MPEG to DASH Silverlight future?

MPEG DASH's Future: Unified Format or DASHed Hopes?

MPEG DASH is a hot topic for the online video industry right now, but will it actually lead to a single format future?

By Eric Schumacher-Rasmussen
Posted on December 7, 2011

Without a doubt, the most popular session at Streaming Media West was a panel discussion called MPEG DASH: Driving the Growth of Streaming Using the New HTTP Standard, which pulled a standing-room, overflow crowd, some of whom had snuck away from the HTML5 Video Summit to see what the hubbub was all about.

That should tell you something right there. HTML5 is the hottest topic in online video right now; there was little chatter and exactly zero sessions about MPEG DASH in May at Streaming Media East. But unlike HTML5, HDS, HLS, or any of the other acronyms that have demanded our attention in recent months, MPEG DASH actually holds the promise of bringing some consistency to the oft-fragmented world of streaming — if it passes ratification and if it is fully implemented.

In brief, MPEG DASH is an emerging ISO standard for dynamic adaptive streaming over HTTP, one that could potentially supersede the proprietary approaches of Microsoft’s Smooth Streaming, Adobe’s Dynamic Streaming, and Apple’s HTTP Live Streaming. The standard should be finalized in early 2012, and all three of the aforementioned companies — as well as others including Netflix, Inc.; Qualcomm, Inc.; and Cisco Systems, Inc. — have been integral contributors to the DASH working group.

The appeal of MPEG DASH is obvious. Content publishers would be able to generate a single set of files — based on DASH’s standardized “manifest” — that would play on all devices, reducing both technical headaches and transcoding costs and preventing consumers from having to worry about whether the devices they have in their hands or living rooms will be able to play the content they want to watch.Streaming Media contributing editor Tim Siglin wrote a white paper on the fragmented MPEG-4 format for Transitions, Inc., with sponsorship by both Adobe and Microsoft, that provides a good deal of technical underpinnings of MPEG DASH; it’s available on Tim’s Workflowed blog.

Siglin and fellow contributing editor Jan Ozer both spent a considerable portion of late November obsessing, arguing, and writing about MPEG DASH and the accompanying common encryption and file formats. The fruits of their labors can be found in various articles on StreamingMedia.com (see CFF: One Format to Rule Them All? and What is MPEG DASH? in particular). So it’s clear that the promise of “one format to rule them all” is one that thrills most of us in the industry.


But as encoding expert and master of analogy Ozer points out in a post on hisStreamingLearningCenter.com blog, unless MPEG DASH considers a few key issues during ratification and implementation, it “has more issues than Lindsay Lohan.” Among them is that while Microsoft and Adobe have indicated their plans to support MPEG DASH in practice as well as in theory, Apple hasn’t. Given Apple’s aversion to embracing any technology that limits its ability to tightly control the iOS/iDevice ecosystem, it’s not unreasonable to think that Apple will refrain.

The bigger issue is that even though MPEG DASH standardizes a single delivery protocol, it supports two video codecs, H.264 and WebM, and two file formats, MPEG-2 Transport Stream (M2TS) and a fragmented MPEG-4 version of the ISO Base Media File Format known as the Common File Format (CFF). CFF is royalty-free, but if MPEG DASH includes H.264, Mozilla won’t support it in Firefox, a browser that still holds more than 20 percent of the browser market share.

Even if DASH is implemented correctly, content producers will still have to deliver content in both DASH-compliant players (with H.264 and either M2TS or FMP4) as well as a WebM-compliant player, as DASH does not yet include profiles for WebM. Without WebM support, content producers might not be able to reach Firefox users at all. And Apple’s iOS with HLS delivery will likely remain a world unto itself. All of which means that MPEG DASH’s promise of total interoperability will likely never become a reality.

This article appears in the December/January issue of Streaming Media magazine, out soon.