Making Metadata Work

Bruce Devlin - new

Published 10th September 2019

Making Metadata Work

I dare say that if you’re a keen reader of my column then you will be brushing up on your Dutch pronunciation and acclimatising yourself to mayonnaise on your chips rather than ketchup. If you’re really keen then you’ll also be making sure that you practise your Dutch jokes about the Flemish to ensure you don’t accidentally tell a Flemish joke about the Dutch at 2am on Sunday morning in a local bar somewhere near the red light district in Amsterdam.

I must admit that my command of the Dutch language is non-existent, but my French is OK and I can recognise but not replicate some of the differences between French French, Belgian French and Canadian French. In the increasingly global media industry, addressing these subtle differences in a common language can make a real difference to the monetisation of content. Territories like France and Canada are big enough that different voice talent might be appropriate for a big title to ensure that a dubbed version of your content sounds more local.

It would be even better if there were some way of reliably labelling your content so that you knew exactly which variant of what language was used in the dubbing. Additionally it would be even better to know exactly which script for what language was used for the subtitles. Not so important in the UK, but is some countries there may be several different ways of writing the language down.

This is not a new problem and we have standards like RFC 1766 which was obsoleted by RFC 3066 which was replaced by RFC 4646 that has transformed into BCP 47 – The IETF’s Best Current Practise for language labelling. This is a great document, but when you look at all the option in it, you realise that there are millions of labels and attribute combinations for every niche dialect and script.

Not all of these language tags are used to buy and sell content, but many of them are more common than you think. To a non-native, it can be quite confusing to know whether a piece of Mongolian content should be labelled mn or mn-Cyrl-MN (Mongolian written in the Cyrillic script as used in Mongolia)

This is where a very nice piece of work done within MESA (the Media and Entertainment Service Alliance) comes in. The Language Metadata Table is a reference of what is actually used to label content rather than what is theoretically possible. It’s freely downloadable and very straightforward to understand. Whether you’re labelling content within a SMPTE ST 377-4 Multi-Channel Audio Framework for use in IMF or just putting a tag in your EIDR database, this is a piece of work you should be excited about. You really should. Start using it now and it will pay dividends in the future.

https://www.mesalliance.org/language-metadata-table

Which brings me to my traditional “What should I be asking vendors at IBC?” question. This year it’s very simple. “How do I transport my metadata in your product?”

This is important because cloud workflows are popular, financially viable and consist mostly of keeping the content at rest and searching / interchanging metadata. Your metadata is the new currency of media knowledge – spend it wisely and ensure your vendors don’t waste it.

Heb een fantastische september in Amsterdam

Bruce

Related Listings

Related Articles

Related News

Related Videos

© KitPlus (tv-bay limited). All trademarks recognised. Reproduction of this content is strictly prohibited without written consent.