Open topic with navigation
Media Server analysis engines produce records, each of which describes something that happens in a video. A record might describe a recognized face, a scene change, or a word spoken in the audio. When you index data into some systems, such as IDOL Server, you might want to combine the data from many records to produce documents that represent video segments or clips.
For example, the speech-to-text analysis engine produces a record for each word in the audio. If you are indexing data into a SQL database, you could insert a separate row into a database table for each word. If you are indexing data into IDOL Server, you might prefer to combine records so that each document contains the transcription from an entire news story or interview.
The following sections describe the indexing modes that you can choose when you configure Media Server to output data.