The Transformation of Transcoding
What was once an easily defined process has become much more complex, as have the solutions that perform it
In the past 3 months, I’ve spent more time with transcoding and live encoding solutions than a full-time compressionist should.
As part of a series of comparative reports, a consulting firm I co-founded a few years ago sponsored a bake-off between professional transcoding and live encoding solutions. From May to July, Transitions, Inc. team members and I spent time peeking under the hood at technical details, running workflow scenarios, and gathering data—lots of data—based on a series of handpicked video clips.
The invitation-only bake-off, in which every company committed to an equal sponsorship amount and 2 days of on-site prep before delivering thousands of transcodes, provided both a micro view of various products and companies willing to verify their marketing pitch against competing systems and a macro view of the transformation transcoding is undergoing.
Between finishing up the raw data gathering on transcoding solutions and moving into detailed analysis, I’d like to share a few preliminary thoughts. These thoughts aren’t limited to just companies that were tested, though, as the goal of this article is to help readers understand which of the five market segments best fits their unique needs.
The first thought is that defining transcoding is becoming more difficult. As the market shifts from simple transcoding (conversion of a video or audio file from one codec to another) to more complex protocol conversions and content fragmentation, there is a move afoot to perform all of these facets of transcoding in both file-based and live run-time encoding scenarios.
The easiest way to discuss transcoding solutions, and how they differ from encoding solutions, is to parse solutions across a five-pronged segmentation.
From the outset, transcoding followed a traditional path, starting with desktop solutions—some integrated into nonlinear
editing and capture tools and others as stand-alone products—that later morphed into workgroup solutions, providing more horsepower for large batch-transcoding jobs.
More often than not, these workgroup solutions were based around a master node handing off jobs to a variety of nodes, but the lack of workflow integration within workgroup solutions led to more robust enterprise-grade solutions, many of which can be easily manipulated to handle complex analysis, transcoding, and distribution scenarios favored by broadcast and media companies.
Some of these enterprise systems take a more hardened approach, in the form of carrier-class solutions, making units that move a step beyond the standard Linux- or Windows-based “pizza box” server into more real-time operating systems and Network Equipment Building System (NEBS)-compliant devices that can be DC-powered to fit within the standard telecom central office infrastructure.
Finally, there’s a move to push forward with software-as-a-service (SaaS) solutions that push the whole transcoding scenario into the cloud. As with every market segmentation, there are exceptions to the rule and companies with products that span multiple segments. Rhozet’s Carbon Coder software tools, for instance, can be purchased in desktop, workgroup, and even
multinode versions, placing it in the range of an enterprise-class solution.
In the same way, ViewCast, whose primary focus is live encoding, integrates the Telestream FlipFactory transcoding tool into its new VMp workflow. ATEME, Digital Rapids, Envivio, and Media Excel all share similar approaches to supporting transcoding in their respective marketing verticals.
Sorenson, best known for its desktop application Squeeze, has been moving forward into the SaaS space of online video transcoding, but it has also announced Squeeze Server, a multinode workgroup and enterprise solution.
For ease of workflow replication for the inaugural transcoding report, Transitions invited primarily enterprise- and carrier-class solution providers rather than desktop- or workgroup-level transcoding tools.
Why limit testing to just two market segments? On the lower-end solutions, Jan Ozer and others have done a good job of examining those products, running most desktop and a few workgroup solutions through their paces. We felt that workflow integration was lacking in many of these products, though, which makes comparisons to mid- and high-level transcoding tools difficult. Likewise, on the online SaaS front, our initial reviews showed a lack of robust workflow integration—short of highly customized programming—that allow online transcoding solutions to compete with enterprise workflow scenarios.
Ability to automatically deliver video to Brightcove for transcoding will streamline workflows.