Hello
I am aware of the Shared Libraries when the whole collaboration is within one physical network, but how about a real remote cloud based collaboration ?
I saw a pipeline built for Davinci users to collaborate overseas using Davinci cloud to exchange … everything, between two Davincies.
My ambition is quite humble, how to exchange work between two Flames that are physically afar ?
For now we use a Google Drive that can handle up to 2To of media to download locally, and share Batch nodes tree as exported folders on the same cloud, of course there is the archive solution, but to be honest, I never trusted this solution for a daily workflow, I always find its options kind of weird.
What do you guys use when it comes to work in different locations ?
I’m working on a job now with four artists in different locations. We are using Lucid Link. Everyone has the same mount point and pathing, and I’m publishing all the shots out and using open clips and connected conform. Lucid had a feature called pinning, which forces a local cache of pinned files. I keep the publish folder pinned, so as renders come back from my other artists they are automatically copied to my local storage and I get realtime playback. Worst case, I have to scrub/play one of their renders once and it is added to the pinned Lucid cache.
So far, it’s been a wonderful experience! I’ve done the same kind of workflow with synology, but it requires all the artists to have synology units.
This is the way. The missing link at this point is a USD style container that can coexist in multiple places for editorial exchange—openclips are relegated to shot only work of course.
That being said. If you need to move and edit and you’re working non-managed, make an archive and save your edit uncached. Restore it on the other machine and it’s a super simple exchange that costs you 10megs of cloud storage.
Lucid is demoing shared libraries via a shared framestore which each user pins. @Slabrie has said it’s not a supported workflow for managed media but it does work albeit a little clunky depending on internet speed. Also a little warning, those workflows won’t work with SSO Lucid environments.
I’ll just never understand why people do any of this instead of just having cloud/co-location/private cloud/machine room gear and then teradici the artists in. Anything else is a headache, filled with gotchas, usually on sub-par gear.
In this case case, I’m leading a team of freelancers for an agency in-house post facility. Everyone is on their own machine, whatever it is. The agency has no infrastructure at all, except for providing the Lucid account. Before I stepped in and did the publishing with the same pathing, we were sending renders and archives to each other via Lucid.
The viability of the approach isn’t based on the methodology—the methodology is sound. Instead it hinges on the specifics of how you choose to deploy and maybe that’s what you’re saying Alan.
You’ve got 5 freelancers with MacStudios finishing a commercial show from 4K graded plates (a-la 99% of commercial work in LA)? Then there are no real issues. Your success is then ultimately determined by the quality of the internet connections of your freelancer nodes and how competent the artists actually are—much the same as would be the case in terms of latency on a remote connect scenario with some obvious differences.
If you get puritanical enough, sure, you can swing that one way but it comes at a cost. For a few people we work with consistently (who have good hardware) it’s an easy decision. For others it makes sense to spin up an instance and avoid the trouble. But that’s also a differentiating factor of this workflow—you have some flexibility in your deployment. The structure isn’t super rigid which is helpful for smaller organizations.
Ultimately, however strongly you feel about cloud/on-prem/co-locate/remote-in versus lucid and distributed filesystem workflows you can’t over-come the latency issues of working with artists in a significant geographical disparity. Remoting-in just doesn’t work sometimes as a deployment simply because of distance.
In my part of this weird world, it is very common for studios and agencies and brands to not own anything. There is no rack. There is no colo. As in, they have zero capex invested.
As soon as you need to have a stable artist connection from China to US/EU, anything fails or is slow af. So even Internet based solutions aren’t working all over the world sadly due to politics and long distance cables etc.
Besides that I’m rly happy with a mix between lucid and remote control so far, depending on the job, artists and licenses.
Just a few good men. A few talented artists that have good gear, and you don’t have to worry about latency or all the nonsense of trying to run Flame through an internet connection across the world.
For big facilities, maybe that could make sense. For most of us, remoting is a hell that neither the artist nor shop owner wants to deal with.
I’ve seen studios offer both. If they’re based in the US but want to hire talent across the world PCOIP isn’t realistic. The Lucid Link workflow works incredibly well from what I’ve seen.
The delightful situation that we find ourselves in right now is that hybrid workflows are entirely acceptable, and that technology choices for creating same said hybrid workflows are greater in scope and performance than they ever used to be.
Absolutely! Media Cache location (i.e. your Volume) should not be on a storage that is not available all the time to avoid seeing red frames when the content is not available due to internet speed latency or local replication not being completed.