Bent nn,

Exercise in Databending

The text collages and concrete poems I made in 2021 are mostly products of processes in which sine functions are defining position and size. What suggests itself is to make one of the results present in an audio format. And one way to achieve this is to do databending.

Excerpt of a concrete poem

Converting & Importing

First, I saved the image in the uncompressed BMP (Bitmap) format. With ImageMagick I converted the BMP to PCM format. PCM stands for Pulse-Code-Modulation and it basically describes the structure of an uncompressed audio file. I chose 8-bit depth.

convert -depth 8 bent_nn.bmp rgb:bent_nn.pcm

The next step is to import this PCM file in an audio editing software. In Audacity: File > Import > Raw Data. Import settings: Encoding [Signed 8-bit PCM], Byte order [Little Endian], Channels [1 Channel (Mono)]

Detailed instructions: Antonio Roberts, Databending using Audacity, 2009

The input picture has twenty lines and they are clearly visible in the waveform of the audio.
Waveform. Screenshot from the program Audacity.
Waveform. Screenshot of the program Audacity


Tempo speedup by 160 %. Some filtering. One ‘tone’ is one line of the poem.

CC BY 4.0 – 02/2021