This session consists of 4 presentations and a joint Q&A with the presenters. The session contains:
➺ Léa Morin, Deva Pereda, Mariana Torres, Paul Bonnarme, Laura Alhach Castro - CinémArabe, Asia, Africa, Latin America (1975-1979): Material Research and Circulation Approach for the Preservation of “C3 Non-Aligned Film Archives” (Short presentation)
➺ Juan Antonio Casado - From echoes to ecosystems: Transforming sound recordings into knowledge through Artificial Intelligence (Long presentation)
➺ L'Armari de la Memòria, Pablo Hernández Miñano, Bruno Laurent, Inés Ferrando Vidal - La Pinteta Rebel: Queer waves for historical memory. A case of radio archiving in an LGTBQ+ archive (Short presentation)
➺ Filippo Mengoni - Like Never Before: Leveraging AI to Promote Access to the Original Oral Sources of the Cineteca di Bologna (Short presentation)
**Abstracts:**
➺ CinémArabe, Asia, Africa, Latin America (1975-1979): Material Research and Circulation Approach for the Preservation of “C3 Non-Aligned Film Archives”
--
Léa Morin, Deva Pereda, Mariana Torres, Paul Bonnarme, Laura Alhach Castro (Short presentation)
--
In the search for the traces, stories and archives of the Third Cinema Movement, we evidenced a disconcerting lack of resources in official archives. Nonetheless, in face of this unexplored theme, an abundance of documentation, publications and even films, almost totally unpreserved, appeared in private and transnational archives. An immense material memory in danger of disappearing.
CinémArabe magazine is one of these almost unknown traces of the existence of a Transnational Third Film Critic Movement. Dedicated mainly to Arab cinema, also to Asian, African and Latin American, CinémArabe was created in Paris in 1975 by a group of young critics and filmmakers, fighting for the existence of a film critic committed to the cinema of their countries.
From 1975 to 1979, it proposed unpublished interviews with African filmmakers, covered tricontinental festivals, and was also the space for the publication of manifestos translated into French and Arabic.
With new interviews conducted with the magazine contributors such as Magda Wassef, Ali Akika, Abdoul War, Ahmed El Maanouni and Khemais Kheyati, and engaged in a collective work methodology for studying close-ups and multiple micro-stories published in the magazine, we tried to understand the context of the existence and disappearing of the magazine, how was it circulating, and what can film criticism do in a context of political and aesthetic domination? This, with the aim of giving new light to non-central narratives of militant gestures and film critics from the 70s and pondering on the echoes in present day political and aesthetical proposals.
➺ From echoes to ecosystems: Transforming sound recordings into knowledge through Artificial Intelligence
--
Juan Antonio Casado (Long presentation)
--
The project, submitted to the National Library of Spain, is an initiative to contextualize the wealth of information compiled in more than 200 Subterfuge Radio podcasts titled "Sympathy for the Music Industry". These podcasts feature interviews with key figures from the Spanish music industry scene, aiming to share their insights, experiences, and outlooks on the recording industry.
Interviewees dissect their roles within this professional landscape, essentially crafting a comprehensive encyclopedia filled with diverse data, including previously undisclosed information. This proves valuable beyond the confines of the interview format. The collection encompasses individuals, events, dates, companies, and statistics, accumulated over more than 200 interviews, each lasting around an hour, that the program has released so far.
This project originates from these audio sources. After converting them into text and employing graph theory alongside ontological analysis principles, it compiles all discussed data into a semantic model. Once processed and interconnected, the narrative's components become navigable on a web platform through interactive visualizations. These visualizations span chronological, geographical, documentary aspects, and the relationships among the elements discussed.
Artificial intelligence plays a crucial role in the project's two primary phases: the conversion of speech to text, utilizing Natural Language Processing (NLP) techniques, and the recognition and classification of entities.
➺ La Pinteta Rebel: Queer waves for historical memory. A case of radio archiving in an LGTBQ+ archive
--
L'Armari de la Memòria, Pablo Hernández Miñano, Bruno Laurent, Inés Ferrando Vidal (Short presentation)
--
Among the most interesting audiovisual and sound archive documents currently kept in l’Armari de la Memòria, there is a set of audios from the radio program La Pinteta Rebel, broadcast on Radio Klara (Valencia, 1982-present) between 1984 and 1994. La Pinteta Rebel was a pioneer radio program in Spain for its content: it included news, interviews and fiction dedicated to the LGTBIQ+ community, so invisibilized and repressed in the 80s. The protagonist voices, were the activist Miquel Alamar and the members of the music-hall group Ploma-2, Rampova and Clara Bowie.
In addition to being a space for transgression, discussion and critical look at the current political and social norms in the last years of the Spanish Transición and entry into democracy, La Pinteta Rebel appears in a year when the gay liberation movement had been dismantled in the city of Valencia, assuming an important role in the struggle and the claim for the rights of LGTBIQ+ people.
In this presentation, the professional team of l’Armari de la Memòria will explain the process of recovery of these sound archives and their treatment as a documentary source to reconstruct a collective memory. It will address the problems that have arisen in this process and what challenges remain to be faced in the reconstruction and dissemination of these documents for their historical use and promotion.
➺ Like Never Before: Leveraging AI to Promote Access to the Original Oral Sources of the Cineteca di Bologna
--
Filippo Mengoni (Short presentation)
--
Despite being one of the earliest methods of historiography, oral sources have continuously struggled to assert their value. Thanks to technological and informatic advancements, we initially moved to recording sound on physical media and subsequently dematerialized it into a vast series of bits. These progressions have allowed us not to settle for transcriptions and to use timed indexes that assist users in searching for information, overcoming one of the major inconveniences that has hindered the widespread dissemination of these sources outside this research area.
Despite these significant changes, interviews still struggle to find its place in historiographical research and in archival management as sources able to reproduce the sonic nature of spoken language, rather than written text. The reason is due to the considerable amount of resources needed to create timed indexes for a multitude of available sources.
Through the use of Automatic Speech Recognition and Automated Indexing systems, AI software will be utilized as a means to promote access to more than fifteen hundred oral sources primarily involving well-known figures and unpublished recordings at the Cineteca di Bologna. For these reasons, the project represents a unique challenge and atypical features: both for an archive specialized in preserving film heritage and for oral history in general.
After physical media and digital formats, we are facing another important, perhaps decisive, step: assessing the implications of using AI to finally allow the voices confined in our archives to emerge and be heard by all.