Datasets:

syvai
/

p1

license: other
language:
  - da
size_categories:
  - 100K<n<1M
task_categories:
  - automatic-speech-recognition
  - audio-classification
tags:
  - audio
  - radio
  - danish
  - dr
  - p1
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*.parquet
dataset_info:
  features:
    - name: id
      dtype: string
    - name: title
      dtype: string
    - name: start_time
      dtype: string
    - name: end_time
      dtype: string
    - name: duration_ms
      dtype: int64
    - name: year
      dtype: int32
    - name: month
      dtype: int32
    - name: broadcaster
      dtype: string
    - name: creator_affiliation
      dtype: string
    - name: genre
      dtype: string
    - name: genre_sub
      dtype: string
    - name: categories
      sequence: string
    - name: kaltura_id
      dtype: string
    - name: flavor_id
      dtype: string
    - name: file_id
      dtype: string
    - name: platform
      dtype: string
    - name: holdback_expired_date
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 48000
  splits:
    - name: train

DR P1 Audio Archive

Danish public radio (DR) P1 audio recordings sourced from the kb.dk DR-arkivet (Royal Danish Library DR archive), covering roughly 2006–2022.

Format

Audio: Opus, 24 kbps, mono, in OGG container (transcoded from DR's mp3 archive)
Parquet shards (~500 items each), small row groups for streaming compatibility
Sortable by year / month / start_time

Schema

Each row is one broadcast item with the full audio bytes inline plus rich metadata from the kb.dk Solr index (title, genre, start/end time, broadcaster, etc.).

Source

Catalogue: https://www.kb.dk/ds-api/bff/v1/proxy/search/ (kb.dk DR-arkivet Solr index)
Audio: Kaltura-served serveFlavor mp3, transcoded to Opus on the fly

License

Original recordings copyright Danmarks Radio. Made available via kb.dk's public DR-arkivet for research purposes. Check kb.dk's terms of use before redistribution.