LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

sound intensity Measurement

Hi there!. I'm developing an application in LabView to calculate Sound Intensity (indirect method, meaning frecuency domain method). I know that there is hi-end and low-end components to calculate cross spectrum. Ok, quiestions:
1.Which one is the component to calculate Cross-Spectral Density (CSD)?. I realized there is 5 or 6 components related to this, but no one of them literally says CSD. I would say that the component should be Cross-Power Spectrum, but not quite sure.

2.I want to get the results using 1/3 octave filters, but the components fot 1/3 octave are not compatible with cross-spectrum components, no matter which component comes first in the process.

3.The formula used to calculate indirect sound intensity has a division by "w" (omega), wich is frecuency, thet means that somehow i have to accomplish this. How do i do it in LabView?

Extra info:
.I'm using a professional audio interface to acquire the
signals.
.I'm kind of new to labview, so i would appreciate lots any help.

Thanks to averyone helping me here!!
0 Kudos
Message 1 of 6
(3,673 Views)

Hi duende this information might be useful

http://zone.ni.com/devzone/cda/tut/p/id/1469 

http://zone.ni.com/devzone/cda/tut/p/id/4278 

Also you can find this forum interesting

https://forums.ni.com/t5/LabVIEW/power-spectral-density-vi/m-p/255879?requireLogin=False

With your question there is a PSD you can configure form the IV Express

And with the formula you do not need the formula directly use the link above

, for the question 1 there is a vi for that and is called And you can make your post directly in Spanish in the Spanish community message board

 

Benjamin C
Principal Systems Engineer // CLA // CLED
0 Kudos
Message 2 of 6
(3,640 Views)
Ok, muchas gracias Benjamin, los links ya los habia leido por lo tanto no despejaron todas mis dudas. En este preciso momento necesito saber como aplicar la siguiente formula I(ω)=-(1/ρ*ω*Dr)*Im{Gp1p2(ω)}.

La parte de esta formula que me complica es que como I es funcion de (ω), no se de donde sacar el valor de ω del componente cross spectrum VI. Lo demás de la formula esta solucionado.

Gracias por la info anterior!

Ok, thanks Benjamin. I have read the links yo gave me, but i still have some questions. Basically right now i need to know how to implement the following formula
I(ω)=-(1/ρ*ω*Dr)*Im{Gp1p2(ω)}.

Specifically what i need to know is how to get the value of ω from the cross spectrum VI, because I is a function of (ω).

Thanks for the information above.
0 Kudos
Message 3 of 6
(3,622 Views)
 

Hola Duende perdón por la tardanza, pero hay bastante información y dan diferentes o

 

 

Por ejemplo aquí http://ccrma.stanford.edu/~jos/mdft/Cross_Correlation.html

 

 

Mencionan que cross correlation es básicamente lo mismo que el cross spectral density en este caso podrías usar el de correlation

 

 

Sin embargo me parece mejor la definición en la siguiente pagina

 

 

http://zone.ni.com/devzone/cda/tut/p/id/4278 

 

 

aquí no tienes que calcular la w simplemente es (FFT(a) x FFT (b) )/N^2

 

 

en donde N es el numero de puntos que tomaste en el dominio del tiempo.

 

 

Espero que esto te sea de utilidad

 

 

Saludos

 

 

Benjamin C
Principal Systems Engineer // CLA // CLED
0 Kudos
Message 4 of 6
(3,588 Views)
Gracias por la info. Aunque creo que no entendiste bien mi pregunta. Lo que necesito saber es de donde saco la informacion de w, o como implemento la funcion mencionada, ya que esta depende de la frecuencia y esta por si sola esta en el divisor de la ecuación. Es decir que
I(w) al ser funcion de w significa que para cada valor de w obtengo un valor de I. Esto es lo que no se como implementar en Labview.
 
De todas maneras te agradezco nuevamente y estare atento a una nueva respuesta.
Atte peter
0 Kudos
Message 5 of 6
(3,525 Views)
Hi,
I had calculate real part cross spectrum which says me something about phase between two signals. When two signals are exactly in phase sign is +, when out -. I try to get a result of real cross spectrum in octave bands i.e. third without loosing sign information. Is that possible in LV ?
0 Kudos
Message 6 of 6
(3,276 Views)