Saturday, August 29, 2009

Audio content

We now have six populated content areas (Fiction, History, Law&Order, Medical, Performing Arts, and Self Help) for each paradigm (Disability Consciousness, Medical/Media). Background sound will be provided in each case by repetitive phrases or sounds – in the case of Disability Consciousness, the voices of demonstrators chanting, in Medical/Media, the sound of medical/hospital equipment. The basic premise for the work is encapsulated in a phrase rendered from text by AT&T’s text to speech on line tool:

Which makes it clear that for some, disability is the problem, while for others, the problem is the way disability is viewed and responded to.

There are lots of terrific resources on line. One great series of articles including history and some basics,

Another about the MDA telethon,

A good article articulating points of view on the Clint Eastwood film “Million Dollar Baby” (which we quote from),

And the organization that has helped organize persons with disability and helped articulate their point of view, ADAPT.

Wednesday, August 19, 2009

Optical Sensing of Elevator State

The two optical methods I had in mind for sensing the elevator's state were video analysis of the elevator display (using a web cam pointed at the display) or mounting eight photoresistors in a frame to surround the display. Since video analysis involves less work I decided to try that first. I made a test patch in PD with simulated elevator display images (see below). My first test was to see if "blob tracking" would be able to distinguish the various elevators states. Clearly, blob size alone will not do. The display images with the same number of segments (2 and 3, either up or down) have the exact same blob size. Vertical blob position values for those same images are indentical too. Fortunately there is sufficient difference in horizontal blob positions to distinguish the six floor states. Now we need to record video of the actual display in the elevator to confirm my simulation and note the dynamics. - Drew

Saturday, August 15, 2009

Cab Position and Movement

Since tapping into the elevator control system will not be possible (see August 11 post) we will be using sensors in the cab to detect position and movement. This information will then control the sound files. My initial thought was to use an arduino microcontroller with an accelerometer to sense the cab movement and try to keep track of what floor we are on by counting stops. This will require "homing" the system at, say, floor 1. Here is a data log graph of a test I made using an accelerometer in the elevator of our condo building. Although somewhat noisy, there are clear spikes for starting and stopping. Seems promising but accuracy over time is my concern.

Here's the patch I made for this test using Hans-Christoph Steiner's Pduino object/firmware and Chris McCormick's [s-totalrecall] PD abstraction to record the data.

My other thought for detecting elevator position and movement is to optically read the display inside the cab. A combination of techniques may be necessary. More on that in a future post. - Drew

Concept Development

The content for this installation comes primarily from recordings of interviews, disability rights protests, synthesized speech and media clips. Our intention is to musicalize these sources by breaking them down into rhythm and melody (tempo and pitch). We will be mixing our various sources in real-time with the elevator's state (cab position and movement). Aspects of the audio layers to be controlled will include start/stop of clips, volume and mix. Elements, such as interviews, will be foregrounded while others, such as protest chants, will be backgrounded. Musicalizing voices has been fun and challenging. I've been listening to the early pioneers of electronic music like Charles Dodge which takes be back to my sound art classes in grad school with Bob Snyder. I'm using PD (a real-time graphical programming environment) in my experiments with beat and pitch tracking. - Drew

Tuesday, August 11, 2009

Technical Progress - Physical Installation

Today was a significant juncture in the technical development of the piece. We met with our cultural center point person, Nate, building engineer, Brian and elevator engineer, Mike. Although we had hoped to access the elevator control room and tap into the elevator data (movement and position info) as well as the speaker system, we learned that would not be possible for liability reasons. Our backup plan was to put the equipment on top of the lighting canopy in the elevator cab and use sensors to detect movement and position.

Mike positioned the elevator so we could see exactly what was above the canopy and above the elevator cab. As he climbed on top of the elevator cab it looked like a scene from Die Hard. He showed us the hatch doors where we could run power down into the cab. I really hope I never have to be hauled out of a stuck elevator - there are many obstacles - not sure how it could be done.

Unfortunately there was very little room above the canopy inside the elevator so Mike suggested installing the equipment above the cab. The problem with that is we would only be able to access the equipment with his help and we need unlimited access for development and troubleshooting. However, Mike was very helpful in finding a solution within the cab. He showed us a small but adequate space behind a wall panel. Perfect! We will have an accessible and secure location for the installation.