SMPTE creates Language Metadata Table with MESA

In collaboration with MESA, SMPTE is bringing a human- and machine-readable Language Metadata Table (LMT) into its public review process, an early step towards SMPTE standardisation.

Published as an SMPTE Public Committee Draft (CD), the vetted and approved list of language codes will be readily available for public comment, implementation and validation.

The LMT register is intended to give media companies, content owners, video service providers and others a controlled vocabulary and standardised set of codes for accurately and consistently identifying spoken and written language, in turn supporting more efficient interchange of media worldwide.

Reflecting many thousands of permutations, LMT codes support numerous applications including audio, written and timed-text (closed captions and subtitles), accessibility, licensing, content localisation and international distribution.

“The Language Metadata Table was started at WarnerMedia in 2017 to normalise language codes within the organisation, and IETF BCP 47 was selected due to its flexibility,” said Yonah Levenson, LMT chair at MESA, the LMT sponsor. “As interest in LMT as the M&E industry’s language code solution increased, SMPTE recognised the value of the LMT and came on board as the technical partner/adviser.”

Work on the LMT register has begun in SMPTE Technology Committees (TCs), which will produce an SMPTE Public CD in the first half of 2021. The Public CD process allows SMPTE to put the LMT register into the public domain quickly and then start the work of gathering feedback and making improvements to both the register and guidelines for its independent management by multiple stakeholders.

Bruce Devlin, SMPTE standards vice president, who worked with MESA to develop the prototype LMT register and online interface currently being reviewed by SMPTE TCs, commented: “If you buy and sell media, you understand that a common vocabulary for language tagging is sorely needed. The LMT register accounts for all languages as well as dialects and scripts. As we see the register through the Public CD process, our hope is that the LMT register will become a canonical resource that serves the needs of all media organisations and ecosystems. Accessing this data will be as simple as clicking on a link or using an API to grab required codes.”

SMPTE TCs are reviewing the prototype LMT register to determine if the structure of the dictionary is correct and if the process for updating that dictionary is correct. After this step is complete, the dictionary and update process will enter a public review period, during which people and organisations can try out the register and use a dedicated GitHub repository to provide real-world feedback that will help the improvement of the register. The Society will leverage the SMPTE knowledge network, which is built on a flexible Microsoft Teams environment with integrated apps including the Microsoft 365 suite and GitHub, to bring agility and efficiency to the Public CD process.

Devlin concluded: “My hope as the SMPTE standards vice president is that ultimately we will have a structured document in SMPTE that defines the LMT, presents the data itself in both human- and machine-readable form, and provides a new administrative guideline that describes how we’ll manage controlled vocabularies and ontologies for third parties. As we undertake the Public CD process for the LMT register, I encourage any organisation or individual with a stake in the internationalisation of content to join the appropriate SMPTE TC and contribute their requirements and expertise.”