John
TRhe room modeling stuff in Soundeasy is FEA? Does it do complex calculations, i.e. complex disipative modes, or are the modes assumed to be real (i.e. complex eigenvalues and eigenmodes)? Can abosorptive boundaries be modeled in the time domain?
I tried to get this info on line, but its not obvious.
TRhe room modeling stuff in Soundeasy is FEA? Does it do complex calculations, i.e. complex disipative modes, or are the modes assumed to be real (i.e. complex eigenvalues and eigenmodes)? Can abosorptive boundaries be modeled in the time domain?
I tried to get this info on line, but its not obvious.
Hi,
I hope this thread is the right place to ask.
Markus pointed me to http://www.gedlee.com/downloads/Cum laude.pdf
There it says "Modes are what make a room a room as opposed to a free field"
My question: What about non-modal influences? Why are they not mentioned? A peak or 'node' in the frequency response can also result from simple interference with a single room boundary. Is that also called a 'mode' in Geddes' terms?
What am I missing?
Thanks and greets.
I hope this thread is the right place to ask.
Markus pointed me to http://www.gedlee.com/downloads/Cum laude.pdf
There it says "Modes are what make a room a room as opposed to a free field"
My question: What about non-modal influences? Why are they not mentioned? A peak or 'node' in the frequency response can also result from simple interference with a single room boundary. Is that also called a 'mode' in Geddes' terms?
What am I missing?
Thanks and greets.
It is not valid to consider single surfaces in rooms at LFs. Basically the waves see all the surfaces and this creates modes. Single surfaces could only be considered as a valid approximation above the modal region where geometrical acoustics operates.
This error is, of course, quite common, where people talk about corner loading etc. But I caution that such discussions have no validity below the Schroeder Freq. I am continually making this point and many people object to it, but that is the way I see it. How can you talk about what a wave will do at only one surface when its wavelength is so large that in encompases the whole room? This just isn't reasonable.
This error is, of course, quite common, where people talk about corner loading etc. But I caution that such discussions have no validity below the Schroeder Freq. I am continually making this point and many people object to it, but that is the way I see it. How can you talk about what a wave will do at only one surface when its wavelength is so large that in encompases the whole room? This just isn't reasonable.
gedlee said:
The room simulations posted several days ago were made with my own code in which I have modeled complex modes and disapation, though my model is rather simple. My code is only applicable to rectangular rooms. SE used steady state FEA and can be applied to room with complex floor plans. It used steady state FEA, that is, solved in the frequency domain. It includes some type of absorption model but I don't think it includes general complex eignvalues and mode. You could download chapter 15 from the website which is the only documentation on the SE FEA. But the FEA part of the code is rather limited in resolution and is very slow.
In your model, how do you handle absorption? This is a very difficult thing to do right. In the most general room with damping, the modes are not orthogonal and the simple series solution which you use is not valid as it assumes orthogonality in its derivation. For light damping the modes are nearly orthogonal and the series is OK with an assumed complex kn to handle the damping and you can even use this complex kn in the eigenmode cosine term, but this is only an approximation that is only valid for light damping. In a very highly damped room, like I suggest, a solution does not actually exist in general.
This has been a limitation in ALL of the multiple sub simulations that I have seen, including my own and Welti's.
This has been a limitation in ALL of the multiple sub simulations that I have seen, including my own and Welti's.
I'm pretty sure you're missing something here. There are non-modal peaks and nulls in the frequency response of every small room. Some are easily predictable: Just play a 86 Hz sine and you'll have a null at 1 m distance from a reflecting back wall because of 1/4-wavelength cancellation.
Ethan Winer made some simple but excellent experiments to prove that "the basic principle of acoustic boundary interference is indeed valid":
Video: http://www.realtraps.com/video_wave.htm
Text: http://www.realtraps.com/quarter_wave.htm
Greets
Ydope said:
This is a mode. 1/4 wave cancelation, mode, its all the same thing. It's the sound field in the room due to its boundaries - all of the boundaries. Some modes involve only two boundaries (in theory, although this in practice doesn't really happen.)
I think from the previuos discussion it must be clear that the modal situation in a small room is anything but simple. Looking at only certain boundaries is just TOO simple.
gedlee said:
Agreed. Mine is no different. But often in engineering, as you know, we can push the assumptions well beyond their range of theroetical validity and still get a fair idea of what is happening. For rectangular rooms my code gives results which arein pretty good with SE simulations and that was my objective. Ser takes hours, my code takes seconds.
john k... said:
Of course, but sometimes those extrapolations come back to bite us in the ****. My room is so heavily damped that I really don't think that the low damping assumptions are valid. One reason that I have put off publishing my sims is that I really want to find a way to use more damping in a realistic way as that alone is a situation worth investigation.
In your code do you use a laguage like C+ or VB or something like MatLab or MathCAD.
markus76 said:<ot>
In the video you can clearly see the eigenlife of eyebrows
Why does Ethan never use a nearfield mic when recording himself? I mean, he should know that this can help increase audio quality.
</ot>
He wants us to hear the untreated room sound quality. It was primarily a visual and measurement demonstration.
The vocal intelligibility (not Fidelity or sound quality) was okay: it's easy to understand his speech. Ethan would agree that it was not broadcast quality.
What would have been more meaningful, however, would be a repeat of the demonstration after all treatments are brought back in.
It's also worth noting that if the demonstration were done with a stereo dummy head, rather than a single mic, the results would have been even less dramatic - the video makes this clear by how hard he has to mic-fish a small, pre-marked area just to find the bottom of the null. It would be much less audible, even unnoticed, by two ears, listening to music, in stereo.
Not to say this stuff doesn't matter, but to say that spatial averages are the only thing worth chasing, unless your head is literally in a fixed listening position. Even the guru at the mix desk moves around quite a bit more than than that test mic did.
I would bet that, in the end, reasonable treatment of the studio's modal and reflection issues took care of it, and that specific treatment for those very hyper-specific locations was neither practical, nor pursued.
--Mark
gedlee said:
Here are some significant differences between the two:
1/4 wave cancellation:
a) not a resonance effect but simple interference
b) not related to room dimensions, only related to distance from boundary
c) only one boundary necessary
d) instant decay
e) happens at every frequency
mode:
a) resonance effect
b) related to room dimensions
c) at least two boundaries necessary
d) slow decay
e) happens only with certain frequencies derivable from b)
So how are they the same?
Greets,
Jules
spatial averages
How do you do sptial averages? Sorry for the newbie question.
I have also googled and have not found a simple answer. I understand that if we use multiple microphones at various positions... But I have one microphone and am too lazy to measure it multiple times then sum them up.
Can I simply set the MLS gate to 22ms up to 44ms and hand hold the microphone and move it around my sitting position slowly and get the "spatial averages" result?
Thanks in advance.
Regards,
Bill
How do you do sptial averages? Sorry for the newbie question.
I have also googled and have not found a simple answer. I understand that if we use multiple microphones at various positions... But I have one microphone and am too lazy to measure it multiple times then sum them up.
Can I simply set the MLS gate to 22ms up to 44ms and hand hold the microphone and move it around my sitting position slowly and get the "spatial averages" result?
Thanks in advance.
Regards,
Bill
Ydope said:
a) Resonances are interferences
b) "distance to boundary" is the same as a room dimension
c) no room has only one boundary
d) whatever that means
e) it can't "happen at ever frequency" for a fixed point
I'm not going to argue this point any more. You simply cannot look at a room as having only one boundary. Its just not reasonable.