Research Assistant-Professor
Dr. Carloalberto Treccani
Hong Kong Baptist University
Noise-free moon: Noise-cancelling technologies for smartphone cameras
In photography, noise is an unwanted visual element that often causes concern and dissatisfaction. To address this, noise-cancelling techniques were introduced in photo cameras and later in smartphone cameras. Initially, these techniques aimed to reduce noise from technical limitations. However, with advancements in computing and machine vision in the late 90s, they began to be used for more complex tasks, such as the automatic removal of red-eye—the appearance of red pupils due to the use of a flash, commonly seen in many photographs until the mid-2000s—and other denoising processes to improve image quality. The last wave of AI capabilities for smartphone cameras in the second half of the 2010s, thanks particularly to new machine learning techniques, has once again expanded the applications of such technologies. A case in point is the recent Samsung Space Zoom case. The Samsung S23 Ultra smartphone's 100x 'Space Zoom' feature allows the user to shoot detailed photographs of the Moon. In its 2023 video advertisement prior to the smartphone's release, it is possible to see a highly detailed and sharp image of the Moon captured by the smartphone, followed by the phrase 'epic nights are coming' (Samsung 2023). However, a Reddit post showed how such images of the Moon appear to be more the result of a series of complex, sophisticated, and opaque operations rather than the smartphone's powerful new zoom capabilities. A Reddit user, ibreakphotos, devised an ingenious yet simple test to demonstrate that the S23 Ultra generates artificial details in photographs to 'ameliorate' them. According to ibreakphotos, the smartphone did not just restore lost or noisy data; it had created 'a new Moon—a fake one' (Vincent and Porter 2023: n.pag.). The increasing attention to photography and the emergence of new AI possibilities have marked a profound shift in the use of these technologies. Yet, despite their presence on nearly every commercial smartphone and the new concerns, doubts, and issues that this new and noise-free type of photographic image raises, noise-cancelling technologies remain largely unexplored in photography and visual cultural studies. Through a case study—i.e., the Samsung Space Zoom case—this paper provides a) an overview of noise-cancelling technologies, b) investigates their functioning as image-making technologies, and c) examines their role in redefining the contemporary photographic model.
BIO
Since I was a child, I had the need to understand how things worked, trying to understand the working principles that regulate them, breaking them up to the point of no return. Needless to say, my parents weren't always happy with this curiosity of mine. Today, as a Research Assistant Professor at the Department of Humanities and Creative Writing of the Hong Kong Baptist University, my work explores the development of visually 'intelligent' machines and their implications on how we see the world. My academic work has been published in several academic journals, including Leonardo (MIT Press), the European Journal of Media Studies (Amsterdam University Press) and AI & Society (Springer).
<< previous | next >>
Dr. Carloalberto Treccani
Hong Kong Baptist University
Noise-free moon: Noise-cancelling technologies for smartphone cameras
In photography, noise is an unwanted visual element that often causes concern and dissatisfaction. To address this, noise-cancelling techniques were introduced in photo cameras and later in smartphone cameras. Initially, these techniques aimed to reduce noise from technical limitations. However, with advancements in computing and machine vision in the late 90s, they began to be used for more complex tasks, such as the automatic removal of red-eye—the appearance of red pupils due to the use of a flash, commonly seen in many photographs until the mid-2000s—and other denoising processes to improve image quality. The last wave of AI capabilities for smartphone cameras in the second half of the 2010s, thanks particularly to new machine learning techniques, has once again expanded the applications of such technologies. A case in point is the recent Samsung Space Zoom case. The Samsung S23 Ultra smartphone's 100x 'Space Zoom' feature allows the user to shoot detailed photographs of the Moon. In its 2023 video advertisement prior to the smartphone's release, it is possible to see a highly detailed and sharp image of the Moon captured by the smartphone, followed by the phrase 'epic nights are coming' (Samsung 2023). However, a Reddit post showed how such images of the Moon appear to be more the result of a series of complex, sophisticated, and opaque operations rather than the smartphone's powerful new zoom capabilities. A Reddit user, ibreakphotos, devised an ingenious yet simple test to demonstrate that the S23 Ultra generates artificial details in photographs to 'ameliorate' them. According to ibreakphotos, the smartphone did not just restore lost or noisy data; it had created 'a new Moon—a fake one' (Vincent and Porter 2023: n.pag.). The increasing attention to photography and the emergence of new AI possibilities have marked a profound shift in the use of these technologies. Yet, despite their presence on nearly every commercial smartphone and the new concerns, doubts, and issues that this new and noise-free type of photographic image raises, noise-cancelling technologies remain largely unexplored in photography and visual cultural studies. Through a case study—i.e., the Samsung Space Zoom case—this paper provides a) an overview of noise-cancelling technologies, b) investigates their functioning as image-making technologies, and c) examines their role in redefining the contemporary photographic model.
BIO
Since I was a child, I had the need to understand how things worked, trying to understand the working principles that regulate them, breaking them up to the point of no return. Needless to say, my parents weren't always happy with this curiosity of mine. Today, as a Research Assistant Professor at the Department of Humanities and Creative Writing of the Hong Kong Baptist University, my work explores the development of visually 'intelligent' machines and their implications on how we see the world. My academic work has been published in several academic journals, including Leonardo (MIT Press), the European Journal of Media Studies (Amsterdam University Press) and AI & Society (Springer).
<< previous | next >>