[music-dsp] Patent: Efficient convolution without input/output delay

[Christopher Weare]

I am stating the NRC does not serve as prior art because it does not
reference either the specific technique or application described in the
patent.  I did not imply anything about convolution for reverb in
regards to other sources.

Furthermore, the lake patent does NOT "patent" convolution for reverb.
In fact, in the "Prior art" section of the patent convolution via FFT is
described.

Application of an invention is important as it is stated in the "FIELD
OF THE INVENTION" section of the patent.  If it is reasonable to assume
that the application of a pre-eqisting technique to a new field is a
novel application, i.e., not an obvious application of the technique to
a practitioner of the art AT THE TIME OF INVENTION, then the applicant
has a case for patenting his/her invention.  It is easy to claim
something as obvious after the fact.  But if you wish to contest this
point after a patent has been issued then go hire some lawers and get
ready to pay.  Everything is obvious after the fact.

However, if you have prior art showing that invention being used in the
same fields as claimed in the "Fields of the invention" section then you
have a pretty strong case.

Now, if you wish to have a chance at contesting the patent then you need
to produce prior art that describes just such a system "In the Field of
the invention" (which in this case is actually broader than audio, they
mention radar etc.)" that was published prior to July 7, 1992.

A very similar patent discussion was dragged through the mud some years
ago in reference to a patent by IVL where a key component of the
invention was resampling to affect a formant change.  Many of us
complained that the application was obvious.  Clearly the work by Lent
constituted prior art.  Certainly resampling was a well established
technique.  If ya wanna set RBJ off ask him about this one ;-).

Now is this system fair?  

*shrug*

Having been on both sides of the fence I now see it as less evil than I
did before.

-chris

-----Original Message-----
From: Jon Watte [mailto:hplus at mindcontrol.org]
Sent: Tuesday, March 06, 2001 10:48 AM
To: music-dsp at shoko.calarts.edu
Subject: RE: [music-dsp] Patent: Efficient convolution without
input/output delay



> Numerical Recipies does not describe zero delay convolution, only
block
> convolution.  Numerical Recipies also does not discuss using this
> technique for reverb.

Are you seriously suggesting "using convolution for reverb" 
is somehow novel in the field?
Wasn't that what experimental computational music people did 
in the sixties?

Cheers,

				/ h+


dupswapdrop -- the music-dsp mailing list and website: subscription
info,
FAQ, source code archive, list archive, book reviews, dsp links
http://shoko.calarts.edu/musicdsp/


dupswapdrop -- the music-dsp mailing list and website: subscription info,
FAQ, source code archive, list archive, book reviews, dsp links
http://shoko.calarts.edu/musicdsp/