Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Activity Forums Square Box CatDV Metadata standards (Video)

  • Metadata standards (Video)

    Posted by Scott Goddard on April 8, 2014 at 10:58 am

    There seem to be very few industry adopted metadata standards for video. Leaving aside the technical fields, I am curious as to if fellow CatDV users are trying to confirm to any one standard for some of the more descriptive information based elements, at least for the core information?

    XMP/IPTC?
    EBU CORE?
    DUBLIN CORE?

    I am thinking longevity, with archives in mind. It would be great to hear what others are using.

    Scott Goddard

    Neo Verite Limited
    https://neoverite.com

    Dave Clack replied 12 years, 1 month ago 4 Members · 6 Replies
  • 6 Replies
  • Isaac Musselwhite

    April 9, 2014 at 5:26 pm

    I’d be curious of this as well!

  • Bryson Jones

    April 11, 2014 at 3:46 pm

    It’s a tough one out there. We’ve seen some publishing standards and then there are web standards such as Dublin Core (the dc: you see in Adobe and other products) but they might be hard to manage in a workflow.

    Taxonomies are a bit more available and are included in some specs.

    You can see a ton in the wikipedia entry for metadata standards.

    Here’s a few you might peruse if you need to catch a nap.

    Dublin Core
    EBUCore
    NISO MIX
    PBCore
    MPEG-7

    The main issue with “meta data” as we discuss it, is that we oversimplify in the post world. Meta data is always contextual. You have md that you need for masters and then separate md for a raw shot, and then md for an image file and then different md for an image that has a license for the people in that image. And then there’s the md for the music in the master that we mentioned above and then there’s…….. it only goes on. So, you can find standards for each part of your workflow but it’s hard to find one for all aspects of the generalized post production process as we touch so many parts of the data flow.

    Ironic that the larger you go in the industry, the easier it gets. Million dollar DAMs regularly support far fewer types of data than CatDV workflows do in a smaller shop.

    bryson

    bryson “at” northshoreautomation.com

    northshoreautomation.com

  • Scott Goddard

    April 11, 2014 at 4:21 pm

    I agree about the contextual element. Photos have it easy!

    Video has far more stages and outlets where metadata needs to be specific. I am thinking of using the Dublin Core across all assets even if a lot of it is irrelevant. I am thinking that somewhere down the line the main things like date, title and creator should hold through. It’s really about it’s use beyond CatDV, especially when you are looking from a long term archive point of view.

    I have been meaning to work on a Worker task to write metadata to multiple fields. This way we could keep clients on their bespoke field names but have that data transfer to Dublin Core fields. So a mapping task.

    User Field 1 (Asset Number) maps to User Field 52 (Dublin Core: Identifier). This would be a simple worker server query that is automated upon field entry. This could then be fields that the clients never need to see, a hidden tab where they are actually adding to ‘standards’ in the background. The advantage of this is that you could write to multiple standards at the same time so if one ‘standard’ does win out in the future the archive will be semi covered for this.

    XMP temporal metadata interests me but getting this standardized across all of the apps that will touch a typical video file workflow is a long way off.

    https://www.philiphodgetts.com/2012/04/why-is-temporal-xmp-metadata-so-exciting/

    Eventually we will have metadata written into every frame. Imagine watching a finished edit and being able to see where every frame and shot came from originally, it’s original shoot date, gps location, creator, file location etc etc.

    Scott Goddard

    Neo Verite Limited
    https://neoverite.com

  • Bryson Jones

    April 11, 2014 at 4:34 pm

    Scott, you are right on.

    One note, maybe keep an eye on CatDV 11 and Server 7 as it will handle the Adobe md way better and split into individual fields among many other md changes in the backend. (All I can say for now but betas will be coming soon.)

    We do the “copy fields” thing a lot, especially when delivering to outside systems. As part of your archive action, for instance, you can prep the md and get all those fields set up.

    Most of our more complicated workflows include field translations and often concatenations or merges as well. You can set this up easily and have the best of both worlds, but be careful of changes and sync of data.

    bryson

    bryson “at” northshoreautomation.com

    northshoreautomation.com

  • Scott Goddard

    April 11, 2014 at 6:06 pm

    Excellent, thanks for the heads up. Adobe XMP is looking very promising.

    Scott Goddard

    Neo Verite Limited
    https://neoverite.com

  • Dave Clack

    April 12, 2014 at 6:45 pm

    Standards are driving part if CatDV 11 – global metadata fields in CatDV plus our recent tutorial on writing metadata back to source media (using xiftool) all give CatDV the capability as standards evolve

    Dave (at Square Box)

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy