AFMG Network Forum

Oriol de Pagès · Member

Hello,

I'm trying to validate the measured reverberation times in EASE .
I'm pretty desperate because different calculations as the Local Decay Time, Local Ray Tracing and Ray Tracing Impacts ( or provided Schroeder ), all give different results . Which is correct?

If you have measured a reverberation time of 1.2 seconds , it is necessary to set a time of 1000 ms in Ray Tracing Impacts ?

To calculate by Schroeder , it is necessary to add a Random Tail? If I export the impulse response to EASERA , the reverberation times are different.

You must have the module AURA to get reliable results ?

I am very confused on how to validate a model in EASE.

regards,

AFMG Pedro Lima · Forum Moderator

Hi,

If you have a Sabine or Eyring rooms (sound field is diffuse enough), you can compare them with Optimize RT (which will use use the formulas for Sabine/Eyring above). That will depend also on your goal. One would want to use it e.g. to calibrate the model based on real measurements. Then Optimize RT could be enough.

But, in general:

Local Decay Time: will give you a only rough estimate so that you can check if you have a diffuse field or if it varies considerably from one spot to another. It should not be used for assessment.

Local Ray Tracing and Ray Tracing Impacts should converge as you increase the number of particles and reflection order. The random tail need will depend on the Schroeder curve integration time. First, you must be sure you are already in the diffuse part of the response. Then you can compute the random tail if need (i.e. if the integration time is greater than the response length).

AURA has some nice tools that help speed that up. For instance, from AURA mapping you can directly view the T10, T20 and T30 and compare more directly with the measurement software output. In EASERA, for instance, you can select the above measures. AURA has accuracy gains, yes, but as you go higher with Ray Tracing the results should converge. The problem is actually the calculation time for that. That´s why you would add the random tail after you have diffuse enough field, or get AURA (where this addition is not needed).

But you can also insert a custom integration time in EASERA by setting the right and left markers (from.. to... setting in Schroeder curve details tab) and compare with the same integration time used in EASE.

HTH,
Pedro

Oriol de Pagès · Member

hi again,

In the project I'm doing the field it is diffuse and absorption coefficients are very low and very similar.
The makeshift that I have is this. The results by statistical theory differ significantly from the measured results , about 1 second. While the results for ray tracing are similar, especially in the tone curve , but they are not very accurate to be higher than those measured.
With the method of Schroeder results are lower than those measured but more like the tone curve .

I think it would be wrong to validate the model using statistical theory as it differs a lot. But to validate the model by ray tracing not know whether to modify the absorption coefficients of surfaces or scatering , as results by ray tracing are higher than those measured and Schroeder are lower.

Ragrds