Source camera identification tools assist image forensic investigators to associate an image in question with a suspect camera. Various techniques have been developed based on the analysis of the subtle traces left in the images during the acquisition. The Photo Response Non Uniformity (PRNU) noise pattern caused by sensor imperfections has been proven to be an effective way to identify the source camera. The existing literature suggests that the PRNU is the only fingerprint that is device-specific and capable of identifying the exact source device. However, the PRNU is susceptible to camera settings, image content, image processing operations, and counter-forensic attacks. A forensic investigator unaware of counter-forensic attacks or incidental image manipulations is at the risk of getting misled. The spatial synchronization requirement during the matching of two PRNUs also represents a major limitation of the PRNU. In recent years, deep learning based approaches have been successful in identifying source camera models. However, the identification of individual cameras of the same model through these data-driven approaches remains unsatisfactory. In this paper, we bring to light the existence of a new robust data-driven device-specific fingerprint in digital images which is capable of identifying the individual cameras of the same model. It is discovered that the new device fingerprint is location-independent, stochastic, and globally available, which resolve the spatial synchronization issue. Unlike the PRNU, which resides in the high-frequency band, the new device fingerprint is extracted from the low and mid-frequency bands, which resolves the fragility issue that the PRNU is unable to contend with. Our experiments on various datasets demonstrate that the new fingerprint is highly resilient to image manipulations such as rotation, gamma correction, and aggressive JPEG compression.