Projects
Ongoing research with Bob Lester developing new performanc interfaces and electronic music solutions, more here.
-
DK II (Custom, Hand-built Pedal Controller)
The DK II is a pedal control interface designed specifically for live music performance. It is intended to allow performers to control many aspects of their sound quickly and efficiently. All switches are momentary type so that, through software programming, the timing of a button press can be significant (if desired). A rack-mounted, custom-built computer interface powers and communicates with the DK II. Thanks to Zach Weaver for help building the steel enclosure.
Features:
• Steel enclosure
• Finest quality Carling momentary switches
• 4 jacks for expression pedals• LED backlights• Phantom powering and data transmission via one 5-pin XLR cable
• Unique shape and switch placement to maximize foot controllability
-
Bassing with Max/MSP (Effect Control Software)
This ongoing work seeks to rethink the live performance possibilities of an electric bass guitar (though it could easily translate to any instrument). Using Max/MSP, and a foot-controlled performance interface of my own creation, visual feedback is provided to the performer while analog and digital processors can be intelligently controlled in a way that is still intuitive. The immense possibilities provided by this software interface for creating new sounds and idioms on a single instrument make it rather unique. Latest software interface:
Each button onscreen corresponds to buttons on a foot-controller. The setup is based around general patches but can be customized on the fly for use in improvisational/compositional settings. In a special “Safe” editing mode, patches can be created, edited, and loaded seamlessly while performing. Any continuous controller can modulate any number of parameters.
-
Pitch Shift 3 (Real Time FFT Pitch Plugin)
I built this plugin with the Pluggo environment as part of an independent study with UM Prof. Jason Corey. Unlike many existing real-time pitch shifters, it does not use any pitch tracking. It is FFT (Fast Fourier Transform) based and can shift any source. It has a massive four-octave range with accuracy within in one cent. The plugin can be used in VST and RTAS environments and can be automated manually or with internal LFO’s.
Audio Examples of the Pitch Shifter
-
Kerrector (Linear Phase Filtering Correction Program)
A few years ago, my father suffered an inexplicable, sudden hearing loss in his left ear. Because he still maintains good hearing in his right ear, the loss does not significantly impact his daily life. However, the experience of listening to music, as one might imagine, is less enjoyable with only one good ear. A recent audiogram showed that he still had some low frequency hearing is his bad (left) ear. I attempted to make a standalone computer application that would intercept audio by played anywhere on the computer, boosting the frequency content in the left channel to exactly compensate for his hearing loss. I designed the linear phase, FIR (finite impulse response) filter in MATLAB and created the standalone application with Max/MSP. Cycling ‘74’s Soundflower application is used to pipe all computer audio to my filter. Using a script, the output settings and levels are automatically configured as the program is launched. The program reports the approximate correction percentage and dB SPL for both ears, given a known pair of headphones. Below are images of the application and the audiogram I was correcting for:
-
Blumlein Panner 2 (Better Panner Plugin)
Although it is currently difficult to accurately present complete 3D audio in a practical way, it is theoretically possible to provide most of the needed information with two audio channels. The localization cues which form the basis for 3D sound spatialization fall into three main categories: Interaural Level Differences (ILD), Interaural Time Differences (ITD), and Effects of the Pinnae.Unfortunately, most audio panners can only provide interaural level difference cues for spatialization. The “Blumlein Panner 2” provides both ILD’s, ITD’s, and phase shifting (based on pan angle), creating much more realistic, spacious panning. The intelligent manipulation of time differences for spatial placement is a natural progression in recording technology, and it’d be pretty cool if mixing desks and recording software all included a similar device.
-
“From Fourier to Everywhere” (Real-time Video Piece)
With a specially programmed joystick, a performer controls images that actually generate frequencies, notes, and sounds pixel by pixel. The software is somewhat of a visual Fourier transform, except that the bands are constrained to simple harmonic relationships. A complex Jitter/Max/MSP patch is at the heart of the piece, allowing the performer to select chords with one hand, while controlling the size, shape, and placement of the images with the other.The blue bar moves in time while activating frequencies that correspond to the distribution of the white objects on the frequency axis. In this exact position, the listener would hear a somewhat lower tone:
Performance of “From Fourier to Everywhere” with Bob Lester at 2007 Scholarship Showcase:
-
Characterizing Dynamic Sensorimotor Cues In Human Listeners (Kresge Hearing Research Institute)
Comments are off for this postPoster from AAS Convention (Scottsdale, AZ) [pdf]
As a partner of the principal lab investigator, I have helped implement a unique human hearing study seeking to characterize how humans localize sound sources with the use of head movements. Many studies have shown the inability of humans to accurately localize narrow-band sources (particularly in front-back discrimination tasks) when the head’s position is fixed. It has been proposed that in the real world, these difficulties are possibly resolved through movements of head, revealing more sonic cues for localization. Preliminary results have been extremely encouraging, showing that head movements (and the degree of those movements) play a significant role in human hearing.Much of my work has been in the designing and programming of the software used to present specific sounds to a subject in an anechoic chamber while their head is moving at a desired speed. Using a head-tracking device, we are able to play a sound while the head is moving through a definable window of size theta (in this case, 2.6°- 20°). The subject is then asked to localize the sound using an automated head-pointing response system. Without going into unnecessary detail, the system is very efficient and allows for many trials in only a few minutes.
Below are some of the preliminary results from many trials from all the subjects in the study. The x-axis is the actual azimuth angle of the sound that was presented, and the y-axis is the subject’s response. For wide-band sources, the subjects are quite accurate, regardless of the window size (theta) in which the sound is played. For narrow band noises (500 Hz – 1000 Hz absolutely filtered noise), front-back reversal errors are abundant for smaller windows, but as the window gets larger, the responses become more accurate. This is a significant result: it shows that humans are able to use head movements to localize sounds that would otherwise be ambiguous in locus.
