Created by Leandro Garber and Tomás Ciccola in 2019, the Analysis and Sensitive Representation of Data research team focuses on the development of software tools for the study, experimentation and generation of images and sounds using artificial intelligence techniques.

Director: Leandro Garber (UNTREF)

Research Students: Luca Belloti (UNTREF), Florencia Pasquali (UNTREF), Ramiro Arsanto (UNTREF), Dai Perséfone Miauro (UNTREF) y Santiago Fernandez (UNTREF).

Graduate thesis students: Tomás Ciccola (UNTREF), Máximo Signorini (UNTREF) y Luis Zunino (UNTREF).

Assistants: Agustín Spinetto, Juan Cruz Amusategui y Tatiana Cuoco.

AudioStellar is an experimental AI-powered sampler. It is free and open source. It is a musical instrument that allows you to explore a collection of thousands of sounds and play them in creative ways by interacting with the map it generates from the corpus of sound data that users can enter.

It offers the possibility to explore and play back audio samples in innovative ways: parameterizable autonomous agents, rhythmic constellations, free trajectories in space, or custom heuristics using MIDI and OSC.

 

 

Director: Leandro Garber (UNTREF)

Research Students: Luca Belloti (UNTREF), Florencia Pasquali (UNTREF), Ramiro Arsanto (UNTREF), Dai Perséfone Miauro (UNTREF) y Santiago Fernandez (UNTREF).

Graduate thesis students: Tomás Ciccola (UNTREF), Máximo Signorini (UNTREF) y Luis Zunino (UNTREF).

Assistants: Agustín Spinetto, Juan Cruz Amusategui y Tatiana Cuoco.

 

The project focuses on the development of software tools for the study, experimentation and generation of images and sounds using artificial intelligence techniques.

AI Music Creativity 2022 (AIMC2022) in Tokyo. Audio-visual piece by Agustín Spinetto and Andrés Duarte Loza using AudioStellar.

September 14th, 2022

For this work, Andrés Duarte Loza created a large set of audio samples using the Japanese musical instrument Koto. Agustín Spinetto assembled these samples with Japanese percussion and used AudioStellar to create a performance in which the various melodies looped by the Koto are supported by AI. They are also modified by an autonomous generative process that gradually replaces the loops with other samples, which the software assumes to share some sonic characteristics.

---

AudioStellar stand at “DIGI LAB: local brands of Technology and Instruments” within the framework of MUTEK.AR 2022.

September 9th, 2022

MUTEK is an international festival of electronic music and audiovisual arts held in different countries throughout the year. At the 2022 edition of MUTEK.AR, Buenos Aires, the AudioStellar project team was invited to present their software in the Artlab space. Two datasets were showcased that visitors could play using different MIDI interfaces, including the first customised interface resulting from a collaboration with Yaeltex, ESTELA.

---

Sonic Constellations: experimental luthiery and artificial intelligence workshop at Fundación OSDE

July 2022

Leandro Garber and Julio Nusdeo gave a workshop that sought to trace a common ground between the musical scene of sound artists – who use as instruments objects they intervened or created themselves – and the search for new sonic strategies through the use of the so-called artificial intelligence techniques.

To this end, the workshop hinged on the production of analogue sound objects that communicate with AudioStellar, thus establishing exchanges between digital and analogue situations. This real-time feedback made it possible to design a sound ecosystem with elements that start from dissimilar universes and mutually condition and expand.

---

AudioStellar at the BILIA – Latin American Biennial of Artificial Intelligence in Mexico City.

April 5th to10th, 2022 

The members of the AudioStellar project were invited to the first edition of the BILIA by Julieta Agriano, founder of the platform “WIP Arte Digital”, Latin American Biennial of Artificial Intelligence. At this event, they conducted a hackathon for three consecutive days. The attendees, ranging from musicians and performers to economics students, were excited to use the creative possibilities of AudioStellar.

During those days basic software concepts were taught and the visitors met in interdisciplinary groups, discussed ideas and built different prototypes: a motion sensor where performers can play AudioStellar by means of body gestures, a sonification of GDP data from different countries and a connection between Arduino and AudioStellar to enable the use of proximity sensors.

There was also a concert where two pieces were performed with AudioStellar as the main instrument with various musicians from the Live Coding scene in Mexico City sharing the stage.

---

Installation of AudioStellar in the Agora Cycle

April 9th, 2022

The members of the AudioStellar project were invited by the Ágora cycle to be part of its “Human Gen” edition in the Estudio 1+1 space. On this occasion they set up an interactive installation in a booth where users were invited to play an AudioStellar session using a MIDI controller and body movements tracked with a webcam fitted with an artificial intelligence algorithm.

 

---

AudioStellar in Vertigo. Residency at the Centro de Arte Sonoro

October 2021

“AudioStellar in Vertigo” is a sound collaboration between Leandro Garber and Julio Nusdeo. It is a blend of AudioStellar and free improvisation and feedback with instruments made by Nusdeo. This ensemble takes place between digital and real-world situations, inter-modulated and intervened by springs, strings. and waste material fed back by the actions of AudioStellar.

The residency included a talk, an open radio rehearsal and ended with an 8-channel participatory sound installation in the Patio of the Casa del Bicentenario. The participants were invited to interact with the different instruments created by Nusdeo, which also activate the AudioStellar sound environment.

---

AudioStellar at NIME 2019

June 2019

Leandro Garber and Tomás Ciccola traveled to Brazil to present a demo version of AudioStellar. The International Conference on New Interfaces for Musical Expression (NIME) were held at the Federal University of Rio Grande do Sol between June 3 and 6.

---

AudioStellar at the Tokyo University of the Arts

June 2019

AudioStellar project was presented at the workshop on Artificial Intelligence applied to Music Composition held at the Tokyo University of the Arts. Agustín Spinetto gave a talk about AudioStellar sampler-like software instrument and how it uses AI to generate a 2D intelligent audio map to organize audio samples by their audio spectrally characteristics.