[PD] spatialisation/convolution-related

J. Scott Hildebrand jshildebrand at ucdavis.edu
Wed Aug 14 23:01:46 CEST 2002


      hi mandy,

          first of all, you live in bristol? i love it there! i went by
myself from london just to see a band that i'm in love with. i got a bit
of a chance to explore the multimedia stuff, but for the most part i was
record shopping. i didn't get to the bridge though; i really wanted to see
that.

          anyway, i'm working on implementing HRTF convolution in Pd. i'm
not a VERY experienced programmer but i have some skills, and i have a lot
of people in the spatial audio field helping me out. i actually wrote my
own convolution patch but it was brute-force, and in the end it took too
many CPU cycles. it just plain old didn't work, so now i'm using FFTs to
do it and hopefully the CPU can handle multiple sources. how many HRTFs do
you have? i'd like to eventually modify the external so that it would be
very easy for anybody to use with however many HRTFs with however many
taps.

                   scott



On Wed, 14 Aug 2002, MLN Kazem, Aerospace Engineering wrote:

> Hi people,
>
> I suppose I'm a complete newbie and therefore inevitably will ask dull
> and repeated question.. but.. I am looking to produce as near as
> effective 3D sound in real time on a PC (windows I'm afraid). At the
> minute I am using a sdk which sits on top of the DirectSound API which
> sounds pretty good (to me anyway) but has anything to do with HRTFs
> been implemented in pd? I've heard of spat~ but I am presuming that
> this is a.)max related b.)relies on external DSPs and c.)not free.
>
> If nothing else newbies like me must at least create a little
> amusement....
> thanks
> Mandy
>
> ----------------------
> MLN Kazem, Aerospace Engineering
> M.Kazem at bristol.ac.uk
> Tel. 0117 9288219
> Fax. 0117 9272771
>
>




--------------------------------------------------------------------

	"640K ought to be enough for anybody." -- Bill Gates, 1981

--------------------------------------------------------------------





More information about the Pd-list mailing list