3 minute read
How you handle metadata through your workflow depends on the type of metadata we’re dealing with. Technical metadata always relates to media files and travels with the media file itself, because technical metadata is always about the file itself.
Content Metadata does not live with the media file. It generally lives in a Project or Library specific to the NLE that was used to enter log notes, keyword ranges, and other metadata about the content of the clips.
Because Technical metadata travels with the file, there’s very little work involved in maintaining the technical metadata until you re-encode a file. Until that point, file name and timecode (in particular) were constant. When we re-encode a file, the file name may be lost, modified or in a specific file path (to avoid naming conflicts). Every re-encode changes the creation date and generally also changes the file size and codec.
This becomes important when we’re working with a proxy workflow. Tracking the relationship between original/source media and the proxies used for editorial on feature films has been big businesses for companies like Deluxe and Technicolor. Traditionally, the preparation of proxies, tracking relationships, and matching back to original media after editorial is complete has been outsourced to these companies, or companies like them.
This is still likely to be the workflow for most feature films, particularly those edited with Avid’s Media Composer. It’s also the least stressful workflow for editors and allows a wide selection of codecs or frame sizes for the proxy media.
Final Cut Pro X and Premiere Pro CC both have native proxy workflows, where proxies are created and managed by the host NLE. This also has low stress for editors, but does not provide much flexibility on what is used as a proxy.
For those who want control and want to manage the process themselves, or internally to their organization, Blackmagic Resolve provides a lot of flexibility in forcing relinking.
Content metadata does not live with the file. It exists only in an NLE or Media Asset Management System (MAM). This makes the connection between clips and content metadata rather fragile. In most cases, simply copying the media is not going to bring the logging associated with that asset or clip.
What makes this more complex is that there is no universal method of handling Content Metadata. The way the information is stored in each NLE is proprietary to that NLE. Projects from Premiere Pro are structurally different from Media Composer projects, Resolve projects or Final Cut Pro X Libraries. Because of these different structures, it’s not possible to do a complete translation from one NLE’s Content Metadata to another’s structure.
For this reason, it is always best to remain within one NLE from beginning to end of the project. The exception is going from any NLE for editorial, to Resolve for color grading. While that is generally a smooth transition, there’s no real focus on retaining Content Metadata because Content Metadata is particularly useful in the editing process.
Adobe does embed metadata added to the XMP Metadata Panel into the media files. Once embedded, another install of Premiere Pro CC will read that metadata from the file and make it available in the XMP Panel.
Something similar can be achieved for FCP X using Intelligent Assistance’s FindrCat. FindrCat embeds all the logging information from FCP X into the media file and translates Keywords to Finder Tags. The full metadata can be returned to FCP X from the file using the Pro version of FindrCat.
Another way of storing and managing Content Metadata is using an asset management system known as Digital Asset Management (DAM) or Media Asset Management (MAM) systems. These tools read in Content Metadata from NLE’s via XML and store it within the DAM or MAM in their own data structure.
Most asset management systems support multiple NLEs, so asset metadata can be read into the MAM from one format of XML and clips can be exported with metadata to another format of XML, making an MAM or DAM a means of managing metadata during post-product
Unlock all 100,000 words of the Frame.io Workflow Guide and learn how the pros do workflow.