As the volume of media increases exponentially, companies need to optimise content for more efficient preparation workflows. From a Media Asset Management perspective, the saturation of metadata can cause confusion and multiple actions undertaken by different teams can impact the quality of metadata. There is a need to locate assets and take action quickly, and to do this, companies need to become smarter about how teams search for and identify content. Modern media workflows are complex chains, with lots of intertwining actions involving many teams. Inefficient processes can lead to duplication of effort and can also result in poor quality and confusing metadata.
Given the fast pace and high volumes of content that flow through media organisations, the need for efficient optimised working is greater than ever. To achieve this, clearly defined processes and metadata schemas need to be implemented consistently throughout the content chain.
A name is more than just a name
When it comes to metadata schemas, the first thing that comes to mind is usually naming conventions. If used inconsistently, naming conventions can make workflows incredibly inefficient. Vague descriptions can also lead to all sorts of technical problems around video, audio and subtitling because key information cannot be easily identified. In QC and Validation workflows that rely on accuracy, it is easy to see how mistakes can be made, without clearly defined metadata descriptions.
Let’s say a media operator marks an asset at a particular time stamp as having “audio distortion” and another operator uses a completely different term such as “audio error” to tag the same problem later on. The end result is not only the duplication of effort for media operators, but also video assets tagged with overcrowded and disconnected metadata. This causes confusion and will also make it harder to locate assets, because using the right description relies on human interpretation rather than a clearly defined metadata schema. If a media operator further down the workflow uses a completely different search term, they will fail to identify the poor audio entirely. This kind of confusion has led to many companies insisting on a stricter approach, using predefined terminology to prevent users from choosing individual tags.
Consistency is key
While it is easy to see why setting up a naming convention is so important, it’s not so easy to define exactly how it should work. This issue is further complicated by the fact that metadata is multi-layered. The identification of key metadata has many purposes, so something important for one team may be disregarded by another. This data is used by different teams in a whole host of different ways, so it is not sufficient to tag assets with generic coverall terms such as “video issues” or “audio issues”.
Instead, tagging needs to specify exactly what the issue is, and this needs to be understood by everyone further down the chain without any ambiguity. There will undoubtedly be times when there is a need to name something that falls outside of the predetermined criteria. In those cases, it needs to be controlled by specific individuals with high-level admin rights. These team members can define the metadata terminology with all the relevant users in mind, rather than having too many cooks in the kitchen.
Metadata’s journey
Alongside implementing effective naming conventions, another key part of making sure metadata works for everyone is understanding that it is not static. Within modern media workflows, assets and their associated metadata travel across multiple systems and organisations. It is therefore critical that both sources and destinations can handle and recognise all of the information. Let’s say a media company receives assets in one location, then uses a recognition service to visualise the metadata, before storing it for later use. When those assets are needed further along the media workflow, the systems must be able to visualise that same metadata.
This can become problematic because different services have different standards for metadata and therefore use different formats. The whole process is further complicated because additional metadata may be added at each stage in the asset’s journey. It’s clear to see why there is a requirement to map information from one format to another to ensure nothing is lost in translation. As the video ecosystem evolves and more stakeholders feed into content production and processing, the need for better interoperability will become increasingly urgent.
The best standard to use?
It’s clear that metadata plays an important role in reducing inefficiencies when searching and indexing, but there are so many different standards out there. So how do you decide which metadata schema is best? That really is a misnomer because there can be no ‘best standard’. Deciding which standard to use is largely down to preference. Although it would certainly make life a lot easier if a common standard was used by all, it is unrealistic to expect that to happen. Instead, for optimised processes, media organisations should determine the criteria they work to. Then ensure all internal teams and systems tag and search using those defined standards. In addition, in order to avoid duplication of effort, effective information sharing between teams is crucial, so that everyone is on the same page.
It is also vitally important that post-production applications and infrastructure are customisable. That way, companies can use the standards and language that they prefer without hindering the media workflow. If solutions are designed with flexibility in mind, a media business can stay in control of its workflow. This is infinitely more desirable than the opposite which sees variables within the workflow, and all the limitations that brings, controlling how a media company operates.
