Recently in Pixel-Based Music Category

Progress and Patterns

Pixel-based musical notation. That's what I'm still experimenting with.
Formerly I was taking a photographic narrative approach and creating a musical score based on a series of images. This was okay, but it turns out people are actually more interested in how changing the graphic changes the resulting audio. It's also interesting to see what type of image (photograph, in this case) creates the most appealing or dynamic range of sounds.

Through breaking down many different photographs into pixel based musical notation, I've found the most exciting audio comes from images that have a varying, somewhat vertical pattern. Diagonally balanced patterns are also very nice.

EXAMPLES OF IMAGES THAT AREN'T SO AWESOME IN AUDIO FORM:

(As a reminder, the audio notation reads from left to right. The pitch is determined by the height of a pixel, and the brighter the pixel, the louder the audio value.)

I will edit these examples to play in the same key for comparison's sake.

A graphic of simple horizontal lines has no rhythm and no variations in melody:
horiz_lines.bmp

horizontal_lines.mp3

A graphic of vertical lines has an obvious rhythm but also has no movement of melody:
vert_lines.bmp

vertical_lines.mp3

So there is definite motivation to use images that actually will make for interesting audio.

For my final end product this semester, I will still be making a video with the images and their corresponding sounds, but instead of using narrative photographs, I have chosen to just photograph things I see within walking distance of my house that I think will sound the best. I am mostly looking for interesting patterns that transform both horizontally and vertically.

Here is one example:
coiledwire_small.jpg

coiledwire2_small.jpg

coiledwire.mp3

Depending on how I change this bitmap image, the sound also changes. To show this relationship, I will also include different variations for each photograph (which will be included in the video).

Compare the following graphic and audio to the previous one. Notice how the more elongated tones in the second example match the longer horizontal groups of pixels.

coiledwire2bsmall.jpg

coiledwire2.mp3

I will also include all the edited graphics of the notation in my final body of work.

Pixel-Based Music Project

I've decided to do some work exploring the relationship between aural and visual language. I have been experimenting with different ways of making sounds based on bitmap images (playing a picture like a musical score).

Through a software program I've been able to open a bitmap file of a photo, where it's then broken down into fewer pixels. The musical "score" is based on these pixels, and played from left to right. The height of each pixel determines the pitch, the color determines where the note will be played on the stereo field (left or right speaker), and the brightness determines the volume.

Within these guidelines, the pitches can be redefined to fit a particular musical scale. Removing specific horizontal lines of pixels accomplishes this by eliminating all notes that do not fit within that scale. Further manipulation of the image results in multiple variations of sounds.

Here are a couple examples, showing the original photograph, the pixelized png (stretched out to match the length of the audio file), the further edited png, and the resulting MP3 file (looped).

Rug

rug1.jpg

rug2.jpg

rug3.jpg

Grandma

gma1.jpg

gma2.jpg

gma3.jpg

Using this method, I plan on composing a longer-length (2-3 minute) song, based on a series of original photographs that will tell a narrative on their own. After this process is complete, I will create a movie file that scrolls through the line of photographs as the resulting audio file plays along in real time, showing both ways of communicating the same story via distinctly different processes.

About this Archive

This page is an archive of recent entries in the Pixel-Based Music category.

Digital Focus is the previous category.

Find recent content on the main index or look in the archives to find all content.