Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Multi-Way Conventional loudspeakers with crossovers

Infinite Line Source: analysis
Infinite Line Source: analysis
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 3rd December 2017, 09:49 PM   #101
gedlee is online now gedlee  United States
diyAudio Member
 
gedlee's Avatar
 
Join Date: Dec 2004
Location: Novi, Michigan
Quote:
Originally Posted by jlo View Post
I don't use any Green function. The simulated acoustic radiation is a brute force approach based on Huygens-Fresnel decomposition of sources and real-time superposition of all radiations at listener place.
I looked up Huygens-Fresnel in Wiki and it appears that these are based on a Greens function. That's what the propagation function is.

Quote:
Why not do it here ?
Personally, I'd like the OP to confirm that. It's his tread and I liked what he was doing.
__________________
Earl Geddes Gedlee Website
  Reply With Quote
Old 4th December 2017, 03:25 AM   #102
werewolf is offline werewolf  United States
diyAudio Member
 
Join Date: Feb 2004
Location: Austin, TX
POST #7

C. Infinite Line Source: frequency-domain pressure response

Coordinates : Imagine, if you will, a rather standard 3-D (rectangular) coordinate system, comprised of the 3 classic, orthogonal axes : x, y and z We'll consider the x-y plane being the "horizontal" plane, with the z-axis extending "vertically".

We're going to build our Infinite Line Source along (coincident with) the vertical z-axis The Infinite Line Source will have a constant volume acceleration per-unit-length of "Al".

But we start with baby steps Here's how we proceed :

First, let's consider a little (VERY little) "piece" of the Infinite Line Source, at some point "z" along the z-axis (meaning, our little "piece" will be at a distance "z" from the origin). Our little "piece" of the Infinite Line Source has a small (VERY small) length of "dz".

Now the clever bit : our little (VERY little) piece behaves just like an elemental monopole point-source ... which we just analyzed ... radiating with constant volume acceleration of (Al*dz)

What we need, to complete our fist step, is to identify the pressure measured at some random point in space, resulting from our little piece/elemental monopole. Let's start by placing our measuring point on the specific x-y plane of z=0, at a radial distance of "r" from the z-axis. The symmetry of our arrangement dictates that all measuring points at a distance of "r" from the z-axis will receive the same pressure from our little piece, no matter where in the x-y plane those measuring points reside. We also know that our little piece of the line is acting like an elemental monopole at some vertical height of "z". What's the distance from the little piece to our measuring point? Let's call it "R" ... Pythagoras reveals the result for us :

R^2 = r^2 + z^2
R = sqrt[r^2 + z^2]

Now we have what we need, to write the pressure from our little piece (VERY little) ... where the little piece is at some point "z" along the line, and we are measuring at some radial distance "r" from the line, in the x-y plane of z=0 We'll simply recall the pressure we discussed from a point-source, to identify :

pressure at "r" from little piece of line = {Al*dz} * {exp[jwt]} * {[rho/(4pi*R)] * exp[-jkR]}

where k = wavenumber = w/c
w= frequency
c = speed of sound

(we'll soon be ignoring that exp[jwt] term, because it's the time-dependent excitation ... which i like to separate from my "transfer function" analysis)


next up : we'll add-up all of our little pieces, to form the infinitely long line
  Reply With Quote
Old 4th December 2017, 03:28 AM   #103
werewolf is offline werewolf  United States
diyAudio Member
 
Join Date: Feb 2004
Location: Austin, TX
Quote:
Originally Posted by gedlee View Post
I looked up Huygens-Fresnel in Wiki and it appears that these are based on a Greens function. That's what the propagation function is.



Personally, I'd like the OP to confirm that. It's his tread and I liked what he was doing.
Gentlemen, thank you for the respect please feel free to take this thread in any direction at all, i rather like the engagement ... including detours

I'm labeling the main direction with "post numbers", so that readers can stay on the main street easily, if they so choose
  Reply With Quote
Old 4th December 2017, 08:26 PM   #104
jlo is offline jlo  France
diyAudio Member
 
Join Date: Nov 2004
Location: france
Quote:
Originally Posted by gedlee View Post
I looked up Huygens-Fresnel in Wiki and it appears that these are based on a Greens function. That's what the propagation function is.
Green's functions are solutions to wav equation in case of Huygens-Fresnel sources, but I don't need to use Green's functions.

I will try to explain the bases : look at post #7 just above and consider just one tiny point source. Signal at listener place is same as source signal but delayed and attenuated depending on this source to listener distance.
Consider another tiny point source : signal at listener is same as this source signal but delayed and attenuated depending on this distance.
Now total signal at listener due to those two sources is simply the real time addition of both signals.
For many sources, just add all signals : divide each loudspeaker in many tiny point sources. And a line array is just many loudspeakers.
Now computers are fast enough to process thousands of signals in real time.
With this brute force approach, you don't need to solve any equation.
It's quite simple but I have not seen this method used before.
__________________
jl ohl
ohl about audio

Last edited by jlo; 4th December 2017 at 08:32 PM.
  Reply With Quote
Old 4th December 2017, 08:35 PM   #105
gedlee is online now gedlee  United States
diyAudio Member
 
gedlee's Avatar
 
Join Date: Dec 2004
Location: Novi, Michigan
Whether or not you realize it you are using exactly the Green's Function approach to find the attenuation and the delay of each element, and the functions that you are using are the free field ones, hence you are not accounting for many aspects of the problem like diffraction of the enclosure, etc. Its an interesting approach to "real time" analysis, but people have been doing this off-line for more than a century. I am not sure that the limit on accuracy required to do real-time is a major advantage to a more accurate model that just takes a few more minutes. BEM and FEA, for example, can do this problem to an extremely high accuracy, just not in real time.
__________________
Earl Geddes Gedlee Website
  Reply With Quote
Old 4th December 2017, 08:50 PM   #106
jlo is offline jlo  France
diyAudio Member
 
Join Date: Nov 2004
Location: france
Quote:
Originally Posted by gedlee View Post
Whether or not you realize it you are using exactly the Green's Function approach to find the attenuation and the delay of each element, and the functions that you are using are the free field ones, hence you are not accounting for many aspects of the problem like diffraction of the enclosure, etc. Its an interesting approach to "real time" analysis, but people have been doing this off-line for more than a century. I am not sure that the limit on accuracy required to do real-time is a major advantage to a more accurate model that just takes a few more minutes. BEM and FEA, for example, can do this problem to an extremely high accuracy, just not in real time.
For attenuation and delay, I just have to calculate the distance. Diffraction of the enclosure is simulated by many secondary sources. Reflections (walls, floor,...) are simulated by image sources. It is just a question of accuracy vs computing power.
Also, for accuracy of primary sources and diffraction sources, you may not allways use point sources : an obliquity factor (see Kirchhof's diffraction) should sometimes be used.

One advantage is a true real time auralization (not a convolution) : you can change a parameter or a position while listening to the result.
__________________
jl ohl
ohl about audio

Last edited by jlo; 4th December 2017 at 08:54 PM.
  Reply With Quote
Old 7th December 2017, 07:36 PM   #107
jlo is offline jlo  France
diyAudio Member
 
Join Date: Nov 2004
Location: france
To explain the method, I just published a (basic) video here :
lapa - YouTube
__________________
jl ohl
ohl about audio

Last edited by jlo; 7th December 2017 at 07:42 PM.
  Reply With Quote
Old 7th December 2017, 08:08 PM   #108
gedlee is online now gedlee  United States
diyAudio Member
 
gedlee's Avatar
 
Join Date: Dec 2004
Location: Novi, Michigan
Looks like a nice piece of software. Why does the screen change for every type of analysis? Does each analysis have to be compiled or do the variables automatically show up as sliders?

I have never been much in favor of auralization except in the case of research, because in valid tests of "sound quality" they must be done blind. However a researcher could setup very nice tests with your software, which I would love to do. The only thing is that we have no money at all for this kind of testing - I typically do all the simulations myself and Lidia just uses her lab. Funding is never available. How would you like to donate this software for studies of diffraction, etc?
__________________
Earl Geddes Gedlee Website
  Reply With Quote
Old 7th December 2017, 08:28 PM   #109
jlo is offline jlo  France
diyAudio Member
 
Join Date: Nov 2004
Location: france
My basic question was : what phenomenas (distortions) are audible and to which level ? And to understand, you need to separate variables.
So I did one software to listen to diffraction, one for crossover auralization, one for reflexions and room modes, one for arrays, one for cardioid loudspeakers in a room, etc....
At the end, I could mix everything into one software only. The main problem would be too many parameters to set.
One tricky simulation is the waveguide (Earl knows it well), I tried to simulate it with image sources but it quickly gets very complicated.

To Earl : for research purposes, I would be glad to give you any software (for free of course).
__________________
jl ohl
ohl about audio

Last edited by jlo; 7th December 2017 at 08:37 PM.
  Reply With Quote
Old 7th December 2017, 08:36 PM   #110
gedlee is online now gedlee  United States
diyAudio Member
 
gedlee's Avatar
 
Join Date: Dec 2004
Location: Novi, Michigan
And what about the head diffraction? Do you have any, a generic model or does it allow for specific HRTFs to be used?

The reason that I ask is that Lidia and I are currently doing some research on very early wall reflections. We could certainly use software like that to generate our test signals. You could generate them and we would credit the software in our paper. That could be a big boost to your marketing. E-mail me directly at egeddes@gedlee.com if this interests you.
__________________
Earl Geddes Gedlee Website

Last edited by gedlee; 7th December 2017 at 08:39 PM.
  Reply With Quote

Reply


Infinite Line Source: analysisHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Short Line Array (line source) build perceval Full Range 226 4th January 2015 02:08 PM
Achieving Optimum Line Array Performance Through Predictive Analysis Patrick Bateman Multi-Way 2 27th June 2014 11:45 AM
Shunyata DTCD power line analysis jackinnj Power Supplies 1 16th December 2011 12:30 AM
Dynamic load line analysis using music and speaker load Michael Koster Tubes / Valves 0 7th March 2008 08:47 PM
line source diffraction david yost Multi-Way 8 5th June 2005 02:01 PM


New To Site? Need Help?

All times are GMT. The time now is 04:07 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki