equivalent impulse response

Hello, I'm Seb and I have a little problem I would like to submit to you.


At the time of an acquisition of an impulse response of a room (room of 
music, canteen or another thing), it is possible that this answer is 
disturbed because of the noise (a car which passes, a person who speaks 
etc). Thus the nature of the noise is not known. It would thus be 
necessary, from various acquisitions of impulse responses, to find the 
equivalent impulse response, ie the one without noise. 
I have no idea how to proceed, please help me. I'm not asking for a code 
but more for a method.

Please forgive me for my english.
Thanks

On 7 Jul, 15:48, System Administrator <in...@bienentendu.fr> wrote:
> Hello, I'm Seb and I have a little problem I would like to submit to you.
>
> At the time of an acquisition of an impulse response of a room (room of
> music, canteen or another thing), it is possible that this answer is
> disturbed because of the noise (a car which passes, a person who speaks
> etc). Thus the nature of the noise is not known. It would thus be
> necessary, from various acquisitions of impulse responses, to find the
> equivalent impulse response, ie the one without noise.

If your measurements are contaminated by noise, it makes
sense to do a number of measurements to obtain a good
result.

Rune

System Administrator wrote:


> At the time of an acquisition of an impulse response of a room (room of 
> music, canteen or another thing),

Pretty useless measurement. Why do you need it?

> it is possible that this answer is 
> disturbed because of the noise (a car which passes, a person who speaks 
> etc). Thus the nature of the noise is not known. It would thus be 
> necessary, from various acquisitions of impulse responses, to find the 
> equivalent impulse response, ie the one without noise. 
> I have no idea how to proceed, please help me. I'm not asking for a code 
> but more for a method.

1. Use bigger caliber gun to generate the impulse

2. Use frequency sweep.  Derive the impulse response from the frequency 
response.

3. Use pulse compression

4. Accumulate the impulse response from several pulses


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com

Google search "maximum length sequence impulse response measurement".

Greg Berchin wrote:
> Google search "maximum length sequence impulse response measurement".

Or you can try the sweep variants, for example Aurora (the price is
right):

http://www.aurora-plugins.com/Aurora_XP/index.htm.

With regard to immunity against noise, I remember taking loudspeaker
measurements while the auditorium was still under construction. The
TEF algorithm (sweep based) seemed to be doing just fine at rejecting
the unwanted noise. I could never hear any constructions workers
swearing or dropping hammers or drilling holes in concrete walls on
the impulse responses :).

Regards,
Andor