Updating to Audio Analysis V7

To take advantage of our latest classifiers, you need to use Audio Analysis V7. If you are currently using Audio Analysis V6 we recommend to switch to V7.

We’ve designed this update to be as seamless as possible. The only required change is updating your GraphQL queries:

• Replace the audioAnalysisV6 resolver with audioAnalysisV7
• Use AudioAnalysisV7Finished, AudioAnalysisV7Failed, AudioAnalysisV7NotStarted, AudioAnalysisV7Enqueued, AudioAnalysisV7Processing instead of their V6 counterparts

Note

Update your GraphQL queries by replacing the audioAnalysisV6 resolver with audioAnalysisV7 and using the V7 status types instead of their V6 counterparts.

Webhooks

Webhooks continue to use the AudioAnalysisV6 event type for both AudioAnalysisV6 and AudioAnalysisV7.

GraphQL Query Changes

The following example illustrates what you need to adjust in your GraphQL queries:

Before:

query LibraryTrackQuery($id: ID!) {
  libraryTrack(id: $id) {
    __typename
    ... on LibraryTrack {
      audioAnalysisV6 {
        __typename
        ... on AudioAnalysisV6Finished {
          result {
            genreTags
          }
        }
        ... on AudioAnalysisV6Failed {
          error {
            message
          }
        }
      }
    }
  }
}

After:

query LibraryTrackQuery($id: ID!) {
  libraryTrack(id: $id) {
    __typename
    ... on LibraryTrack {
      audioAnalysisV7 {
        __typename
        ... on AudioAnalysisV7Finished {
          result {
            genreTags
          }
        }
        ... on AudioAnalysisV7Failed {
          error {
            message
          }
        }
      }
    }
  }
}

Removal of moodMaxTimes in Audio Analysis V7

The moodMaxTimes field is no longer available in AudioAnalysisV7Result. If you currently rely on moodMaxTimes from AudioAnalysisV6Result, you must remove it.

Instead, use the segments field to retrieve detailed mood data. In Audio Analysis V7, mood values are provided as numeric values at the segment level rather than tags.