Hi there,
I'm doing an RT60 calculation of a room at different frequencies. I'm
using 1/3 octave noise for frequencies smaller than 512 Hz and octave
noise for highest frequencies. I've recorded all the reverbs and now I'm
feeding matlab for the analysys. I've some question:
1) Could I bandpass the signals (that now are only short reverberations)
to remove the unwanted noise (the room is not perfectly isolated!) and
if yes could I bandpass It with filters that have the same range of the
original octave or 1/3 octave noise (Example: i've recorded the reverb
of a 512-1024 Hz noise. Before giving it to the RT60 calculation
algorithm could I bandstop frequencies >1024 and <512 safely?)
2) Have you any Matlab function the calculates RT60 from a wav file?
3) Could be the hilbert transform suitable for enveloping purposes?
I'm a beginner, please be patient!
Bye,
Giorgio
Angelo Campanella - 01 Apr 2004 20:40 GMT
> Hi there,
> I'm doing an RT60 calculation of a room at different frequencies. I'm
[quoted text clipped - 9 lines]
> 2) Have you any Matlab function the calculates RT60 from a wav file?
> 3) Could be the hilbert transform suitable for enveloping purposes?
It is important that the amplifier and measurement system (and
microphone) never becomes nonlinear in amplitude transfer. That is
spurious harmonics and noise should never occur when the amplitude
increases a little more.
Generally, frequency filtering (low pass, high pass, band pass) is
always done and it is always acceptable. The nonlinearities, as I have
said, must always be investigated and must always be staisfied (that
they do not ever occur).
jd - 02 Apr 2004 09:03 GMT
Hi Giorgio,
> 2) Have you any Matlab function the calculates RT60 from a wav file?
Currently at "Matlab Central File Exchange" there is only it:
http://www.mathworks.com/matlabcentral/fileexchange/loadFile.do?objectId=1212&ob
jectType=file#
but it seems it doesn't fulfill your requirements.
Try Google - there are some links to pages concerning that problem
Regards,
jd
P.S. As far as I know long long time ago there was at Matlab pages script
which fits to your problem. I will check my files - maybe I still have it.
Please write to my private e-mail.