The Upscaling of the Early Cinema Image through Artificial Intelligence. A New Aesthetics between Continuity and Dissent in Image Theory

1 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 The Upscaling of the Early Cinema Image through Artificial Intelligence. A New Aesthetics between Continuity and Dissent in Image Theory1 Luis Navarrete-Cardero 2; Carlos Ramírez-Moreno 3 DOI: https://dx.doi.org/10.5209/aris.90455 Recibido: 11 de julio de 2023 / Aceptado: 27 de octubre de 2023 / Avance en Línea: 30 de octubre de 2023 Abstract. An active community of users is manipulating the image of early cinema using Machine Learning-based software. This practice, encompassed in the field of Artificial Intelligence, has transformed the materiality of these images, generating a mutated image as a result. Through a qualitative research design, based on a hermeneutic approach, this research aims to delimit the nature of this new image by observing it from the theoretical parameters that have shaped the specificity of the original filmic image. Starting from three historical concerns of film image theory, we have studied the interferences that this mutated image produces in the field of aesthetics, in its signifying relationship with reality and in its own ontology. Surprisingly, the new mutation presents a meagre break with the theoretical tradition of the cinematic image, oscillating between a slight dissidence and a marked continuity of its postulates. Keywords: early cinema image; image aesthetics; image ontology; artificial intelligence; machine learning. (es) La mejora cualitativa de la imagen del cine primitivo a través de inteligencia artificial. Una nueva estética entre la continuidad y la disidencia en la teoría de la imagen Resumen. La imagen del cine primitivo está siendo manipulada por una activa comunidad de usuarios a través de Machine Learning. Esta práctica, englobada en el campo de la Inteligencia Artificial, ha transformado la materialidad de estas imágenes, generando como resultado una imagen mutante. A través de un diseño de investigación cualitativo, basado en un enfoque hermenéutico, esta investigación pretende delimitar la naturaleza de esta nueva imagen observándola desde los parámetros teóricos que han configurado la especificidad de la imagen fílmica clásica. Partiendo de tres preocupaciones históricas de la teoría de la imagen cinematográfica, hemos estudiado las interferencias que esta imagen mutada produce en el campo de la estética, en su relación significante con la realidad y en su propia ontología. Sorprendentemente, la nueva mutación presenta una exigua ruptura con la tradición teórica de la imagen fílmica, oscilando entre una leve disidencia y una marcada continuidad de sus postulados. Palabras claves: imagen del cine primitivo; estética de la imagen; ontología de la imagen; inteligencia artificial; machine learning. Sumario: 1. Introduction. 2. A New Materiality Prefigured from a Double Convergence. 3. Methods and Instruments. 4. The Value of the Aesthetic Experience. The Ages of Looking. 5. The Total Image and its Historical Relation to the Psychic Drive of Realism. 6. The Disappearance of the Subject and the Visibility of the Non-Existent. 7. Conclusions. Cómo citar: Navarrete-Cardero, L., Ramírez-Moreno, C. (2024). The Upscaling of the Early Cinema Image through Artificial Intelligence. A New Aesthetics between Continuity and Dissent in Image Theory, Arte, Individuo y Sociedad. Publicación en línea, 1-17. https://dx.doi.org/10.5209/aris.90455 1. Introduction. The early cinema image (Burch, 1990) is being manipulated by an active community of users through the application of Machine Learning tools. This practice, included in the field of Artificial Intelligence, has changed the materiality of these images —color, format, resolution, and number of frames—, and 1 Financiación Este artículo es un resultado del proyecto de investigación financiado por la Consejería de Universidad, Investigación e Innovación de la Junta de Andalucía, Una manera de Hacer Europa / Proyecto (B-SEJ-370-UGR20), “La industria cinematográfica en el hipersector audiovisual-TIC (CINATIC)”, dirigido por los investigadores Francisco Javier Gómez Pérez and Jordi Alberich Pascual. 2Universidad de Sevilla E-mail: lnavarre[email protected] https://orcid.org/0000-0002-7156-9144 3Universidad Loyola de Sevilla E-mail: [email protected] https://orcid.org/0000-0003-1048-6326 2 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 has favored the appearance of a participatory disposition in an image historically associated to a contemplative regime. Machine Learning’s praxis induces the manipulation of images through the operational interaction by which the program empowers the user through its interface. This tautology of interactivity (Manovich, 2001), typical of a computer-based mediation, allows the transformation of the cinematographic image into the properties indicated. Figure 1. Arrivée d’un train à la Ciotat. Directed by Frères Lumière, 1897. Aspect ratio: 4:3 (1.3:1). Source and License: YouTube, CC BY. Figure 2. Arrivée d’un train à la Ciotat. Directed by Frères Lumière, 1897. Aspect ratio: 16:9 (1.78:1). Resolution: 3840 x 2160 (4K). Source and License: YouTube, CC BY. Figure 3. Arrivée d’un train à la Ciotat. Directed by Frères Lumière, 1897. Aspect ratio: 16:9 (1.78:1). Resolution: 3840 x 2160 (4K). Color: Colorized. Source and License: YouTube, CC BY. 3 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 This transition towards a participatory capacity is due to the extensive power of software in its interference in the field of media (Manovich, 2013). In this way, Machine Learning empowers us to operate a profound intervention on the early cinema image turning it into a new cultural object: a mutated image. This new image produces interferences (1) in the field of aesthetics (Renoir, Debray), (2) in its signifying relation with reality (Burch, Morin, Bazin), and (3) in its own ontological character (Virilio, Mitchell). In the following paper, we try to weigh the gravity or futility of these observations from the field of image theory. When we speak of Artificial Intelligence’s applicability to the early cinema image, we do so in a loose sense. Strictly speaking, we are talking about a computational framework that has managed to assimilate the nature of this image thanks to the previous swallowing of a huge amount of them. This accumulated experience in the form of ingestion subsequently allows the composition of new images. This process is not governed by abductive principles, it is not based on unknown environments (Eco, 1983), but on a preceding contextual work that, inexorably, distances it from the usual meaning of understanding. The task of the software is to find a mathematical formula that, when applied to a collection of inputs called ‘training data’, produces the desired outputs. This supervised learning formula is susceptible to generate correct outputs for other types of inputs, other than those used in the training data process, ‘on the condition that those inputs come from the same or a similar statistical distribution as the one the training data was drawn from’ (Burkov 2019, p. 3; Cf. Eco, 1983, p. 281). Figure 4. Graphical Representation of the Convolutional Neural Network (CNN) Process for Image Upscaling through Machine Learning. Source: own elaboration. If the application of Machine Learning to the early cinema image cannot be considered an intelligent procedure, neither can we define it as an act of learning. If the input is slightly distorted, it is very likely that the output returned will be incorrect. As we know, learning in animals is much more flexible. If a person learns to play a video game sitting in front of a screen, he will be able to continue playing if we turn the screen slightly. However, if a Machine Learning algorithm has been trained to play in the identical position, it will not be able to continue if we rotate the image, unless it has also been trained to recognize the new location of the output device (Eco, 1983; Burkov, 2019). Before continuing, it should be noted that in this work there is a latent opposition between the concepts of Upscaling, a term we apply to Machine Learning work with images, and Film Restoration, a process that starts from a given work to improve its reception in the interest of recovering the original state of the film at the time of its filming (Peters, 2014). In the latter case, the materiality of the early cinema image, which as we will see has been ideologically assumed by image theory as an improvable imperfection, has become one of the resources of humanity to understand the development of the social and historical foundations of the twentieth century (Jones, 2012). Is it convenient to permute the constitution of these images, or should we preserve them as intact vestiges of our historical, aesthetic, technological, and social footprint? Obviously, the 4 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 purpose of this paper is not to invalidate the practice of Machine Learning, whose application has proven successful in various fields, from driverless driving or language translation to coronavirus diagnosis through images (Narin, Kaya and Pamuk, 2021). However, its intervention in the field of the early cinema image may cause problems conceptually close to the paradox of the time machine. If the materiality of the film encloses a certain conception of the epoch and, consequently, of our past, altering it means a rupture with our present because ‘history should not be reconstructed solely on paper’ (Rosenstone 1995, p. 20). Figure 5. Showing the Three Essential Layers of the Convolutional Neural Network (CNN) process. Source: own elaboration. 2. A New Materiality Prefigured from a Double Convergence. The emergence, capacity and operability of Machine Learning software can be traced in the genealogy of media. The change in the materiality of the digital image is described by Manovich in The Language of the New Media (2001) through his famous pentagram. These five characteristics remain valid twenty years after their establishment and are still useful to describe the processes by which two successive convergences, digital in the first instance and via software in the second, have changed the face of media. The first pair of these principles, numerical representation and modularity, deals with problems concerning the specificity of new media; the next pair, automation and variability, affected by the previous binomial, is related to more far-reaching operational issues connected with authorial intentionality and interface ductility. But it is undoubtedly the fifth element, the principle of cultural transcoding, the most important for this research (Manovich, 2001, p. 92). According to Manovich, cultural objects produced by the new media show a double facet where before there was only one (p. 92-95). Thus, 1) all new media objects can interact with culture, as was already the case with those created by the old media, but, in addition, 2) they are able to establish a particular dialogue between them. This dialogue is impossible in traditional media because it is determined by their exclusive informatics layer. For example, a video file shares with an image file a certain data structure and compression method that can be extended to an audio file. This new ability, at once a statement of uniformity exacerbated by the hegemony of software, is changing the relationship of objects to culture, affecting it as never before in the analog era. Manovich’s principles are encompassed in the first convergence of new media, caused by the application of digital processes to the production of cultural objects. But it is also possible to recognize in them the emergence of the second convergence, brought about by the totalitarian irruption of software in the field of media (Manovich, 2013). Machine Learning can be understood as a result of both. 5 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 The first convergence was foreshadowed by McLuhan and R. Powers in the field of communication theory and prefigured by Deleuze and Guattari from the field of philosophy. Influenced by the exuberant terminology of L’anti-Oedipe (1972), McLuhan drew in The Global Village (1989) a technological landscape split into two realities. The Deleuzean concept of line was renamed as visual space, whereas the rhizomatic model was renamed as acoustic space. While visual space possesses qualities such as hierarchy, order, and sequentiality, acoustic space, which is much more fluid, is characterized by decentralization, holism, and juxtaposition. Both spaces are only figures for the intellection of the distinction between the technologies of the past and the technologies of the future. Is not this double spatiality a metaphor for the transition to digitalization that occurred in the nineties of the last century? Digital has transmuted the watertight compartments of the old media into fluid space. The technology of pre-digital video and the technology of analog photography maintained serious ontological connections as they both shared their capacity to mechanically record reality; however, they both inhabited independent technological dimensions. Today we contemplate the spaces of video and photography as an undifferentiated continuum whose boundaries have been erased by the digital. In the same work, the terms enhancement and obsolescence, one of the two pairs of figure and ground of McLuhan’s famous tetrad, serve to explain the phases of a technological artifact in culture. When a new technology becomes a figure, occupying a relevant position in our lives, another becomes a blur in the background. Possibly, this conversion of the early cinema image into a mutated image will come to occupy a prominent position in the study of primitive cinema. Will this be the case in the immediate future? For McLuhan, this is the traditional functioning of any cultural object. However, to the apocalyptics’ relief, the process pointed out by the media theorist is completed thanks to another figure and ground binomial, composed by the instances of retrieval and reversal. Every technology promotes the recovery of a previous forgotten factor or process when it reaches a high level of saturation, showing an inverse behavior for which it was created. At present, digital cinema has not yet reached this stage of recovery, although the so-called material turn is affecting the current discourse of film heritage. This movement pursues a renewed yearning for the purity of the material experience of cinema, a return to film at the expense of the numerical format; currently its traces can be found in the work of filmmakers and artists alike, ‘including Peter Delpeut, Gustav Deutsch, Bill Morrison, Tacita Dean, and most recently Hollywood filmmakers such as Christopher Nolan, Paul Thomas Anderson and Quentin Tarantino’ (Fossati 2018, p. 18). The second convergence, initiated with the software revolution, has ended up imposing modus operandi that further standardize media differences. This new mass media reality has brought about an operational unification due to the impregnation of software logic in the constructive processes of cultural objects, thus affecting our understanding of culture. Therefore, ‘Are there some structural features which motion graphics, graphic designs, websites, product designs, buildings, and video games share since they are all designed with software?’ (Manovich, 2013, p. 4). This question tries to show how interfaces are modifying our worldview in a direction of no return towards the simulation of a reality that is born in the execution of the computing machine itself. The interactive dialectic of software imposes itself in the era of the technical image (Flusser, 2017) making the user participate, manipulate, reconstruct, alter and modify different cultural objects (Lammes, 2017), and generating new dispositions in prosumers that entail profound cultural changes. 3. Methods and Instruments. In this intersection of software with culture, of Machine Learning with the early cinema image, certain questions are foreshadowed that have only been sketched so far. The main question to which this research aims to answer can be stated as follows: What levels of dissidence and continuity does this transformation of the materiality of the early cinema image show with the classical formulations postulated by the theory of the image? 6 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 In this paper we will undertake the analysis of certain configurative aspects of the early cinema image —called instruments— in order to verify their resilience in the mutated image. Despite the intrusion of software in the result of the new image, is the latter still indebted to the ideological principles that helped theorize a certain specificity of the early cinema image? This last question leads us to the definition of the methodology and the instruments chosen for this work. The answer to these questions must be sought through a methodology based on a qualitative design. Qualitative research aims to approach the world out there in order to understand, describe, and sometimes explain social phenomena by analyzing their interactions as they occur. This procedure is based on the observation of the phenomenon and ‘on the analysis of appropriate material, texts, images, films and other similar traces, product of these interactions and of the experiences of individuals in society’ (Flick, 2015, p. 12). The qualitative method is not defined in negative terms —qualitative research is not quantitative and not standardized— but is characterized by “the use of the text as empirical material, starts from the notion of the social construction of the realities under study, and is interested in the everyday practices and knowledge that refer to the studied issue” (Flick, 2015, p. 20). In this sense, qualitative research studies things in their natural environment, tries to make sense of the different phenomena, and interpret them from the point of view of the meanings that people give to them (Denzin & Lincoln, 2005). Several purely qualitative research perspectives can be defined, all with different theoretical positions, data collection methods and interpretation approaches (Flick, 2015). This research has started from a hermeneutic analysis of the underlying structures of the object of study, chosen from a triad of possible approaches (Table 1). The following is a description of its elements shaded in color. The qualitative method chosen consists of three elements. First, the theoretical perspective that allows us to approach two different stages of the image -image and softwarized imageis called Genetic structuralism. By simple image we mean images that have not been generated by the intervention of digital technology or thanks to the power of software. In order to analyze these two types of images, technologically different, we must accept that in both of them, despite the intrusion of this technological component, the same ideology underlies, that which allows the researcher to jump from one to the other, knowing that in both cases, and despite their differences, he is dealing with the same cultural object. Secondly, this qualitative method requires certain materials on which the researcher must operate. These materials, which we have called interactions, are the enhanced images that we have analyzed for the research. Their name, interactions, reveals that they are the result of human agency, either directly or indirectly. Finally, thirdly, we have chosen a hermeneutic, i.e. interpretative, approach. To counteract the subjectivity of the researcher, something very common in any interpretative process, we have proposed certain instruments with the desire to limit it. These instruments are themes recurrently investigated by the authors on whom we have relied to make this research, a selection of themes that has helped us to erect the limits of the interpretation. The selected qualitative method is now described in more detail with the help of tables showing the interactions and the instruments. Table 1. Qualitative approach to work. Focus on subjective points of view Descriptive approach to social situations Hermeneutic analytical approach to underlying structures Theoretical positions. Symbolic interactionism. Ethnomethodology. Constructionism. Genetic structuralism. Methods of data collection. Semi-structured interviews. Narrative interviews. Discussion groups. Ethnography. Participant observation. Interaction log. Collection of documents. Interaction log. Photographs. Films. 7 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 Interpretation methods. Theoretical coding. Content and narrative analysis. Hermeneutic methods. Conversation, discourse and documentary analysis. In-depth hermeneutics. Source: own elaboration. Genetic structuralism is understood as a theoretical position where the critic adopts a historical approach to the analysis of the structure of a phenomenon in the belief that there is a homologation with the collective mental structure of the epoch in which it is manifested (Goldmann, 1980). In other words, we can observe the same categories in the description of the early cinema image and its mutation because elements belonging to the same epoch underlie in both. As stated before, from this perspective we will try to verify if the appearance of the software has also changed, beyond the evident material transformation, the historical and genetic ideas that have helped to define the early cinema image in our culture or, on the contrary, if its mutation into a new image is indebted to its same regulating principles and, therefore, to a projection of the immutable ideology of a certain historical time, in the sense expressed by Goldmann. On the other hand, the interactions section of the research is made up of the different mutations of films from the early cinema period that have seen their materiality altered through software. Table 2. Interactions. Film Enhanced image Repas de bébé. Directed by Frères Lumière, 1895. Operator: Louis Lumière*. Vue nº 88. (Channel Nineteenth century videos. Back to life, 2021, 3m16s). (Cf. References for link to Youtube). Rue Tverskaïa. Directed by Frères Lumière, 1896. Operator: Charles Moisson*. Vue nº 307. (Channel Denis Shiryaev, 2020, 2m10s). (Cf. References for link to Youtube). Arrivée d’un train à La Ciotat. Directed by Frères Lumière, 1897. Operator: Louis Lumière*. Vue nº 653. (Channel Deoldify videos, 2020, 48s). (Cf. References for link to Youtube). Bataille de neige. Directed by Frères Lumière, 1897. Operator: unknown*. Vue nº 101. (Channel Nineteenth century videos. Back to life, 2021, 52s). (Cf. References for link to Youtube). Source: own elaboration. *Data: Catalogue-Lumière.com Finally, the hermeneutic perspective of the method (Ricoeur, 1969; Thompson, 1990) is justified by the place from which we analyze the early cinema image through the selection of certain recurrent themes of image theory. Indeed, depth hermeneutic, an articulation of Ricoeur that Thompson modifies motu proprio ‘with the aim of studying the ideology of mass communication’ (p. 405), seeks to understand cultural objects —in our case, the early cinema image and its mutated reverse— from the construction of its meaning, marked by the historical and social conditions of its production. Let us not forget that the theory of the image, like any other theoretical process, can be considered a sediment, and at the same time a driving force, of the historical and social meaning of images. In a way, what Thompson proposes is the objectification of interpretation, clearly an oxymoron, through the establishment of a methodology erected ex profeso to approach the object of study. This new method is materialized in the choice of a certain number of tools, a series of instruments chosen in a sampling work (Flick, 2015). Thus, we have decided to work with certain recurrent themes of image theory and to disregard others according to their suitability to the interests of the study. Let us not forget that all interpretation is a constructive operation of the interpreter who selects and reconstructs by imputing a meaning among many other possible meanings. All hermeneutics is an open process that does not aspire to total mediation, in the sense of Gadamer (2004), because it can never be closed (Ricoeur, 1969). 8 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 As we shall see, these instruments are historical objectivations undertaken on the image by one or more scholars. Their concreteness responds to the verbalization of the commonplaces of image theory that best fit the objectives of the research. Generally speaking, the triad corresponds to the problems derived from the user’s interventionism as was seen in the introduction to this article: (1) interference in the field of aesthetics, (2) significant relationship with reality, and (3) ontological problems of the image. Table 3. Instruments for a hermeneutic analysis. Instrument Image theory topics Essential authors (1) Aesthetics. (1) The value of the aesthetic experience. The ages of looking. Renoir, Debray. (2) Realism. (2) The total image and its historical relationship with the psychic drive of realism. Burch, Morin, Bazin. (3) Ontology. (3) The disappearance of the subject and the visibility of the non-existent. Virilio, Mitchell. Source: own elaboration. 4. The Value of the Aesthetic Experience. The Ages of Looking. For more than a century, most books and articles that have dealt with beauty in art have done so from a psychological perspective. The focus of this operation has been on the human response to beauty in the field of art, i.e., what has been called aesthetic experience (Tatarkiewicz, 2001). The current concept of aesthetic experience assumes that there is a mental attitude in the viewer that influences the perception of the work. From a historical point of view, this component of subjectivity is to be found in the focus of the idealist critique of the judgment of works of art. For Kant, the taste that judges whether a work is or is not beautiful has the pretension of a universal judgment without being able to rationally demonstrate its correctness, since its determining cause is the feeling of the subject and not a concept emanating from the object (Venturi, 2004). Specialists in the study of the early cinema image have established a certain aesthetic disposition towards its viewing derived from its original materiality. The first argument for this disposition is based on the strong epochal imprint conveyed by the early cinema image. Jean Renoir, son of the painter Auguste Renoir, idolized the image conveyed by the Lumière cinematograph because of its capacity to offer us an impression of a period not surpassed by any history book (Rohmer, 1968). As a second argument, we see how the revelation that the prefiguration of this mode of aesthetic reception of early cinema, associated with a certain sense of epoch, lies in the level of development of the technology of the time; a technological development whose consequences are the number of frames per second —from fourteen to twenty—, the instability of the film resulting from a deficient drag system or an unskilled operator, the black and white image, the absence of soundtrack or, in discursive terms, the exclusion of the spectator subject from a non-closed story (Burch, 1990). The power of cinema and its technological evolution end up transferring to the represented object its own capacity, or inability, to register reality. This fact makes indistinguishable our idea of a certain epoch —the transition to the twentieth century— from the qualities of the image attributable to the technological development of the medium to carry out its task of stenographer. Consequently, as a third argument of this particular aesthetic disposition, the early cinema image has led to a case of iconic fetishism where its imaginary is idolized, making it impossible to separate its own materiality from the meaning that has been culturally granted to it as a metaphor of a certain historical time. 9 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 Figure 6. Rue Tverskaïa. Directed by Frères Lumière, 1896. Ensemble of neural networks used for: 1) FPS boosting – to 60 frames per second; 2) Image resolution boosted up a bit with ESRGAN (general dataset); 3) Resorted video sharpness, removed blur, removed compression artefacts; 4) Color: Colorized. Source and License: YouTube, CC BY. However, all these arguments are now in crisis. A new materiality emerges that commands a different aesthetic disposition, objectified in the characteristics of a hyper-technified era whose tentacles extend over all cultural objects. This new aesthetics, which we could call Pythagorean, since its harmony dwells in numerical representation, carries with it the encumbrance of its own time (Agamben, 2004). Thus, the first argument of this new provision 1) states that the image, whether synthetic or not, must be produced and marketed at the highest possible resolution. Although we do not have time for its development, one should consider how much this idea owes to the obsession with graphic quality demanded in the reception of video games (Amanatides, 1987; Pastor, 2017). As a second argument, 2) this desire or drive is implicit in the very growing and inexhaustible logic of hardware and software performance, a simile of the very infinity of the capitalist system from which it emanates (Navarrete- Cardero & Vargas-Iglesias, 2018). Consequently, as a third argument, 3) the power of the image, in the sense wielded by Freedberg (1989), lies more in its tendency to display the result of the upscaling process than in the value of its content. The exchange value imposes itself on the value of use. Thus, the user demands the fulfillment of certain quality requirements, converted into a new standard of vision, but ignores other ideological considerations. The mutation eliminates all traces of the image’s historical charge, a gesture that nobody seems to care about as long as it is viewed at a horizontal resolution of four thousand pixels and at a speed of sixty frames per second. 16 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 Burkov, A. (2019). The hundred-page Machine Learning book. Andriy Burkov. Casetti, F. (1993). Teorías del cine. Cátedra/Signo e Imagen. Channel Denis Shiryaev. February 26, 2020. Moscow, Tverskaya Street in 1896. [Video archive]. Youtube. https://www.youtube.com/watch?v=6FN06Hf1iFk Channel Deoldify Videos. February 4, 2020. Arrival of a Train at La Ciotat (The Lumière Brothers, 1896). [Video archive]. Youtube. https://www.youtube.com/watch?v=EqbOhqXHL7E Channel Nineteenth Century Videos. Back to Life. January 20, 2021. Snow Battle. Lumiere brothers 1897. [Video archive]. Youtube. https://www.youtube.com/watch?v=UaYfi-A7xY0 Channel Nineteenth Century Videos. Back to Life. January 20, 2021. Baby's Meal (Repas de bébé). Louis Lumière. Lyon, France, 1895. [Video archive]. Youtube. https://www.youtube.com/watch?v=e2dsn07FqUI Debray, R. (1994). Vida y muerte de la imagen. Historia de la mirada en Occidente. Paidós. Deleuze, G. & Guattari, F. (1972). Capitalisme et schizophrénie. Vol. 1, L’anti-Oedipe. Minuit. Denzin, N. & Lincoln, Y. S. (2005). Introduction: The Discipline and Practice of Qualitative Research. In Denzin, N. & Lincoln Y. S. (eds.). The Sage Handbook of Qualitative Research, pgs. 1-32. Thousand Oaks, CA: Sage, Eco, U. 1983. The Sign of Three: Dupin, Holmes, Peirce Advances in Semotics. Indiana Uiversity Press. Flick, U. (2015). El diseño de Investigación Cualitativa. Ediciones Morata. Flusser, V. (2017). El universo de las imágenes técnicas. Elogio de la superficialidad. Caja Negra. Fossati, G. (2018). From Grain to Pixel. The Archival Life of Film in Transition. Amsterdam University Press. https://doi.org/10.1086/710922 Foucault, M. (2002). La arqueología del saber. Siglo XXI. Freedberg, D. (1989). The Power of Images. The University of Chicago Press. Gadamer, H. G. (2004). Truth and Method. Bloomsbury, Contiuum Impacts. García Varas, A. 2010. Tiempo, cuerpo y percepción en la imagen técnica. Paul Virilio y la estética de la desaparición. STVDIVM. Revista de Humanidades 16, 231-247. Goldmann, L. (1980). Essays on Method in the Sociology of Literature. Telos Press. Gorki, Máximo 1981: El reino de las sombras, en Harry Geduld (edit.). Los escritores frente al cine. Fundamentos. Jones, J. (2012). The Past Is a Moving Picture. University Press of Florida. Keightley, E. & Pickering, M. (2014). Technologies of memory: Practices of remembering in analogue and digital photography. New Media & Society, 16(4), 576-593. https://doi.org/10.1177/1461444814532062 Lammes, S. (2017). Digital Mapping Interfaces: From Immutable Mobiles to Mutable Images. New Media & Society 19(7), 1019-1033. https://doi.org/10.1177/1461444815625920 Liandrat-Guigues, S. & Leutrat, J. L. (2003). Cómo pensar el cine. Cátedra/Signo e Imagen. Manovich, L. (2001). The Language of the New Media. The MIT Press. Manovich, L. (2013). Software Takes Command. Bloomsbury. Márquez, R. (2021). Forspoken. Vida Extra [11 de julio de 2023: https://shorturl.at/gknH0]. McLuhan, M. & R. Powers, B. 1989. The Global Village: Transformations in World Life and Media in the 21st Century. Oxford University Press. Mitchell, W. J. T. (1994). Picture Theory. The University of Chicago Press. Morin, E. (2001). El cine o el hombre imaginario. Paidós. Narin, A., Kaya, C. & Pamuk, Z. 2021. Automatic detection of coronavirus disease (COVID-19) using X-ray images and deep convolutional neural networks. Pattern Anal Applic. https://doi.org/10.1007/s10044-021-00984-y Navarrete-Cardero L. & Vargas-Iglesias. J. J. (2018). Can We Programme Utopia? The Influence of the Digital Neoliberal Discourse on Utopian Videogames. Communication, Capitalism & Critique. Open Access Journal for a Global Sustainable Information Society, 16(2),782-804. https://doi.org/10.31269/triplec.v16i2.961 Pastor, J. (2017). ¿Qué importa más en un videojuego, la historia o los gráficos? Webedia. Xataka [11 de julio de 2023: https://shorturl.at/coruG]. Peters, R. (2014). All There is to See: Film Restoration and Access in the Digital Age. Cinesthesia, 3(2), 3. 17 Rev. Arte, Individuo y Sociedad, avance en línea, pp. 1-17 Ricoeur, P. (1969). Le conflit des interpretations. Essais d’herméneutique. Éditons de Seuil: Paris. Rohmer, É (Director). 1968. Aller au cinema / Louis Lumière. Institut Pedagogique National, Cinematheque Française et Télévision Scolaire. Rosenstone, R. A. (1995). Visions of the Past: The Challenge of Film to Our Idea of History. Harvard University Press. Sánchez, S. (2016). Toward a Non-Time Image. Deleuze in the Digital Era, in Denson, S. & Leyda, J (eds.) Post-Cinema. Theorizing the 21st Century Film. Reframe Books. Tatarkiewicz, W. (2001). Historia de seis ideas. Arte, belleza, forma, creatividad, mímesis y experiencia estética. Tecnos. Thomson, J. B. (1990). Ideología y cultura moderna. Teoría Crítica social en la era de la comunicación de masas. Universidad Autónoma Metropolitana. Venturi, L. (2004). Historia de la crítica de arte. Debolsillo. Virilio, P. (1980). La estética de la desaparición. Anagrama. Virilio, P. (1989). La máquina de visión. Cátedra/Signo e Imagen.