- Well the performance (called "Henry's Beard", strangely after
my sister's hilariously obese cat) went great. People liked our cube part,
and I liked it too. Demetrius and I needed to do a better sound check,
but we were a bit rushed. Very fun. And loud. You
can see images of the performance too. No sounds yet, hopefully that
will come soon. Curtis absconded to China with the video footage and will
be back sometime in June. Of course the Percussicube was finished for
the concert. I just didn't have time to update the site. If anyone wants
any info on how to build a similar instrument, email me, I'd be more than
glad to help. Later.
- Added various sample based sound generation thingies. Attempted
to make it much more dynamic and interesting to listen to and to play.
Samples consist mostly of glass/plastic bottles being bounced, crushed
and jumbled about in buckets (real ones, not the max object :) . Thinking
about performance options for Friday. Going to have several different
modes for performance through which I can switch.
- I found that it is really fun to play my instrument with a
hammer and to rip it apart during playing and beat it with its own parts
and step on it and hit it with my head and lift it up and drop it etc
until it falls apart completely and that is the end of the performance.
The WHO all over again! But this one goes to 127!... and can be easily
put back together again. And of course the Powerbook sits a safe distance
- Made beat sensor, a way for it to repeat what I did last measure in
a different form, more sound generation stuff - I'm finally starting to
like some stuff I am getting, though it needs to become more diverse.
Began to integrate mouse preset control. Added "Blow Up" feature
where if the time-density of hits is above a certain level for the last
12 hits, volume increases, overdrive kicks, and a signal rising in pitch
kicks in. Also I think I have fixed some of the issues in the averaging
of hits that didn't work because I had neglected to consider the order
in which Max executed the commands. A few [t b b i] objects saved the
day - I think. We will soon see.
- Reworked data processing AGAIN. All to stuff below just
did not work in the real world and made the thing
way to delicate to play correctly. I have now revised the system so it
averages all the velocities and coordinates of the any triggers that go
off together with added weight to the two that occur first and diminished
weight to those that occur last. This now determines position. I have
thrown out all the other crap. In doing this, the Lobjects
for Max have become my very good friends for list operations. Also I have
developed a new way to sense tempo that uses [lp.stacey] to detect when
I am making a continuous beat (one that has a low standard deviation of
time between hits) and fire only then, instead of just using the mean
of to last x hits within y amount of time which yields unwanted results
when I am playing more complex rhythms.
now I can beat the thing like crazy and it sort of works! This will be
a much better performance than my first ideas. Maybe I can even take it
to where the thing will fall apart by the end of the performance, as the
cube part is nearly indestructible and if so easy and very cheap to repair
- that would be crazy. Now to make it sound cooler!!
- Check out new IMAGES.
Reworked data processing, added system to average two adjacent triggers
together (if I bang on the tube between them) to estimate where hit the
tube and sent the object there. Also added control from the drum unit's
sustain pedal where if it is pressed, and hit sends the object immediately
to its destination, more like real percussion and possibly more responsive
for playing. All of this is very complex for me, and I still don't know
how it will really work. Here's hoping. Also I have reworked the sound
somewhat to sound when a trigger goes off no matter what, so that quick
movements (hard hits) will not go un-sonified due to the delay induced
by calculating velocity and basing attack of the sound on that.
played it for the first time yesterday. It was not too great. "Anti-ergonomic"
to play and the sounds it makes are not that great yet. So I must really
rethink / rework stuff. But I have ideas! And by the time it is "finished",
everything below may be inaccurate. Sorry
- no new Info lately. Project has been on hold while I was in Rome for
the week with the Brown Jazz Band.
digital instrument using a Yamaha PMC-1 percussion to MIDI converter (see
photo below, under powerbook) to 'pilot' a virtual object, the position
of which controls various aspects of the sound produced. The 'brain' of
the instrument will be my powerbook running Max/MSP. Here is a drawing to
give you an idea:
will be probably be generated in two possible modes, one where hitting a
trigger node will trigger sound, the other where sound will be triggered
somehow depending only on the motion of the virtual object through space
- based on velocity and/or acceleration. Another foot pedal will control
a 'freeze' function [no Demetrius,
not a "freeze the computer" function] that will stop the object
where it is when one achieves a sound one wants to keep for a time. A continuous
controller, maybe in the guise of a pressure sensor on one stick will control
intensity/volume in some way. A memory function is also a possibility.
have a simple realtime 3D visualization that can be used by the performer
and/or viewed by the audience. It is done in the free Pixelshox
environment, which accepts MIDI controller data in realtime.
'Physics' of the virtual space currently include imitation of either a frictional
or accelerating force on the object, 'wobble' (oscillation in 3D, the magnitude
and frequency of which can be controlled with the instrument) and a sort
of pseudo-inertia (proportional to velocity) that comes into play when the
object changes direction, making it continue on its previous path for a
small amount of time before swinging around.
instrument can be small and single-player or up to 30 feet square and played
by many people! Right now I am having problems trying to figure out how
the frame that holds the triggers in space will be constructed. I would
like it to be very transparent to the audience (no sides, just edges of
the cube), portable, easily reconfigurable to a different size cube, and
still tough enough to stand up to being hit with drumsticks. Any suggestions?
current state of affairs,
Frame: removable coupling for triggers so they can be easily changed,
reconfigured, put on different size frame etc. Where
can i find cheap XLR couplings? Need to pad triggers also to protect them. Mousepads...
resistors - allelectronics
mounting/covering - rubber feet that I have?
cable - allelectronics
of hitting one pad determines freq of granularity/delays/ rhythmic
elements Begun to implement pattern recognition for specific actions - series of
hitting. Prob. not. 'blow up point' at which number of hits within X time
causes something - granular explosion, self-driving oscillator? Done.
But I have not yet tried the self driving oscillator! It's a bit scary! sound when hitting - goes sounds when note off received
right after note on (within X) - i.e. when PMC sustain pedal is down.
(PMC must be set fast gate to do this) possibly some sampled 'bouncing
object' sounds (cans, pots, pans, boxes, shoes, fruit, animals?) that
can be recognizable to listeners and triggered more instantaneously.
for movement of virtual object - Basically
Implemented with FM, granulation and panning controlled by position in
XYZ, volume controlled by velocity. Needs help. Got
help. Still needs more.
two trigs at once freezes? Implemented
with "two hits at once freezes" (within 10 ms - may need to
be adjusted) Possibilities: (until
pedal is down for a hit? ie note off is within X like above?) Or kept
on by pressure - when pressure is at high or low? value it turns off,
or is on only if pressure is above/below certain value? Removed. Mouse: scroll through presets with wheel - normal is
presets for movement based sound, press down scroll is for percussive
sound for hitting, other buttons can either be used for press-scroll to
set a value (unfortunately these values have resolution of 0-10 as that
is most you can scroll in one finger motion.) or for trigger/toggle of
things that need not be used always when playing. mouse position will
not be used. This would not be useful here. Too much. Just a source of
buttons for now that are far from the computer.
In progress. May also get a numeric keypad to use for controls or
to take apart and put into other buttons/switches/foot pedals/digital
sensors. For now, no. Visual: lighting - lights up when hit (need to figure
out how to do this nicely using 3D modeling/lighting), changes colors
of background and/or object on preset selection through spectrum so u
can remember - "I liked sound of green, lets go back to that - it
was between yellow and blue". The audience
will probably not need to see this part. We shall see.
info and maybe Max/MSP and pixelshox patches coming soon.