Audio Content

Recordings

A Recording in soundevent is a singular audio file, typically an unaltered capture of original audio recorded by a device. Uniquely identified and linked to a file path, each Recording preserves its authenticity.

Accompanied by metadata, Recordings reveal crucial details like duration, sample rate, and the number of audio channels, providing technical insights for accurate processing and analysis. Temporal and spatial context can be enriched with date, time, latitude, and longitude coordinates, facilitating organization and comparison based on the time of capture and geographical location.

Optionally, Recordings can include ownership information, valuable for compiling datasets and ensuring correct attribution. A rights field specifies usage permissions. Metadata can be further enriched with additional tags, features, and notes. Tags offer categorical details like recording site or equipment used, while features provide numeric values quantifying specific characteristics. Notes contribute additional context, insights, or pertinent details, fostering a deeper understanding of the recording.

erDiagram
    Recording {
        UUID uuid
        string path
        int samplerate
        int channels
        float duration
        string hash
        date date
        time time
        float latitude
        float longitude
        string rights
    }
    User
    Tag
    Feature
    Note
    Recording ||--o{ Feature : features
    Recording ||--o{ Note : notes
    Recording }|--o{ Tag : tags
    Recording }|--o{ User : owners

Recording Hashes

Recording names can change, but their content remains the same. To ensure we can always recognize a specific recording, regardless of its name changes, we employ a helpful tool: the Hash. In soundevent, we use the MD5 hashing function. Imagine it as a unique fingerprint for each file, derived solely from its content. This way, even if the name switches, the essence of the recording remains unmistakably identified.

Audiovisual Core Metadata Schema

In soundevent, we prioritize compatibility with established standards, specifically the TDWG Audiovisual Core metadata schema. All essential fields required by the schema are integrated into the Recording object. While there may be variations, particularly in the attachment of taxonomic terms, a thoughtful selection of Tags in soundevent enables a smooth alignment of most Recording fields with the Audiovisual Core's standardized terms. This ensures a cohesive collaboration within the metadata realm.

Datasets

In soundevent, Datasets serve as purposeful collections of related audio Recordings, sharing a unified context, like a specific deployment or field study. Each Dataset receives a meaningful name and is accompanied by a descriptive summary, facilitating distinction among multiple datasets. The suggested descriptive information should delve into the dataset's origin, context, and relevance, enriching the understanding of its content. Although soundevent doesn't enforce specific criteria for defining datasets, maintaining cohesive audio collections is recommended. This practice enhances source comprehension and enables more nuanced analysis of the dataset's contents.

erDiagram
    Dataset {
        uuid UUID
        string name
        string description
        datetime created_on
    }
    Recording
    Dataset ||--|{ Recording: recordings