Take a look at your computer desktop. For most of us, it can be cluttered with a bunch of folders, screenshots, and files you haven’t had a chance to organize yet. It’s daunting to even think about going through and consolidating all of it. So it just sits there, unnecessarily taking up storage space.

If you have a hand in post-production workflows, you know that this same problem exists with media libraries. You’re dealing with hundreds of versions of the same title accumulated with each new delivery. Even for slight audio or timed text variances, an entirely new full-length file is created and stored. This issue is magnified with every new OTT service and international distribution point. It’s not just an issue of the intensive workflow process, but it also means a high level of redundancies amongst the different versions. The ramifications on time and cost are huge.  

Ateliere’s Deep Analysis, with our proprietary FrameDNA AI/ML, automatically identifies the duplication amongst multiple versions of your content, retaining only the differences between these versions, eliminating storage and processing redundancies, and allowing you to consolidate the necessary material into IMF packages automatically.

Deep Analysis: Breaking New Grounds 

Ateliere’s Deep Anlaysis uses FrameDNA, an artificial intelligence-driven feature focused on analyzing content and identifying similarities and redundancies. In tandem with the Interoperable Master Format (IMF), this capability allows our clients to reduce the number of mezzanines and deliverables that need to be stored, usually by more than 70%. 

How Does it Work?
There are two key technologies at work in Deep Analysis. The first is driven by artificial intelligence.
FrameDNA works by fingerprinting every frame in the image track of a file upon ingest into Ateliere Connect, and based on structural similarities, identifies the scenes that are different, and allowing for easy comparison of those differences. Deep Analysis can then automatically extract the clips with the differences without having to manually scan through the entire file.
Because FrameDNA engages this workflow in a parallel processing and auto-scaling way, the whole process of scanning and identifying the clips takes minutes.

The second key technology at work in Deep Analysis is our house specialty – IMF generation in the cloud.
Deep Analysis can convert the results of its’ scan of that into a base CPL (the Original Version or OV) that contains your original material as well as various supplemental Composition PlayLists (CPLs) that describe how to combine your original material and deltas together to compose your different versions.

How Can I Apply Deep Analysis to My Workflow? 

Many of our customers use Deep Analysis for de-duplication, including comparing and consolidating texted and textless versions of the same title.
Productions wrap with a base textless version of a title, and over time, the title is augmented with English texted, Spanish texted, daytime, and uncensored equivalents—just as a start. The majority of all that content is the same–it’s just the parts that show in-video text that are different. Storing all of those versions creates unnecessary costs, especially as most workflows are moving to cloud-native digital media supply chains. 

In just minutes, Deep Analysis consolidates all that content into just the base textless source material and all the texted clips, generating IMF Packages (IMPs) with CPLs to represent your versions along the way.  If you’ve got Petabytes of data that you’re storing and handling in the traditional way, Ateliere’s Deep Analysis feature within Connect can take that number and reduce it to Terabytes.  


Contact us for a demonstration or allow us to show you how your content can be de-duplicated and consolidated quickly and easily using Ateliere Connect’s Deep Analysis feature.