impulse responses??
Hello
a bit offtopic, but that's for one my patch...
Does anyone know any free libraries of impulse responses captured NOT from acoustical spaces but from anything else - electrical circuits, resonances in materials or anything else?
Thanks a lot
Regards, Alex
I think the process you are looking for is convolution, where you map
one source onto another. If I remember there were some good examples
from the summer workshops from CNMAT (I think that's where they were
held)...past workshop patches I think are still on-line. Also
soundhack had a pretty good non-realtime utility for this. I use to
play with mapping some of my clarinet playing with sounds of traffic
this way (a very Los Angeles thing to do :)...back in my youth...
I don't have links handy, but they shouldn't be hard to find.
David
On May 24, 2007, at 12:01 AM, Alex wrote:
>
> Hello
>
> a bit offtopic, but that's for one my patch...
> Does anyone know any free libraries of impulse responses captured
> NOT from acoustical spaces but from anything else - electrical
> circuits, resonances in materials or anything else?
>
> Thanks a lot
> Regards, Alex
thanks for your messages!!!
enjoyed that IR library on efnet site, some nice samples there.
you're right, i'm assembling a sort of convolution algorithm using buffir~ (if anyone can recommend the how to make the same thing via fft i wil say a REALLY BIG THANK YOU :)
and i want to use some recognizable filter responses with complex morphing algorithm.
and for the mother lode:
http://www.kvraudio.com/forum/viewtopic.php?t=62458
Can't guarantee that they aren't all acoustic spaces,
but I remember finding some interesting non-space
impulses amongst the Acoustic Mirror stuff - trouble
is that they're in some proprietry format, though i
did manage to convert /some/ of them using SoundHack's
header munging function,
cheers
Roger
--- Alex wrote:
>
> Hello
>
> a bit offtopic, but that's for one my patch...
> Does anyone know any free libraries of impulse
> responses captured NOT from acoustical spaces but
> from anything else - electrical circuits, resonances
> in materials or anything else?
>
> Thanks a lot
> Regards, Alex
>
don't know if they're good but:
http://www.xs4all.nl/~fokkie/IR.htm
hey, could you share your work? the buffir~ limited size thing is
disappointing , it would be nice if you'd find a workaround.
happy patching : )
isjtar
On May 24, 2007, at 10:41 AM, ROGER CARRUTHERS wrote:
> http://www.memi.com/echochamber/responses/index.html
> (German)
> http://www.cksde.com/f_6.htm
>
> and for the mother lode:
> http://www.kvraudio.com/forum/viewtopic.php?t=62458
>
> Can't guarantee that they aren't all acoustic spaces,
> but I remember finding some interesting non-space
> impulses amongst the Acoustic Mirror stuff - trouble
> is that they're in some proprietry format, though i
> did manage to convert /some/ of them using SoundHack's
> header munging function,
> cheers
> Roger
>
> --- Alex wrote:
>
>>
>> Hello
>>
>> a bit offtopic, but that's for one my patch...
>> Does anyone know any free libraries of impulse
>> responses captured NOT from acoustical spaces but
>> from anything else - electrical circuits, resonances
>> in materials or anything else?
>>
>> Thanks a lot
>> Regards, Alex
>>
>
Hello again!
First thanks everyone for the help!
Of course, i'll share, just will write some comments to that patches. Give me a couple of days.
The problem is very high CPU consumption (i'm on 4.5.7 on PC) of buffir~
I'm using a pair or 4 buffir~ with the parts of IR and the delays to the audio signals. It's still not useable for creating a reverb unit, i'm using it just for filtration. And sometimes 256 is enough for that purposes. I've also attached multislider for drawing a buffer and pattrstorage for making snapshots and morphing.
MSP Tutorial 26 shows how to do convolution using FFT's.
Yes, i know that, i'm using several modules based on such way of signal processing. but not sure how to use that for impulse reverb or convolution with long static impulse, not for convolving 2 audio sources.
As i know, all that IR reverbs as Waves IR 1 or anything else use FFT for less CPU consumption, but i don't know the algorithms how can this is done...
Alex schrieb:
> The problem is very high CPU consumption (i'm on 4.5.7 on PC) of
> buffir~
That's the nature of convolution, it will eat CPU exponatially with the
ir size. 512 will eat 4 times as much as 256...
And daisy chaining is not the solution, but might lead to "interesting"
sounds...
> As i know, all that IR reverbs as Waves IR 1 or anything else use FFT
> for less CPU consumption, but i don't know the algorithms how can
> this is done...
As I know there are some patented algorithms... But some info should be
available on the net. And if its patented you should be able to read the
description, as this is what patents are about...
Stefan
--
Stefan Tiedje------------x-------
--_____-----------|--------------
--(_|_ ----|-----|-----()-------
-- _|_)----|-----()--------------
----------()--------www.ccmix.com
well, i can say i sometimes need "interesting" sound rather than technical exactness, but this time i thought i was right to get the desired result. can you point me where i'm wrong thinking that chaining o buffirs~ helps in this way?
if anyone needs "results" of my work - see the attach.