<p>The unprecedented volume of real videos, as well as the recently emerged challenge of (video) forgery, pose a challenge to the discipline of digital forensics and multimedia security. This research focuses on the distinctively engineering, comprehensive survey of the computational theory and digital video forgery detection. A digital forensics pipeline is proposed and each of the digital forensics pipeline phases; acquisition, features extraction and machine learning, as well as deep learning analysis is scrutinized. The paper offers a thorough and systematic classification and a comparative review of the detection approaches, including handcrafted features, machine learning, and deep learning methodologies across the spatial, the temporal, and the combined spatio-temporal dimensions. A significant and original contribution of this paper is the introduction of XAI digital forensics engineering for the legitimate, legally-defensible, and irrefutable digital forensics engineering system. Practical engineering aspects of the proposed computational system; modularity, efficiency, and performance with standard digital forensics datasets, to the proposed engineering system are incorporated. The paper proposes measures to overcome the most significant engineering challenge identified in the survey, real time edge computing adversarial machine learning, as well as measures to overcome the lack of digital forensics design methodologies.</p>