Early this year I went to watch a perforance of Kentic sound artist Ray Lee at Kinetica.
Taken from the site http://www.kinetica-museum.org
Kinetica is the UK’s first museum of kinetic art. It will actively encourage the convergence of art and technology, providing an exhibition space in central London where the most important examples of kinetic, technological and electronic art, both past and present, can be properly stored and displayed.
Ray Lee was giving a performance with his collection of instillations, creating strange sights and sounds. There was an odd apparatus in the centre of one of his popular instillations, Sirens. It had a rod on either side. He pressed a hidden switch and began to do some very precise hand movements, adjusting what he needed to on the machine at the same time. He moved closer and further away from the metal rods, the closer he became the higher the pitch and vise versa and looked as if he was playing with invisible forces. He went on for a while, making random noises then a violin started playing in an oriental theme. His hands were trembling, his eyes were closed tightly and he was really focused. It was so beautiful, almost romantic. Another sound was also present, it sounded like a dinosaur snoring with its mouth closed, a heavy base.
Ray Lee used a kinetic sound machine to generate electromechanical sound. He had many different installations but this particular instillation caught my attention. From mechanical structures to mechanically produced sounds.
Image of Ray Lee playing on his instillation.
Inspired by Ray Lees performance I will also like to play music using artificial sound. I looked through the web for two days and stumbled across a video. I can’t seem to upload a video to this blog but you could find it on this link http://www.billaboop.com/
You should also check this out http://www.youtube.com/watch?v=YqyjnFNQ96Q
Watching these videos reminded me of Ray Lee’s performance
Information from www.billaboop.com
KTDrumTrigger is a VST plugin with custom editor that triggers MIDI notes based on the sound level of the incoming audio stream in different frequency bands. It allows you to “detect” occurrences of percussive sounds in an audio stream and send out a MIDI event whenever that happens.
It works like this: incoming (mono) audio is split up into different frequency bands using three tweakable filters. The signal in each of these bands is sent through an envelope follower. When the envelope reaches a specified threshold level, a MIDI note on message is generated. The trigger module has some built-in hysteresis and minimum inter onset time settings to avoid fast successive triggers due to noise on the envelope curves. Visual feedback of the filter and trigger settings is available….
Note that this is a rather simple system for finding drum onsets in an audio stream. It will probably only work reliably on isolated drum tracks (no mixed-in music), it’s quite sensitive to volume changes and it might not always be possible to separate the different types of drum sounds using simple filters. However, if you just need a fast way of transcribing your isolated drum tracks into MIDI so you can tweak them further or apply the groove to other parts, it might well do the trick and save you a lot of time 
That said, no one says you *must* use this on drum tracks! You might come up with some other applications where this thing comes in handy or use it to extract a rhythmic sequence out of an instrument track (although this will only work with instruments that have clear onsets). As always: tweak those knobs and experiment!
This is a great demo made by ModuLR and demonstrates how KTDrumTrigger can be used for turning beatboxing into synthesized percussion. More info on how he did this can be found in this thread on KVR.
In the picture below you can see a very simple experimental setup I did at home on my desk, just to see if it could work. I routed the audio picked up by the microphone through a single instance of KTDrumTrigger with one low pass filter and one high pass filter and recorded the generated MIDI back to a track in Cubase.
- An empty shoe box with a mouse mat on top (use the backside of the mat to get a muffled sound).
- An almost empty can of transparent metal varnish (anything that will produce sounds with higher frequencies when hit).
- A simple microphone.
- Two old drum sticks.
CONCEPT PROPOSAL
(Assuming you have watched the videos on the links provided) I intend on creating similar Virtual artwork using real-time expressive performance (like Ray Lee) using VST or Virtual Studio Technology (plug-in) to create music and effects.
Information from soundonsound.com
The VST (Virtual Studio Technology) system was developed by Steinberg to enable a complete studio to be created in software. Even in its earliest incarnation, it allowed third-party developers to produce real-time effect modules that could ‘plug in’ to the host application. However, when Steinberg introduced the second version of the VST plug-in standard it also became possible to send MIDI data to and from such effects. This enabled developers to add more features, such as MIDI control of effect parameters and locking of effect settings to tempo. The inevitable result of this advance in the protocol was that this MIDI information was also used to run synth engines, rather than just effects processors. It is these synths, masquerading as effects plug-ins in order to fit directly into the sequencing environment, that are called VST Instruments.
To begin I will need a VST compatible host that is able to load the plug-in. My set up will be very similar to the image above but also using my body to create sounds. The audio signals (sounds) will pass through the microphone and will be analysed and converted into MIDI events. I would need a MIDI device or MIDI VSTi (VST instrument) that will be triggered by the plug-in.
A VSTi can create a huge variety of sounds from MIDI and Audio sequencers.
The host I have chosen to use is called Wavosaur. Wavosaur is a software useful for mastering, sample looping, sound designing, audio mangling etc. It’s
- Fast and efficient :selecting, zooming etc.
- Reliable waveform representation (without parasitic phenomenons due to graphic aliasing).
- Real-time manipulation for effect (no more “few seconds modal preview”).
- ASIO & VST support.
Image from www.wavosaur.com
My initial idea was to use sound create interaction virtually with a PC. The Virtual Space Technology that creates real-time expressive performance using sound couldn’t be more relevant to the research I have developed. Although I saw Ray Lee playing nothing and just moving his arms in the air, I heard a violin play. The relationship between my eyes and my ears were reassessed. I know from memory and life experience that if I hear a violin play I’m bound to see someone playing the violin. I saw no violin but I heard it play, its was as thought the sound was plucked out of space to play in the air around me.
Posted by dartiste 
