modeWizard: Yet Another Mode Calculator but with a twist

Post and discuss acoustic topics, Studio design, construction, and soundproofing here

modeWizard: Yet Another Mode Calculator but with a twist

Postby jcgriggs23 » Fri Apr 21, 2006 1:35 am

Greetings StudioTipsters,

Attached is my latest entry into the mode calculator parade.  I think this one is unique in being the only GUI-based mode calculator that allows calculation and comparison of multiple rooms (although it does single rooms too.)  For those of you familiar with my previous entry into the mode calculator sweepstakes - RMmapper - this one - modeWizard - is the successor to that program with more features, better math (thanks mainly to Eric Desart, Jeff Szymanski and others from the old Yahoo! acoustics site), a much easier to use interface that is much more interactive and much better output (in HTML format.)

The attached zip file contains the program itself (a Java .jar file - you must have Java run time version 1.4.2 or newer installed to run it, but it should work on Windoze, Linux or Mac) and some minimal documentation for the program and the math it implements.  I still have a number of features I would like to add to the program before I would consider it "finished", but it is functional and, hopefully, useful in its current form.  Eventually I hope to have the time to complete development, but that doesn't seem likely in the near future, so I thought I'd post it as a work-in-progress.

Anyway, please try it out and let me know what you think.

If someone with appropriate privileges wants to add this to the "Calculation Tools" section of the site (or tell me how to do this myself), I'd be much obliged.  Also, if anyone is interested in helping to develop better documentation (and some on-line help) for the program, please let me know...

Regards,
  John Griggs
Attachments
modeWiz.zip
(329.8 KiB) Downloaded 2361 times
jcgriggs23
 
Posts: 257
Joined: Tue Feb 10, 2004 2:53 am
Location: Ottawa, Canada

Postby Bob » Fri Apr 21, 2006 3:33 am

jcgriggs23:

On my main computer I have .jar files tied to WinZip, so double clicking didn't work for me.

So I
a) unzipped your modeWiz.zip into a c:\temp\StudioTips\jcgriggs
b) used this statement to run it against jdk1.4.2_06
 java -jar c:\temp\StudioTips\jcgriggs\modeWiz.jar

I used your default dimensions and let it run (next, next, next, ).

Nifty. :)
Bob
 
Posts: 4358
Joined: Sun Feb 08, 2004 4:37 am
Location: Oakville, Ontario, Canada

Postby Scott R. Foster » Fri Apr 21, 2006 3:40 am

Way cool.
SRF
Scott R. Foster
 
Posts: 3854
Joined: Mon Feb 02, 2004 12:41 pm
Location: Jacksonville, FL USA

Postby Eric Desart » Fri Apr 21, 2006 9:52 am

Bob wrote:jcgriggs23:

On my main computer I have .jar files tied to WinZip, so double clicking didn't work for me.

So I
a) unzipped your modeWiz.zip into a c:\temp\StudioTips\jcgriggs
b) used this statement to run it against jdk1.4.2_06
 java -jar c:\temp\StudioTips\jcgriggs\modeWiz.jar

I used your default dimensions and let it run (next, next, next, ).

Nifty. :)


Hi Bob,

If I where you I just should go to WinZip and untie jar files with whatever compression program.
You could always reverse that, or if necessary you can decompress them by simply opening WinZip or whatever decompression program and opening them from there.
This linking of Jar Files to compression programs, as a standard user causes always such annoying workarrounds while the tie itself by most never or hardly ever is been used.

Eric
Eric Desart
Moderator
 
Posts: 1174
Joined: Tue Dec 13, 2005 10:04 am
Location: Antwerp/Belgium

Postby Bob » Fri Apr 21, 2006 9:59 am

Eric Desart:

I know how to do that.

In my case I find WinZip handy to see what's in a jar file without exploding it -- so I did it on purpose.
Also, the installation of my JDK is a bit non-standard too.

Or to put it another way, it's vastly better for me in an hours-per-day kind of way, to double-click-on-jar files to run WinZip.
Bob
 
Posts: 4358
Joined: Sun Feb 08, 2004 4:37 am
Location: Oakville, Ontario, Canada

Postby Eric Desart » Fri Apr 21, 2006 10:01 am

Bob,

Understood.  You're a programmer yourself.
I only think for standard users, having the same problem (......damned why doesn't this work) they better untie this extension link.

Eric
Eric Desart
Moderator
 
Posts: 1174
Joined: Tue Dec 13, 2005 10:04 am
Location: Antwerp/Belgium

Postby jcgriggs23 » Fri Apr 21, 2006 12:31 pm

Bob wrote:
[quote]jcgriggs23:

On my main computer I have .jar files tied to WinZip, so double clicking didn't work for me.

So I
a) unzipped your modeWiz.zip into a c:\temp\StudioTips\jcgriggs
b) used this statement to run it against jdk1.4.2_06
java -jar c:\temp\StudioTips\jcgriggs\modeWiz.jar [/quote]

Bob,

D'oh! - I meant to add this to the documentation, but forgot.  Thanks for catching it!!  I usally run by double clicking on the JAR in Windoze and by using a Makefile ("make run") under Linux (where all the development happens.)

Regards,
  John
jcgriggs23
 
Posts: 257
Joined: Tue Feb 10, 2004 2:53 am
Location: Ottawa, Canada

Postby Scott R. Foster » Fri Apr 21, 2006 2:35 pm

John:

Love the program and all the info it auto-generates, but a quibble if you don't mind.  When you go to enter a new room by absolute dimensions the program comes up with data in the fields - no biggy on the large unit field, but if the small unit field were zero it would save an entry step every time the large unit was a whole number or a known decimal if the small unit was empty and did not need to be wiped.

==============================================================================================

On to a basic theory question:

Is there a potential for non-axial modal ringing between corners - mode calculators like the one in question - and also the one made by Bob - show the long axial as the lowest mode.  

Example: a 12 x 14 x 8' room [that'd be 4.27 x 3.66 x 2.44m for you heathens] will have a lowest mode of +/- 40 Hz which is based on the 14' axial.  But, what about the dimension from corner to corner [18.4' feet = +/- 31 Hz] , can this dimension - even though is represents a limited surface area - support modal ringing enough to be audibly unsmooth?  How about the dimension from a spot on the floor in one corner to the spot at the ceiling in the corner diagonally opposite [20.1' = +/- 28 Hz] ?
SRF
Scott R. Foster
 
Posts: 3854
Joined: Mon Feb 02, 2004 12:41 pm
Location: Jacksonville, FL USA

Postby bert stoltenborg » Fri Apr 21, 2006 2:57 pm

These waves have to be seen as plane waves, Scott.
They won't bounce from a corner to a corner, I guess.
They need a solid surface to bounce from.

Bert
.
If you view life with the knowledge that there are no problems, only opportunities, you are a marketing manager.......this is my personal philosophy
bert stoltenborg
 
Posts: 4566
Joined: Sun Apr 18, 2004 11:03 am
Location: Achterhood, Netherlands

Postby Scott R. Foster » Fri Apr 21, 2006 3:01 pm

solid = flat

correct?
SRF
Scott R. Foster
 
Posts: 3854
Joined: Mon Feb 02, 2004 12:41 pm
Location: Jacksonville, FL USA

Postby Scott R. Foster » Fri Apr 21, 2006 3:02 pm

...and this is relative to size - the lower the wave, the biger the flat surface required to support modal ringing / plane waves - correct?

Thanks
SRF
Scott R. Foster
 
Posts: 3854
Joined: Mon Feb 02, 2004 12:41 pm
Location: Jacksonville, FL USA

Postby jcgriggs23 » Fri Apr 21, 2006 3:14 pm

Scott,

First of all, thanks for the kind words about the program.

Secondly, the dimension fields convert their values between metric and imperial if you switch measurement systems - this makes it hard to know when the program should ignore or zero-out the details.  It's hard to distinguish between "noise" and fine detail in values due to conversions.  I could add a control to clear the fields - would that help?

Eventually I intend to add a method for the user to alter the program defaults and constants (Speed of Sound, etc.) - then you could just create a set of defaults that zero these fields.

Again - thanks for the feedback.  It's very gratifying to have other people using this thing!!

Regards,
   John
jcgriggs23
 
Posts: 257
Joined: Tue Feb 10, 2004 2:53 am
Location: Ottawa, Canada

Postby bert stoltenborg » Fri Apr 21, 2006 3:53 pm

Scott R. Foster wrote:...and this is relative to size - the lower the wave, the biger the flat surface required to support modal ringing / plane waves - correct?

Thanks


yep  :D
If you view life with the knowledge that there are no problems, only opportunities, you are a marketing manager.......this is my personal philosophy
bert stoltenborg
 
Posts: 4566
Joined: Sun Apr 18, 2004 11:03 am
Location: Achterhood, Netherlands

Postby krasmuzik » Fri Apr 21, 2006 5:00 pm

Have you researched Walker's BBC ratios for rooms - I think that accomplishes the same thing as "GriggsCriteria" to avoid integer ratios.

http://www.tmlaboratories.com/RoomDimen ... lator.html


This is much much faster than your old program - of course with so many filters and step sizes you can spend hours dithering between room choices down to the mm for this or that criteria- which is likely not even achievable in reality - and who is to say one criteria is better than the other?
krasmuzik
 
Posts: 165
Joined: Tue Feb 10, 2004 1:49 am
Location: Newport, VA

Postby jcgriggs23 » Sat Apr 22, 2006 12:13 am

[quote]krasmusik wrote:
Have you researched Walker's BBC ratios for rooms - I think that accomplishes the same thing as "GriggsCriteria" to avoid integer ratios. [/quote]

krasmusik,

Yes I'm familiar with Walker's work.  If you look at the research behind this recommendation you'll see that it is actually based on a plot of mean square mode spacing to 120Hz for some sets of rooms of various volumes and that it maps an "area" similar to Bolt's famous plot while explicitly excluding bands where two dimensions coincide (defined by the formula at the site you linked.)  It is also the basis for the EBU Techncal Document 3276 recommendations, which includes this same formula for restricting dimension ratios along with some floor area restriction for rooms for various purposes.  The original BBC Research document for Walker's work is BBC-RD-1993/8.   It used to be available here: http://www.bbc.co.uk/rd/pubs/reports/index.shtml, but I couldn't find it in a quick search - I have a hard copy from the NRC Technical Library here in Ottawa, Canada.

I have my reservations about this approach - I'm not sure that mean square mode spacing is a very good choice as a single figure of merit for acoustic behaviour of rooms and the choice to limit measurements to 120Hz makes it hard to compare rooms of differing volumes reliably.  But I'm not going to suggest that I know better than Dr. Walker or the EBU, so the program implements both the room dimension formula and the underlying mean square mode spacing calculation as Filters.  You can use modeWizard to compare mean square mode spacing, the EBU spec (which is basically the formula implemented by the calculator you linked) and the "Griggs" criteria - I think you'll find that the results are actually quite different for my approach vs the EBU/Walker formula and that the EBU mapping over the mean square spacing breaks down when you look at rooms with smaller or larger volumes than Walker considered (100 and 200 cubic meters.)

[quote]This is much much faster than your old program - of course with so many filters and step sizes you can spend hours dithering between room choices down to the mm for this or that criteria- which is likely not even achievable in reality - and who is to say one criteria is better than the other?[/quote]

I'm glad that you feel the modeWizard improves over RMmapper.

Yes - both programs are capable of much greater accuracy than makes sense for real world applications, but I feel that this is useful for examining the models and math behind the various filters and criteria.  

Like most arrogant programmers looking into acoustics for the first time, my original idea back 6 or so years ago was to write a program that could select the one "ideal" room from a set all on it's own.  Of course, once you start to learn what you're up against you quickly realize how much art is involved in selecting a real world room that real people will actually be happy to use, how hard it is to define an "ideal" room (especially using a model as simple as the wave model of room modes) when different applications (Live room, Control room, Home Theatre) all have different requirements and how naive the original goal actually is.  So RMmapper and modeWizard evolved with a dual purpose - to allow comparisons of rooms, even if there is no "ideal" room possible in the search space, based on the users idea of which criteria are important and which are not and to allow investigation of the various simple room selection criteria that have been proposed across sets of theoretical rooms.  I also tried to keep the program as flexible as possible - different applications might lean more heavily on one criteria or another or might involve combining various criteria to make a selection.

I highly recommend that users of either of my programs resist the temptation to be seduced by the apparent detail of the results they produce - the model itself is just not robust enough to make such fine distinctions meaningful, let alone translating such fine tolerances into real-world construction.  Of course, if the program had proper documentation it would have to include a section outlining these considerations for evaluating the program's output (and credits for Dr. Walker, et al, although I have included attributions for some of the math in the existing documentation)

The bottom line is that modeWizard leaves it up to the user to decide how to weigh and apply the various criteria it implements.  My own preference is for the Bonello criteria, but that is a topic for another discussion...

Regards,
   John Griggs
jcgriggs23
 
Posts: 257
Joined: Tue Feb 10, 2004 2:53 am
Location: Ottawa, Canada

Postby krasmuzik » Sat Apr 22, 2006 3:56 am

I did not look to closely at your criteria - I just remembered that was one of the walker rules - no close integer ratios.

And of course there is the reality that all modal calculators based on modal spacing metrics - are ignoring modal ampltude/phase throughout frequency and spatial domains - the resulting interaction of nearby modes - and the dependency of all this on wall impedance.   Nobody should be using any of these calculators to tell their carpenter the wall has to move out 2mm - that is for sure.   Rather they should be used to make first cut choices, and identify potential problems and not search out forever the most optimal solutions.

I really get tired of seeing forum posts like - "my house is 3m high and the modal calculator says I should be 3.1m high - I am doomed and cannot build my room - should I dig out my basement - should I move - someone help me please before I jump out the window!"

I like you implemented all the criteria you could think off though - that should make it fun to play with!
krasmuzik
 
Posts: 165
Joined: Tue Feb 10, 2004 1:49 am
Location: Newport, VA

Postby jonessy » Sat Apr 22, 2006 8:46 am

John,

Congratulations!  :D  :D  :D
I can't wait to come home and play with this a little bit.

I'll report back once I've tried it out!

Kind Regards,
Jonathan.
jonessy
 
Posts: 580
Joined: Wed Jan 04, 2006 3:39 pm
Location: Israel

Postby jcgriggs23 » Sat Apr 22, 2006 2:50 pm

Krasmuzik wrote:
[quote]I like you implemented all the criteria you could think off though - that should make it fun to play with![/quote]

Krasmuzik,

I think this sums up my approach to modeWizard pretty well - it tries to be a flexible calculation tool that implements as many criteria as possible but leaves it up to the user to decide how to apply the criteria and interpret the results.  Hopefully this is useful and not just confusing 8^)

BTW, another application of the extra precision that modeWizard offers is to give yourself a warm and fuzzy feeling about the stability of the criteria you are using for room selection within the tolerances that you can build to.  For example, if you settle on a set of room dimensions using "reasonable" increment values and you figure you can build to within +/- 2cm of those dimensions you can use modeWizard to see how much the criteria you use vary within the tolerances by setting up a Map using the space that goes from X - 2cm to X + 2cm with some small increment for each dimension.  This can give you an idea of whether the selection criteria you are using vary greatly within the "tolerance space" you can realistically build to.  It can also give you an idea of which dimension(s) you must be more exacting with.

Cheerz,
   John
jcgriggs23
 
Posts: 257
Joined: Tue Feb 10, 2004 2:53 am
Location: Ottawa, Canada

Postby jcgriggs23 » Sat Apr 22, 2006 2:55 pm

Jonathan,

I look forward to hearing your opinions!   I also plan to return the favour with your iRoom program, but I need to take the time to see if I can get .Net running on my windoze(98) machine first - I'm mostly a Linux guy.  From the screen shots you posted it looks pretty impressive!!

Cheerz,
  John
jcgriggs23
 
Posts: 257
Joined: Tue Feb 10, 2004 2:53 am
Location: Ottawa, Canada

Postby krasmuzik » Sat Apr 22, 2006 5:49 pm

Oh it most certainly will be confusing to acoustic noobs.   They want acoustic calculators to give them the ANSWER - and then when it does not they start posting to the forum saying someone tell me which settings give me the ANSWER.   Then someone like me up and says the optimal ANSWER does not exist - and even if it did - the ANSWER does not fit your house....and the ANSWER is 42 - everyone knows that already! :D
krasmuzik
 
Posts: 165
Joined: Tue Feb 10, 2004 1:49 am
Location: Newport, VA

Next

Return to Discussion

Who is online

Users browsing this forum: No registered users and 4 guests