Quickplay has struck its first acquisition since being carved out of AT&T back in Q1 2020, to take the Canadian cloud-based OTT video software vendor into the realm of media services and media asset management (MAM). An unfamiliar name to many, the target is 440 Digital, a business bringing automation tools into Quickplay’s ranks to simplify the content onboarding process for video customers.
While Quickplay has long been in the CMS space, Quickplay is now in the MAM game. The two can be loosely demarcated as CMS manages content in one archive, while MAM manages digital files across an entire organization.
We aren’t saying that Quickplay has sold us a red herring, yet our numerous conversations with company executives over the years, as both Firstlight and Quickplay, have rarely ventured away from encoding, CMS, cloudification, personalization, and data analytics. A lot of people did not see this one coming.
Such is the significance of this deal that Quickplay is not burying 440 Digital’s assets somewhere inside the existing platform, but is using the acquisition to grow a whole new arm – Quickplay Media Services.
As many readers will know, MAM is a minefield for content distributors juggling different sources, formats, nomenclatures, and QC variations. Where 440 Digital is adept is in normalizing all this supporting metadata for content in an automated process, versioned for the specific needs of customers.
However, MAM itself is evolving. We reported earlier this year on Sinclair Broadcast’s cloud media migration project, spearheaded by Telestream, in which the US broadcaster decided to strip the term MAM from its vocabulary, presumably in some sort of throwing down of the gauntlet to its legacy technology and recognition of the pivot to centralized content sharing between its two stations.
Whatever way you slice it, this is fundamentally still media asset management. The issue for broadcasters is that they have been sluggish to embrace the cloud, building in-house systems for content preparation workflows that have struggled to migrate to cloud infrastructure. And for some reason, many want to leave MAM behind.
The need for a standalone Quickplay Media Services wing is obvious, given the breadth of services on offer. Complementing Quickplay’s expertise in video and audio transcoding, the acquisition takes on onboarding of content providers and distributor endpoints, managing ingress and egress of drop-off locations.
At its core is the identification, tracking, and quality control of media assets, which extends to pre-processing and enrichment of content, and then onto the all-important stage of metadata aggregation and normalization. Crucial to this is support for subtitles and closed caption conversion, plus avails, planners and content rights management, and finally preparation for and delivery of media packages.
Presumably, once Quickplay Media Services has completed its MAM and media processing tasks, content will be conveniently positioned to hand over to Quickplay proper for delivery, personalization, and front-end customization, all via cloud-based software. This is likely to come at the cost of cutting off former MAM partners, but at the benefit of differentiating from rivals.
As far as we can tell, the likes of Kaltura, Brightcove, Vidmind, perhaps even Wowza, don’t have the same MAM credentials that Quickplay now boasts. Imagen is a traditional MAM provider that springs to mind, along with Nagra’s DVnor division for content preparation and post-production services.
There is also likely to be overlap between the new Quickplay Media Services wing and VionLabs, the Swedish AI-based video metadata firm that Quickplay recently teamed up with to disrupt the business of metadata tagging.
Vionlabs, founded in 2018, extracts information directly from audio and video files using AI – based on trained neural networks to understand positive and negative emotions from viewers, as well as how we perceive sound. This produces a rich and emotional analysis that is summarized into a highly dense data point, where machine learning is used to encode heaps of information within a video asset into a single coordinate – aka a media fingerprint.
Without access to third-party APIs or sources, Vionlabs’ fingerprint platform provides video clients with mood, predicted genre, high-dimensional content similarity embedding, and keywords/video story descriptors.
Part of the issue is that traditional manual metadata tagging is not analytical enough to know the difference, and poor recommendations are mostly due to a lack of quality metadata. You can pump a title full of keywords until the cows come home, but it won’t necessarily make a title more relevant.
The addition of Vionlabs is timely because AI is a powerful addition to any MAM system – as an accelerant for content discoverability, content sharing, and content monetization through the mediums of AI-based transcription, translation, optical recognition, and even facial recognition.
Increased demand for MAM systems in various guises has grown on the back of ballooning content libraries. This in turn has required more powerful and intuitive systems for enabling partners to quickly acquire content assets from these portfolios.
Backed by parent company Highview Capital Partners, after enjoying a brief period of two years frolicking as Firstlight Media, the business returned to its Quickplay Media roots for the sake of clarity, as well as clients’ sanity.
Much of that period as Firstlight was spent emphasizing the cloud, buddying up ever closer to Google to build pay TV-grade redundancy and resiliency on cloud infrastructure, while expanding in analytics (including a dedicated sports data team), and monetization methods (like FAST), including adding video stitching and manifest manipulation for server-side ad insertion.
Originally, Quickplay’s content delivery capabilities were positioned to complement Firstlight’s AI-based personalization software and user engagement platform to create a one-stop OTT shop for premium video delivery. While times have changed, the ambition to build a one-stop shop has not. The shop is just selling slightly different products.