-
What Ever Happened to Metadata?
Starring Bins, Folders, Libraries, Events, Keyword Collections and more. I asked Bette Davis and Joan Crawford if they would appear but they said the story was too tragic for them, besides which they’re both dead.
So here’s the thing – and some of you will spot that this is a reprise from several other places, but I was hoping we could widen the conversation beyond FCP X alone.
For more than half a decade (at least), we have been talking about how metadata is changing the world around us, and yet in one important area, we seem not to be using it that much, if at all. And that’s media organisation.
We are still contentedly throwing our material into various containers and sub-containers, and sub-sub-containers – but what if there was a different way? What if we could use metadata do that organisation for us?
Obviously there are editing applications that do allow us to leverage the power of metadata to access our material – FCP X and Resolve are the most obvious examples, but the other NLEs do similar things too if not quite at the same level. These allow us to use filtering options to bring specific media (or groups of media types) to the fore and hide the rest.
So my question is: will we always still be using containers of one kind or another, alongside these metadata filtering options, or can you see a time when we might move over entirely to the metadata-driven model?
It’s worth pointing out that when we drop a file into a container of any kind, we are applying a kind of virtual metadata to it – we are describing the file in a particular way such that we will be able to recall it at will. I can make a folder called “Boston Shoot, 05-15-2016” and drop all my relevant camera dailies in there such that I can later recall when and where they were shot. However, if I take any of those files out of that container, the virtual metadata doesn’t follow along with it – it’s volatile and ephemeral.
If instead I embed the place and time of the shoot into the file, I can use a filtering method to access it regardless of where I happen to have put it. And that seems to me to be a major advantage.
You will have spotted of course that your camera may well have embedded the date and even place (if it uses geolocation) of the shoot into the metadata already, along with a whole lot else that is potentially very useful if we were contemplating a filtering method of organisation as against a container-based one.
We can also embed metadata into our files outside of the editing application in a number of ways, the most obvious of which is simply by editing the filename, but there are other possibilities.
And inside FCP X in particular we have ways of adding useful metadata – we can add it into the text fields of individual files, of course, but there is a really powerful method that’s a lot better, and that’s Roles. Assigning Roles to Dialogue, SFX, Music, etc. gives us an easy way of accessing them ad filtering them from different parts of the application and we can do this in batches, which obviously makes it a lot easier. Thus we don’t actually need to make a container called Music or SFX and we can let the Roles metadata do the work for us.
Of course, we don’t have to stick to the Roles conventions that Apple have built in – we could for example have a Role called “Boston” in order to create a metadata “flag” for our Boston shoot. To me that seems potentially more powerful than simply having a container called “Boston” and if you are a power user of Roles I think you’ll see why I say that.
So this is really all speculation at this point. I’m absolutely not suggesting there is a right way of doing it, or a better way. I’m just inviting the question whether in the not too distant future we might be moving towards a more metadata-based approach and away from the classic container-based approach.
I don’t think we have the ideal tools for this just yet, but I would be very interested to see them developed. In particular, I’d like to see better and easier ways of batch embedding metadata at the Finder level. In OSX we now have batch rename which is great, but what about batch embedding anything and everything? Applications like FCP X let us see some of the metadata embedded by our cameras but not all of it. So I just want to stress that this is a question about the future of editing, not a practical suggestion necessarily about how we should be editing now.
I really look forward to hearing your thoughts … thanks for reading this far.
Simon Ubsdell
tokyo productions
hawaiki -
96 Replies
-
Unfortunately the trouble with media and metadata is that camera manufacturers are inconsistent and contradictory. Most have little or no valid metadata at all. Most – many all – NLEs (other than FCP Classic) have no way to actually inject metadata into the media file itself. So while the FCPX methods are very useful, they also only live within the confines of FCPX itself.
Oliver
Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com -
Agreed. It should all be a lot better implemented by now, especially after all this time. So in many ways we are hampered from using filtering methods to access our media because development of metadata whether in our cameras or elsewhere hasn’t kept pace.
Really though, my question is not so much about what we can do now, which clearly has limitations, but what would we like to happen in the (near) future?
Simon Ubsdell
tokyo productions
hawaiki -
Thinking about the second part of your post, there’s presumably then a market for a product (or products) that can externally embed many more kinds of metadata than is currently possible … metadata that travels with the file rather than lives solely inside the editing (or other) application.
EDIT: And most importantly do this as an intelligent batch operation.
Would you agree?
Simon Ubsdell
tokyo productions
hawaiki -
Absolutely. That would be great. But then it’s a question of file formats or more correctly wrapper formats. MOV, R3D, MXF? It’s a minefield. In theory MXF carries the most potential, but it’s also the most opaque.
– Oliver
Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com -
[Simon Ubsdell] “I’d like to see better and easier ways of batch embedding metadata at the Finder level. In OSX we now have batch rename which is great, but what about batch embedding anything and everything?”
Exactly. While FCP X works well with media WITHIN the library. Once you get outside the library all that work you did painstakingly keywording evaporates. If you have to search for a clip where you don’t know what library it is in, you are SOL.
This is one of the primary reasons I don’t spend a lot of time organizing within FCP X. It just doesn’t have a lot of down the road advantages. It’s also why I use the folder level organization you describe, even with it’s flaws. Of course, I’m a bit of a weird animal on the board here being an institutional video producer. I actually have to care about footage and projects I worked on 5-10 years ago. I think too much of our workflows are centered around a once and done mentality. Maybe justifiably so.
At the moment I’m considering Keyflow Pro since it does seem to work with Finder level tags. And it would allow me to search across projects. One rule for me is not being reliant on any one application because in 20 years it’s likely that application will not exist anymore.
-
I’m only vaguely aware of what KeyFlow Pro can do so I’m hoping someone will explain how it may or may not be useful in this context.
EDIT: Way too cumbersome, would be my instant verdict … I’d need something a lot faster and leaner. Great product for its designated purpose though.
Simon Ubsdell
tokyo productions
hawaiki -
[Brett Sherman] “Exactly. While FCP X works well with media WITHIN the library. Once you get outside the library all that work you did painstakingly keywording evaporates. If you have to search for a clip where you don’t know what library it is in, you are SOL.
“FWIW Library manager seems to claim it can search keywords across libraries. But I don’t have the he advanced license to try it out. https://www.arcticwhiteness.com/finalcutlibrarymanager
-
I’m thinking more of how tags now work in OSX – they’re readable across the OS in many Apple applications, and of course are very useful for use with FCP X. (Not readable in Motion though as far as I know …)
Once you’ve tagged your files in the Finder with single or multiple tags each, importing them automatically creates keyword collections from any and every tag. Notice the difference here in that I don’t have to create keyword collections and add the files to each of them separately. I could have a Boston tag, a date tag, a camera lens tag, an ISO tag, etc., etc. all applied to the same file or files, and each would create a keyword collection automatically on import. It then becomes easy to filter for those keyword collections.
What would probably work for me would be an extended version of the Finder tagging option with a dedicated interface that made it much easier to add and edit the tag metadata. In other words, what I would be looking for in terms of external metadata management is very easily achievable by Apple. All it would take is for the developers of non-Apple applications to provide the option to read the tag metadata in the same way the Apple applications can.
-
[Simon Ubsdell] “So my question is: will we always still be using containers of one kind or another, alongside these metadata filtering options, or can you see a time when we might move over entirely to the metadata-driven model?”
I don’t think there will be. kInd of like there will always be filling cabinets even though we have computers.
I think there will always be a multi level approach to most things.
I will use some pics to help describe what I’m talking about this time.
I start off with over 800 b-roll shots from various locations. Some locations have many shots and some just a few. In this example of Coldwater Creek, there are only 10 shots at this location. If you notice, I didn’t even bother to label the individual clips. I didn’t really need to because after one click I went from 800 to 10 and I can see what they are so I don’t waste time naming all these clips. I couldn’t search for these clips by name because I didn’t name them but in this case I can see what they are so I don’t need to.
So when do I use more meta data?
Interviews.
Unlike B-roll or stills, I can’t see what he is talking about by just looking at the interview clip. So when I’m reviewing the interview I make favorites and I label those with notes. Then I can search a subject like Silkwood, and find the section of his interview where he or anyone else is speaking on that topic.
So there you have it. I’m glad Apple was smart enough to give me two options because one size does not fit all methods. Sometimes I need to group with folders because that’s faster and sometimes it pays to add more meta data because that’s faster. It just depends on what you are doing and what becomes more efficient after experience.
-
This is a great conversation. I’ve been interested in metadata for a long time. In my mind MXF carries the most potential as the standards are more set, although mov is good too, but less universal. If you browse an FCPXML of an imported Alexa or Red file to see how much data is there that FCPX reads but doesn’t really allow you to use it’s pretty astounding (can’t make a smart collection on ISO, for instance, or lens length, or fps, or camera position, or color space, other metadata that is already present in files). So having separate metadata files (such as XML) seems to be a safe way to go, but it’s harder to track as you then have to track the media AND metadata files, and if the media and metadata is shared, you have to track the changes between all the systems.
As far as injecting metadata in to files, I think it’s simply a very complicated task due to all the varying file formats we have today. Adobe products do this with XMP and it used to chase fits in FCP Legend. I know we’ve had covers actions before about standardizing metadata, but it’s not s straightforward endeavor. The ACES standard is an attempt to do this with color and other metadata, but is a DJI or GoPro product going to conform to an ACES standard? Probably not.
[Simon Ubsdell] “EDIT: Way too cumbersome, would be my instant verdict … I’d need something a lot faster and leaner. Great product for its designated purpose though.
“There’s a different style of cataloging software currently in beta called Kyno. It sits on top of the Finder so there’s no library file. It sends to X, 7, and Pr, exports XMLs for all of those, and is worth checking out: https://lesspain.software/kyno
Reply to this Discussion! Login or Sign Up