Characterisation of Large Audio and Video Files

Status

on standby

Contact

Bolette Jurik (SB)

User Story

As the owner of a large collection of video files I need a digital preservation system that can characterise very large audio/video files to enable me (or a watch system) to evaluate the collection for preservation risks and ongoing risk management.

User Requirements/Components

1. We need to be able to characterise very large video (8GB+) files
   1. This includes identifying container formats and contained streams
   2. Features extracted need to be decided - we need to consider what is useful here to extract and include here as requirements. See:
      1. http://www.jiscdigitalmedia.ac.uk/movingimages/advice/metadata-and-digital-video
      2. http://www.jisc.ac.uk/media/documents/programmes/preservation/spmovimages_report.pdf
2. Needs to perform characterisation quickly and efficiently
3. Would be good to be able to validate video format compliance with specification

Experiments

Create experiments as child pages and they should appear automatically here

Execute ffprobe across content (NBR)
Data: Available. Sharable.
Workflow: No.
Issues: Tool too fast to give scale problems.

Developer Notes

The original scenario expressed some concern that existing characterisation tools (JHOVE for instance) do not seem to work well on large files.

This story provides the opportunity to compare tools over time, such as the Tika/DROID/etc. as performed by PC.CC. (See also Incubator)

Related Documents