I suppose you need to apply deconvolution to get
the impulse. Fons Adriaensen 
is your man  to give the exact details (I have
not done that myself yet). He's got a software for
IR measurement.

Victor


----- Original Message ----- 
From: "Oeyvind Brandtsegg" 
To: 
Sent: Wednesday, June 04, 2008 11:56 AM
Subject: [Csnd] Re: Re: pvsmaska


> Ah, you're right. Thanks. I did not think of that.
>
> As for the sine sweep method of recording an impulse response,
> how do I post process the recording of the sine sweep to leave only
> the "response" and get rid of the original sine sweep ?
> Could I just subtract a delayed version of the sine sweep to phase out
> the original signal in the response recording ?
>
> best
> Oeyvind
>
> 2008/6/3, Victor Lazzarini :
>> Oeyvind,
>>
>> 1. instead of using pvsmaska, you could try pvsfilter
>> that does the same thing but with two fsigs. Note that
>> both pvsmaska & pvsfilter will not affect the phase
>> of the signals, so the transformation is magnitude-only.
>>
>> 2. I would have thought you could record the impulse
>> response of your system and then use pconv to
>> process the audio. That will use mag+phases
>> (full-featured convolution).
>>
>> Regards
>>
>> Victor
>>
>>
>> At 10:24 03/06/2008, you wrote:
>> >
>> > Hi
>> >
>> > I'm trying to write an instrument that will analyze the frequency
>> > response of my audio system (the signal path from Csound output, via
>> > speakers, microphone, and back into Csound).
>> >
>> > I want to use the result of the analysis to shape an audio signal (all
>> > inside Csound) so that it will sound (as close as possible) like it
>> > has been played through the speakers and recorded back into Csound.
>> >
>> > I figured I could use the pvs opcodes, with pvsftw writing a table
>> > with amplitudes for all frequency bands in the analyzed signal (audio
>> > system response simulated with an arbitrary filter in the attached
>> > csd), then use pvsmaska to shape the frequency response of another
>> > signal.
>> >
>> > This works in terms of frequency response, but the overall amplitude
>> > is very low for the "shaped" signal (aout in instr 2 of the attached
>> > csd).
>> > I wonder if there are some formula for gain makeup in this case ?
>> > It seems to me in theory this should just work "out of the box), with
>> > no need for makeup gain, or am I way off in supposing it should work ?
>> >
>> > best
>> > Oeyvind
>> >
>> > Send bugs reports to this list.
>> > To unsubscribe, send email sympa@lists.bath.ac.uk with body 
>> > "unsubscribe
>> csound"Content-Type: application/octet-stream; 
>> name=auto_calibrate_pvs2.csd
>> > X-Attachment-Id: f_fh0a9i9o
>> > Content-Disposition: attachment;
>> filename=auto_calibrate_pvs2.csd
>> >
>>
>> Victor Lazzarini
>> Music Technology Laboratory
>> Music Department
>> National University of Ireland, Maynooth
>>
>>
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe 
> csound"