Can IMF’s current boundaries be extended?
Let me be more specific; what I’m not talking about here is the linear broadcast of a pre-subscribed set of TV programmes vs the on-demand TV from your favourite app; what I’m referring to is that a programme within itself needs to be tailored to the individual viewer.
Examples: A hard of hearing person prefers a louder foreground to background audio mix; a program distributor might wish to put relevant (to the viewer) product placement within a show to maximise advertising revenue; a cookery show might be selected to comprise the stages of making a 3-course meal (of the courses selected by the viewer) in the order that the ingredients need to be cooked, and so forth.
Digital delivery of content to the viewer means there is no reason why all of this shouldn’t be possible but it does require some steps to be carried out first.
The concept of Object Based Broadcasting in relation to the above challenges has been around for years. In fact the BBC has been quite active in this area in their research labs, has published several interesting articles, and has demoed the concepts at IBC.
I love the technology and fully believe that it has a future – the technical implementation is clean and well thought through at many levels. I believe we are only scratching the surface of the types of programmes that can be made with a more viewer-centric solution.
However, it has to be said that the uptake to date appears to be very limited. Why is that? Let me suggest a few reasons:
- There is limited need for the technology in so much that nothing is breaking at the moment.
- To date, programmes produced for the technology are very hand-crafted
The Interoperable Master Format
Now for something a bit different: The Interoperable Master Format (IMF).
When you watch a programme in the UK it may contain 99% of the same material as the version that gets played in the USA, but may, for example, contain a different title sequence. When you watch that programme on a plane, it may have a different composition again and a different playout format for delivery to that seatback screen. Pretty quickly, one programme might have 100’s of different versions for different regions and devices. That volume of versions is difficult to construct, QC, archive, etc.
IMF for Broadcast and Online has been created by the DPP and SMPTE (Society of Motion Picture and Television Engineers) as the specification that delivers the “Distribution Service Master” or archival master. In brief, in the IMF package there is a collection of files consisting in part of: video tracks, audio tracks, metadata tracks, and many composition playlists. The composition playlist (CPL) simply describes which pieces of audio and video (etc) are played and when. For instance, taking our above example, there might be a CPL for the UK market and a CPL for the USA market. They might be almost exactly the same assets except that the UK CPL might point to a different title sequence asset than the USA CPL. I’ve only touched the surface: Netflix have written a good blog on the IMF which can be found here.
Hey presto: now because distributors just need to keep a CPL per region they don’t need to keep 100’s of copies of the whole programme render.
IMF has achieved good publicity and is being more widely adopted in cinema. It tackles a problem with a practical solution. One that broadcasters will need to solve as more content is sold to fulfil global online video players.
Can IMF be extended beyond its current limits?
Wouldn’t it be great if the practicality and industry adoption of IMF could be applied to the dynamic and user-centric aims of Object Based Broadcast?
Netflix specifically say “IMF is not intended to be delivered to the consumer directly” but it would seem clear to me that, in their own way, these two processes have some similarities. They both realise that broadcast is composed of multiple “objects” (e.g., video, audio and subtitle assets) and that the trick is to have a language to describe the composition of those objects.
Imagine a system that dynamically generates a CPL according to the consumer’s needs or actions. Could that CPL then use the “objects” that have been delivered to the consumer to render a unique playout experience?
There are some major differences between IMF and Object Based Broadcasting. I am not suggesting that the user device should transcode the playout from master assets (as would happen in IMF)!! IMF, as Netflix say, is not to be delivered to the consumer directly.
It is clear that Object Based Broadcasting is desirable. Perhaps it could pick up a few tricks from IMF on its road to market? How IMF has picked up popularity, how IMF splits out the playout list (CPL) from the format in which the playout is delivered – the Output Playlist (OPL), how IMF describes assets and composes them into a deliverable, tools that exist for IMF, and so on.
Or could it be that IMF could learn from Object Based Broadcasting and that its future lies in a more dynamic, component-based user-centric world?
Object Matrix will be at BVE (#Q23 respectively) from 27th February to 1st March and at NAB (booth #SL5005 respectively) from 7th – 12th April.
About Object Matrix
Object Matrix is an award winning UK based software company that pioneered Digital Content Governance (DCG), object storage and the modernisation of digital video workflows. Our media focused private and hybrid cloud solutions are tightly integrated into file based and IP workflows and bring economic and operational benefit to all of our customers. Our flagship product, MatrixStore, is used by the world’s largest organisations that create and distribute video content, including NBC, TV Globo, MSG-N, the BBC & BT.