IMF has nothing to do with the International Monetary Fund. It is an Application Specification of MXF being developed within the SMPTE.

The goal is to create a delivery format that meets the business needs of shipping versioned content around a country and around the world.

This presentation will be in three parts: (1) a review of IMF technology and how it works; (2) a review of IMF workflows that could exist when an IMF ecosystem exists; and (3) some of the savings that might be realised by using IMF.

The presentation will consider not only the file formats and processes, but also the preservation of multi-platform captions, metadata and media life cycles within MAM systems as well as the benefits that can be achieved by considering versioning from the initial concept of a programme.


Speed reading can often lead to misinterpretation. The title of this paper, “USING IMF FOR INTERNATIONAL DISTRIBUTION”, is intentionally ambiguous, hence the subtitle WHAT DOES THAT MEAN?

Assuming that you interpret IMF as SMPTE’s Interoperable Mastering Format (1), the ambiguity comes from the word using. One reader may think instantly of sending IMF bundles to a destination, another reader may think of IMF as a service mezzanine in a facility, and another reader, as a MAM representation of the versioning requirements.

All of these interpretations are equally valid, and in this paper I won’t try to differentiate between them.

What I will attempt to do is review the technology and some of the working practises, then consider how IMF is a standardised technology that can assist in providing workable solutions to a management and distribution problem that has both national and international implications.

The underlying approach to IMF is that of a componentised workflow. All the different elements in an IMF bundle are stored in separable files so that they can be re-used.

Much of the inspiration for the structure comes from current software working practises where different versions of an application can be made using identical code that is localised using specific libraries for different operating systems.

There exists today a huge number of tools for retargeting and reversioning code to get maximum re-use. In the media world, it would be nice if we could get close to that level of efficiency without losing the essential entertainment value of the underlying content.

The basic problem we’re trying to solve is the creation and management of many versions of the same title (for example a movie, sports programme, episodic or documentary) with efficient use of storage, compute, transfer and human resources.

As an example, a movie distributed for television may have hundreds of different versions created due to differences in language, subtitling, compliance editing (removing video and audio content), localisation (e.g. adding dubbing credits or updating signposts to a local language) and even the addition of telephone numbers and hyper-local inserts for regional content.

Today, there is a broad range of workflows that can achieve this. At one extreme we have tape-based workflows where one tape is created per version (and then a spot reel to fill in the errors found when ingesting the first tape). At the other extreme we have the emerging importance of IMF.


The full IMF standards can be obtained from the SMPTE standards library(1). Useful background reading is the “Digital Cinema Packaging” set of specifications (2) and the AMWA AS-02 specification (3), both of which laid the groundwork for the creation of IMF.

Making a great version of media is like baking a great cake, so I will use a food analogy to help explain the role of the various elements of IMF.

All Essence is in Track Files

When one visits the supermarket, speed and convenience is achieved by having a consistent packing and labelling strategy. The same is true in IMF. All essence is in an MXF track file.