There are several challenges to using multimedia data:
- The first challenge is size - one colored image could require 24 Mb. In traditional SQL databases the data types used for storing data are of strictly limited size. The format might be very restricted as well (e.g., DATE). Even for string data the limit for VARCHAR2 is currently 32,768 bytes. This can cause difficulties for storing rich media within the database.
- The second challenge is time - called “real time nature” (e.g., in video, the frame sequence). The components of a video have to be maintained in the right order even if they have been processed separately.
- The third challenge is that the semantic nature of multimedia is much more complex. Data retrieval is much more complex with rich media. With traditional structured data, we can search for a specific value in a straightforward manner using SQL and retrieve exact matches or use BOOLEAN operators. With rich media there can be a great deal of information, but that information if described in words would be different for different people. This causes a great deal of confusion in retrieval and indexing rich media and means we may be looking at near matches instead of exact matches.