I-Simpa icon indicating copy to clipboard operation
I-Simpa copied to clipboard

Published 20 hours ago • verified publisher icon Universite-Gustave-Eiffel

Scattering issue

Open wbinek opened this issue 10 years ago • 3 comments

I am currently working on my bachelor thesis, on AGH University in Poland, in which I am comparing I-SIMPA with CATT-Acoustics.

I've noticed strange issue. The result of simulation is extremely dependant on scattering. With no scattering the results are quite similar to ones from CATT, but when I set proper scattering values, then RT increases 2 up to 3 times. I know that adding scattering changes RT, but I think, the difference is much to big. Especially taking into account, that the same changes done in CATT produce more or less good results in both cases. Have you encountered such behaviour? I attach an image of measured and predicted T20 with different scattering settings.

I am looking forward to receiving your reply.

diff a t20

wbinek avatar Dec 23 '14 21:12 wbinek

I've found solution. Reflected direction vector was not normalised, so the particle speed was decreasing with each diffuse reflection.

The solution is: https://github.com/d-cooper/I-Simpa/commit/5e98f6193eca8a52e8d0904550010ec25cc87452

wbinek avatar Jan 14 '15 16:01 wbinek

Thanks for the feedback. We will investigate as we work on continuous integration (Unit tests + integration of existing validation of Spps).

nicolas-f avatar Jan 18 '16 13:01 nicolas-f

The patch has been applied

However there is no corresponding unit test

nicolas-f avatar Jan 31 '17 14:01 nicolas-f