Digital editing is making the job of detecting altered recordings much more difficult. Because of this, it is very important that the identical CODEC and/or the original storage device be available when attempting to detect alterations in the recorded event. Procedures have been developed by the FBI to provide a comprehensive overview as to what is considered essential to make a valid scientific analysis for detecting alterations.
Start by providing these details:
- Location of recording
- Recording system used in the original recording- brand, model, # of cameras, etc.
- File type provided for examination and type of media containing the recording
- Log sheet or notation of important section/s
- Date of recording and chain of custody
- Type of alteration believed to have taken place
If the original recording was captured via HDD or SSD, and the manufacturer’s duplication process was followed, a digital copy in the same format/codec copied to flash drive or disc must be provided for examination and authentication.
A determination if an image is an original, a reproduction, or even perhaps an artificial manipulation or creation, is a difficult and detailed task. A detailed analysis of the numerous components that comprise the images and sounds must be examined in detail.
Digital files often contain a separate information files known as the file’s metadata. The details contained in the metadata must be examined and cross-referenced for consistency and validity.
The metadata often contains information of when, where, what equipment was employed, and even who captured the image. If the metadata is consistent with all other values comprising the digital image, it can act as a verification stamp for the recording.
But metadata can be easily manipulated to create a believable counterfeit. A detailed examination of the various values in the metadata will align numerous factors in order to resolve or reject the validity of the digital recording.
Other methodologies for authentication include PRNU (Photo Response Non-Uniformity) and PCE (Peak- to-Correlation Energy. These analyses compare an image from a known camera image device to groups of images contained in the video under analysis. These techniques produce tables where patterns are compared from the original to the test image indicating where the authenticity of video recordings is at a variance and therefore suspect.