First off, apologies for cross-posting...
I would just like to share the quick-and-dirty downmix of the premiere of my
latest work "Symmetries" (that took place at the last-week's "Linux
Conference" in Karlsruhe, Germany) with the LAU/LAD as well as Pd community.
Without you guys, this piece would never have been possible :-).
As my token of gratitude, in conjunction with this release I am also
releasing the soundfont that I've built from scratch using exclusively Linux
software (Swami, Rezound) and specifically for use in this piece. For more
info on each of these please see notes below.
As always, your feedback is much appreciated!
Symmetries (for computer and optional violin) is an experiment in relegating
musical structure and expression to the inherently stupid box of
transistors. By concurrently utilizing various GNU/Linux audio software
(Fluidsynth/QSynth, Pd, LADSPA, Jack-rack, JACK) it was composer's intention
to generate a lush interactive texture whose frail balance engenders a
consistent forward drive. In an ever-changing array of hierarchical
probabilities no two instances are expected to ever be the same. The piece
has been designed to be completely modular in terms of computer-driven sound
diffusion and can utilize 2-8 channels.
For its premiere the piece used 8-channel diffusion. However, the recordings
below are provided in a stereo-downmix form.
Hardware used in performance was eMachines m6807 laptop (64-bit AMD 3000+),
RME HDSP Multiface, and a Peavey 1600x midi controller that I used to
control some of the timbral nuances via Pd and Jack-Rack (LADSPA).
The violin part was played by Ania Zielinska (Poland) who commissioned the
There are 3 recordings available:
1) 128-bit (fixed rate) 48KHz OGG recording of the premiere:
2) 128-bit 48KHz MP3 recording of the premiere:
3) 64-bit 44KHz MP3 of the computer part:
The soundfont is a based on series of recordings of solo violin playing
straight tone "con sordino" sound. The recordings have been structured in a
gigasampler fashion (minor 3rd apart, except for the open strings). It has 2
sets of samples, ones without limiter which have already been mapped, and
other with the limiter which are in the soundfont but have not been mapped.
The sample tuning has been adjusted "by ear," so it may not be
mathematically accurate but FWIW it did pass my own scrutiny (which should
be taken with a grain of salt as this piece is not as demanding when it
comes to absolute preciseness of the tuning of individual pitches). The
looping of sounds is measured to provide most seamless transition while
accounting for the change of the direction of the bow. Conceivably one could
create a sense of orchestral sordino by layering the same sound over and
over. The soundfont does have reverb and chorus abilities enabled but IMHO
it sounds the best without any chorus applied to it.
The soundfont is released under the "GPL/Artistic 2.0" license (for more
info please see: http://dev.perl.org/perl6/rfc/346.html
). Btw, I chose the
art-related license simply based on my limited understanding that it is more
appropriately tailored towards something that is not code-based. That being
said, if anyone can explain me the difference between the two licenses, I
would really appreciate it :-).
To download the soundfont please click here:
NB: Considering that the Linux server that is hosting this may be on its
last legs (strange noises from the HD), I would not mind if someone would
consider mirroring this particular file. Many thanks!
Once again, I would like to extend my thanks to the developers and users
alike of the open-source, and perhaps more importantly, Linux audio
Many thanks also go to the organizers of the Linux Audio Conference for
making this performance possible!
Ivica Ico Bukvic, composer & multimedia sculptor