some mode peaking, another mode dipping, it seems challenging to conclude one way or another.

i disclaim by observing that i have measured only 3 normal rooms in my time. actually, the same room 3 different times. always inspired by these firefights. just to see...

Brian

Posted: Thu Mar 17, 2005 1:35 pm

eric_desart:

Re your graph.
I'd been focusing so much on the peaks, I'd forgotten about the valleys. There's a couple of strong dips that correspond to tangental and oblique there.

Posted: Thu Mar 17, 2005 2:04 pm

Bob/Z,

> There's a couple of strong dips that correspond to tangental and oblique there. <

Or maybe it's just simple boundary interference. The exact microphone placement is documented in my report, but it's still not straight forward to determine the peak and null frequencies because of the angles involved. This is where Eric's expertise would come in handy.

The thing about non-axials is there are so any of them that you can find almost any peak or null and proclaim it to be due to a non-axial mode, even if it's not. This is why I offered to post ETF graphs of the same room taken at different microphone positions. In fact, that's exactly why I tested at different places - to try to identify and distinguish true modes from simple interference. That's when I discovered that few/none of the measured peaks and nulls aligned with the predicted modes. Hence my first post in this thread.

--Ethan

Posted: Thu Mar 17, 2005 2:08 pm

Ethan Winer wrote:

Bob/Z,

> There's a couple of strong dips that correspond to tangental and oblique there. <

Or maybe it's just simple boundary interference.

...

The thing about non-axials is there are so any of them that you can find almost any peak or null and proclaim it to be due to a non-axial mode, even if it's not.

...

--Ethan

yes, exactly. situations involving densely spaced modes are never that simple (in my experience from mechanical systems, of which i have measured ... thousands). you'd have interference, and all these different modes (modes aren't exactly 112.4hz, they have a bell sort of shape) interacting and adding in different ways. a peak at 88 might be two modes at 81 and 94 adding to make one big tower, etc.

Posted: Thu Mar 17, 2005 2:30 pm

Brian,

> a peak at 88 might be two modes at 81 and 94 adding to make one big tower, etc. <

Yes, good point. ETF does a good job showing slight peaks farther down the slope of a larger peak. But you're correct that at some point two different peaks can be close enough together to appear as one. In the set of graphs I've offered to post, if anyone cares, you can see this effect much more clearly. In a few cases peaks that are obvious at one location are completely hidden in another location.

--Ethan

Posted: Thu Mar 17, 2005 2:46 pm

Ethan:

Quote:

Or maybe it's just simple boundary interference. The exact microphone placement is documented in my report,

I'd thought of that too.
Although I'd guess that the speaker placement, and that you've got two speakers, would be more relevant. Based on sticking a ruler against my computer monitor against a photo of your room, I'd guess your speakers were about 5.8' apart, but I wasn't able to devine enough information from their placement to guess about wavelength to effect.

Quote:

The thing about non-axials is there are so any of them that you can find almost any peak or null and proclaim it to be due to a non-axial mode, even if it's not.

I thought the rule was that if it sustains for a bit after the speakers stop then it's probably a mode. Or to put it another way, if you shorten the gate time in ETF5 and it goes away then it's a mode. If you shorten the gate time in ETF5 to less than the time it takes to make a ceiling floor room bounce and it's still there then it's SBIR or inter-speaker effects or something.
Shortening the gate time to less than the wavelength you're interested in probably won't work well though.
40hz = 28'
room height around 8'
so that's 28/8 = 3.5 floor ceiling trips

Quote:

In fact, that's exactly why I tested at different places - to try to identify and distinguish true modes from simple interference

I think the only good places to put a microphone for a modal check are each of:
a) back top right
b) front top right
c) front bottom right
i.e. these rules:
1) a tri-corner.
2) if the room is symmetrical to left-right, then there's no point in doing the opposite side
3) back and then front because there's a different distance from the speaker pair
4) top and then bottom for the same reason -- I chose the front because there's a larger distance difference from the speakers in your config than there would have been at the back.

Random microphone positions in the room will give random evidence of modes. (This differs from the absorbtion estimating technique where lots of microphone positions are a good thing)

Posted: Thu Mar 17, 2005 3:01 pm

Brian R:

This is one of those posts I 'm really not sure about.

Quote:

a peak at 88 might be two modes at 81 and 94 adding to make one big tower

That's certainly true of absorbtion -- when two membrane (Q) absorbtion curves with close center frequency are added you get one big tower centered roughly halfway between the two frequencies.

In one dimension when two frequencies are added you get beats and chords, not some new frequency that's halfway between them.
In three dimensions I'm extremely shaky about what happens when two frequencies are added.

Posted: Thu Mar 17, 2005 3:09 pm

well, in two dimensions ( a mechaniacl panel), this happens as a matter of routine, especially as damping gets high and modes get very broad, and very overlapped, and it takes more ... careful analysis to really assess things.

in 3D mechanical structures, same thing. i can't see any reason to believe that acoustic modes are different.

find a thread titled (by ethan) "do room modes even matter?", in the audio theory, setup & chat section at AVS, and look for an entry by Terry Montlick with a picture...

he describes this.

i will make what effort i can to sieze 30 spare minutes with the lab o-scope to make some examples and post 'em.

Posted: Thu Mar 17, 2005 3:11 pm

Ethan Winer wrote:

I deleted Scott's post because it was insulting and inflammatory. Not because it contradicted me!

Nice try........
as unambiguous as the dishonesty shown in the earlier discussed quote here:
http://www.recording.org/ftopicp-208255.html#208255
That's why you """Forgot""" to correct your own misinformation too, to the questioner?

It least you know and remember that the example was there. So you can't tell that you never found tangentials.

It's still 4 examples.

This topic sponsored by: