IMVDb is designed to be a database that collects all sorts of data, and some of the data is media from external sites. For instance, music video shoots are full of people with smart phones posting content about the production process, and we want to list that as behind the scenes material.
To meet this need, we've developed structures where we take in and collect key data about external media that can then be used to produce embeds or other manifestations of the media. If you query the API for behind the scenes data on a video, you might get a an Instragram picture or a YouTube picture, so it's good to know what data we collect on each.
YouTube and Vimeo
YouTube and Vimeo both have identifiers that we store as source_data. These can be used to build embed and query their respective APIs for more information. We also store the title of these videos as it appears on YouTube/Vimeo. That can be found in the title section.
Instagram images have an ID that we store in source_data variable, and we also pull the title. However, the title can get very long since these are more like descriptions.