What are the abilities that are to be taken into consideration if one wants to understand the performance of singular perception, i.e. the incremental perception and tracking of an object as the same unique, or distinct and numerically identical, object? This question formulates what one can term the Problem of Singular Perception (PSP), which presents both philosophical and psychological aspects. Section 1 introduces and formulates PSP. Illustrated by the visual tracking of a flying object which is difficult to identify sortally, section 2 outlines two distinct strategies which aim to resolve PSP, and which are often considered as incompatible approaches. The first strategy appeals to proto- or nonconceptual skills (e.g., visual object files or indexes) whereas the second concentrates on elaborate conceptual abilities (e.g., sortal-based identifications, perceptual inferences). In order to propose a framework for studying PSP and argue that nonconceptualist and conceptualist strategies are not mutually exclusive, this article brings together ideas from three fields that have traditionally been progressing in isolation from each other: the philosophy of reference; the philosophy of object perception; and the psychology of attention. Section 3 presents an ‘argument from updating’ that supports the use of the concept of object file – or, of singular content – to study PSP in an interdisciplinary framework. In addition, it provides the basis for a taxonomical classification of distal properties controlling visual files. While the research on visual object files seems to support a non-conceptualist approach to PSP (and thus to refute a generalized sortalist view), section 4 argues that the object-file framework offers both (1) a minimalist and non-conceptualist and externalist explanation to PSP based on the perceptual anchoring onto a particular distal object and (2) a hybrid conceptualist solution based on perceptual inferences which aim to produce judgments of demonstrative identification.