Artist Statement
With the advent of the computer in the 20th century, there has been a lengthy exploration of the relationship between graphic and audible arts. However, this exploration has been relatively one-sided, with the general focus being on the creation of graphic art from music. Especially guilty of this discrimination is the sub-genre of generative art. The most widely observed form of generative art, birthed from music, may be seen in the spectrum analyzers built into programs such as Winamp and iTunes. While I do not argue with idea or practice behind this form of generative art (I have participated in this discrimination myself), I do think it is time we as generative artists explore the other pole of this relationship.
Audizer (pronounced aw-di-zer) is an "Image to Sound Translation" tool. It is a generative art piece created to explore the relationship between the graphic arts and music. As an interactive piece Audizer is limited – and it’s meant to be. The intention of the project is less to be an audio tool for musicians (or happenstance musicians) and more of a generative art piece that is able to stand on its own. As such, the user is asked only to choose an image to listen to. The user may either put in a link to an image (an image already on the web) or they may use one of the famous works of art listed on the startup screen.
Once Audizer is fed an image, it will analyze the data and create a song based on that image. The image is evaluated on a number of different criteria: The lightness or darkness of the image, the variation of color throughout the image, and the individual red, green and blue values of each pixel in the image. After the image has been evaluated, Audizer will pick a key based on the darkness of the image and the number of octaves based on the variations of color. The song will then begin to play. Initially, with no user input, Audizer will read the red pixel values and output treble notes, the green pixel values and output bass notes, and the blue pixel values to decide how many notes of each bass and treble to play in any given measure (this is the “other data”).
The user may decide to change which color value determines any note at any given time simply by choosing R, G, or B next to Treble, Bass or Other Selection Data. Audizer also allows the user the option of a different treble or bass voice as well. The voices by be chosen by selecting 0, 1 or 2 next to Treble or Bass Voice. The 0 voice is the default, 1 is an oldschool eight-bit sounding voice, and 2 is a wobble or vibrato.
Technically, Audizer uses Flash’s newfound ability to generate audio on the fly. Audio generation inside Flash is still in its infancy and can be a little glitchy. If the project is run in an installation environment, as it would be in a gallery, these hiccups are generally avoided. As it is currently, online and in the browser, you may notice a few pops in the audio.
I would like to thank Andrew Twomey at SCAD for asking the question that birthed this project: “…I wonder what it would sound like if something like the Mono Lisa could be converted into lines and fed into the sculpture?”
Enjoy,
Philip Likens
© Copyright 2010 Philip Likens. All Rights Reserved.