We’ve updated our Terms of Use to reflect our new entity name and address. You can review the changes here.
We’ve updated our Terms of Use. You can review the changes here.

no shell, just a voice

by teresa pelinski

supported by
bionikka
bionikka thumbnail
lia_mice
lia_mice thumbnail
Ángel Faraldo
Ángel Faraldo thumbnail
acoustic-mirror
acoustic-mirror thumbnail
/

  • Streaming + Download

    Includes unlimited streaming via the free Bandcamp app, plus high-quality download in MP3, FLAC and more.
    Purchasable with gift card

      €3 EUR

     

1.
archivo 00:43
2.
voz 01:30
3.
proceso 04:25

about

I generated the artificial voice in 'no shell, just a voice' using a couple of hours of my dad's field recordings and Wavenet, a neural audio synthesis deep learning model. My dad, Ramón Pelinski, was an ethnomusicologist, and he studied the sword dance at la Todolella, a small village close to Castellón, in eastern Spain. In these tapes, recorded during the 90s, he interviews the old and the young who inhabit la Todolella, and asks them about how their lives have changed over time and how cultural identity is represented through music.

I found these tapes in a wooden mountain of boxes in his office. My dad does not live anymore, and I wonder if these tapes will ever be listened to again. Will someone just throw them away when I die? In the process of digitalising the recordings, I listened todolellanos' voices and a voice I had long not heard, my dad's.

This project is about many things, but mainly about listening. I feel the warmth of my dad's voice, and recognise some of the todolellanos' voices – I met some of them when I was a kid. And I wonder who am I listening to when I listen to the artificial voice. Do voices always point to someone? Can they point to more than one person? I discuss the question of the body in artificial voices considering Mladen Dolar, Michel Chion and Pierre Schaeffer's accounts on the voice in the webzine accompanying this work (noshell.teresapelinski.com).

The first track is an archive (dataset) sample. The second track is the generated voice. Finally, the last track collects the algorithm-generated voices during training in ascending order (i.e., the algorithm learns to speak as the track evolves).

credits

released March 1, 2022

license

tags

about

Phonos Netlabel image

Phonos Netlabel Barcelona, Spain

Initiative to support Phonos' artists in residency and related projects.

Phonos is an electronic music lab in Barcelona, Spain, with a long tradition bridging sound experimentation and research in technology.

contact / help

Contact Phonos Netlabel

Streaming and
Download help

Redeem code

Report this album or account

If you like no shell, just a voice, you may also like:

Bandcamp Daily  your guide to the world of Bandcamp

On Bandcamp Radio

Featured guests are the golden-voiced Baby Rose and psyche soulsters Brainstory.
listen now