[Feature] Add ElasticSearch Audio model

[obulat]

Problem

As part of the work on adding Audio to Openverse, we need to make the audio data searchable by indexing the audio data to the ElasticSearch. The models for image are in elasticsearch_models.py

Description

We should look into extracting details that are common for all media, just as we did in the API models.

Alternatives

Additional context

Implementation

• 🙋 I would be interested in implementing this feature.