Academy Software Foundation - Technical Advisory Council (TAC) Meeting - September 20, 2023

Join the meeting at https://zoom-lfx.platform.linuxfoundation.org/meeting/97880950229?password=81d2940e-c055-43b9-9b5a-6cd7d7090feb

Voting member attendance

Premier member Representatives

Project Representatives

Carol Payne - OpenColorIO Representative
Cary Phillips - OpenEXR Representative
Chris Kulla - Open Shading Language (OSL) Representative
Jonathan Stone - MaterialX Representative
Joshua Minor - OpenTimelineIO Representative
Ken Museth - OpenVDB Representative

Industry Representatives

Jean-Francois Panisset - Visual Effects Society

Non Voting Project, Working Groups, Linux Foundation

Other Attendees

Ben Schofield
JT Nelson, Pasadena Open Source consortium / SoCal Blender group
Robin Rowe, Cinepaint
Sean Wallitsch, AWS
Joseph Goldstone, ARRI
Deke Kincaid, Digital Domain
Doug Walker, Autodesk / OCIO
Bill Ballew, Dreamworks
Daniel Heckelberg, Netflix

Antitrust Policy Notice

Linux Foundation meetings involve participation by industry competitors, and it is the intention of the Linux Foundation to conduct all of its activities in accordance with applicable antitrust and competition laws. It is therefore extremely important that attendees adhere to meeting agendas, and be aware of, and not participate in, any activities that are prohibited under applicable US state, federal or foreign antitrust and competition laws.

Examples of types of actions that are prohibited at Linux Foundation meetings and in connection with Linux Foundation activities are described in the Linux Foundation Antitrust Policy available at linuxfoundation.org/antitrust-policy. If you have questions about these matters, please contact your company counsel, or if you are a member of the Linux Foundation, feel free to contact Andrew Updegrove of the firm of Gesmer Updegrove LLP, which provides legal counsel to the Linux Foundation.

Agenda

Swag Store feedback #446
Annual Review- DPEL Project Review- 9/20 #472
Dev Days Update #476
Project Proposal - OpenQMC #434

Notes

Swag Store Feedback (Yarille)
- Announced at SIGGRAPH event, all the projects logos on different items: ASWF Swag Store
- Issue created to provide feedback on the store: Issue 446
- some issues with branding, please provide feedback
- Joseph: ordered from the store, the sizes seem rather too large, maybe we should have a sizing guide in inches / centimeters to have something more accurate
- Yarille: a comment was made about inclusiveness of a product description, some things we have control over, some things we don’t. Will look into sizing chart. Some features we can enable / disable, and maybe we just haven’t enabled some features. But product descriptions we don’t have control over. I purchased a shirt for a project, ordered a Lady’s Small, it can to mid leg, we’ll see if something can be done about sizing. My order came really fast, surprising since it’s print on demand.
- Michael (chat): Sample Size Chat
- Joseph: the quality of the fabric and silk screening was great
- Carol: thanks for limiting the colors, but if we have specific issues for a project, should we reach out to you, or just put it in the ticket? Yarille: can just do it over email.
- Carol: thank you for the work on this!
Dev Days Update (Carol)
- Emails went out yesterday, sent out template, trying to make it as easy as possible. We want as many people from member companies to participate. Two templates for outreach: one for you as your company lead, sending out to your company. Second one is for someone you know who would like to participate, has all the links populated, hopefully will make it easier for people to have all the information they need. We’ve already got 17 registered participants, and that’s before a lot of outreach has been done. Should we expect more?
- Participant Outreach Template
- Project Guidelies
- Dev Days Worksheet
- Cary: I put out a solicitation at ILM and got 12 volunteers. Not the profiles I was expecting: 2 juniors, 10 fairly senior people. Larry: had similar results at my company. Need to “rattle the cage” of some more junior people. Cary: senior people, but without a history of contributing, so not just the usual suspects. Matthew Low: started outreach at Dreamworks.
- Carol: Slack channel good place to coordinate.
- Larry: very happy with the results so far
- Carol: someone signed up from Scanline, they will need to sign CLA, so will need to make sure that happens. Will need to coordinate with people who don’t have a company lead.
- Carol: email was to project leads that have volunteered to participate, but in the form where we ask for sign up, we allow participants to sign up for interest in 2 projects. Got a couple of responses for OpenCue, OTIO, so we could arrange for that to happen if someone wants to help supervise.
- Carol: we’re going to be asking for volunteers, even if you aren’t actively contributing to a project and want to volunteer, please let me know.
- Jonathan: in addition to ILM folks Cary mentioned, sent a notice to MaterialX Slack, have had a bunch of people writing back in DMs, is there a way to channel these answers to make it more visible? Carol: yes, they will need to register through the form website. Jonathan: some may or may not have, will make sure to remind them. Carol: MaterialX has more registrants than any other. So make sure to reach out to your community, it has worked for Jonathan. Jonathan: I may have gone too far and promised direct membership, but I will do so.
OpenQMC Proposal (Eric)
- Talked to John Mertic to get some clarity on some legal points with Linux Foundation
- Fully expect it will work out fine, but have to get sign off internally
- Kimball: we’re holding off on getting back to them until we have clarity? Eric: yes, I think so.
DPEL Annual Review (Eric)
- Digital Production Example Library
  - A library of digital assets - 3D scenes, digital cinema footage, etc. - that demonstrate the scale and complexity of modern feature film production, including computer graphics, visual effects and animation….
  - A license that lets it be used by open source, commercial developers, university researchers. Only thing you can’t do is make a movie out of it.
  - DPEL web site
  - We launched DPEL web site in August 2022, did a presentation at Open Source Days.
- Initial Assets
  - ASC StEM2 - Standard Evaluation Material 2
  - Animal Logic ALab - USD Production Scene
  - AWS Noa Character
    - From in-house team FuzzyPixel
    - A Maya model, fully rigged, with hair groom
  - Intel Volumetric Clouds Library
    - OpenVDB, low / medium / high resolution
- 2 new assets
  - Animal Logic ALab - Trailer
    - full media clips, audio, OTIO timeline
  - 4004 Moor Lane - USD Scene from Intel
    - Constructed to be challenging for path tracers: narrow openings to let light in, architectural feature + foliage
    - 20 built in cameras to highlight challenging shot compositions
    - Around 7.5GB
- Near Pipeline: Picchu Edit
  - Contributed by AWS
  - DaVinci Resolve project, source media, and OTIO assets
- Near Pipeline (infra) GitHub web site
  - Implemented by Dreamworks
  - Convert DPEL website from LF Creative to GitHub
    - uses Astro framework
  - Changes through PR review
  - GitHub Actions to preview, deploy
  - Self server
  - 98% done since beginning of summer, about to push it out
- Far Pipeline
  - 2023(?) AWS Airship - complex vehicle, with UDIMs (from short film Spanner)
  - Proposal: ALab trailer OTIO with links to ALab USD cameras
    - ASWF mashup! Combining the worlds of 3D-CG and Editorial
    - Prototyped by Michael + Joshua
    - Michael: wired the 25 USD cameras into the scene, can jump between them. Complexity comes from trying to line up the OTIO, it’s a “conversation” since there’s no piece of software out there to do this. Could this be put into OpenRV? Did the basic stitching together, but haven’t had time to do more. The point is this is what we’re trying to achieve with DPEL, use assets as part of technical conversation about ASWF projects, we want to serve as an example to inspire others.
  - Proposal: Weta large 3D environment
    - Kimball: propose building something called “Open Spaces” which will be typical Weta materials. Will have a small edit as well with some animation. Eric: sounds you have tools to dial in the complexity? Kimball: yes, may be a separate component, trying to do scale testing, so might be a useful side component / debugging tool. We also want something visually appealing.
    - Eric: in early stages of conversation.
  - (2024) New ASC project StEM3-VP: virtual production!
    - StEM was in early 2000s, reference materials for digital production.
    - StEM2 is extended color space example footage, designed to stress a color pipeline.
    - StEM3 is focussed on Virtual Production, committee working on it is co-chaired by David Morin. Not sure what it will look like.
  - Continue to work with USD-WG and MaterialX
    - ALab in DPEL, Chess set in USD Assets
    - Sometimes an asset will pop up, and we’re not sure where it belongs.
    - Chess set wasn’t “production scale”, so ended up in USD Assets, whereas ALab belonged in DPEL.
  - Long-standing need (infra): Download statistics
    - Donors would like to know if anyone is looking at those files. ASC has been asking about StEM2.
    - Asking LF IT, but they aren’t really set up about this. We’ve gotten as far as collecting weekly statistics in a CSV and putting them in a folder.
    - John Mertic was showing some ways to integrate this, looking at using Google Spreadsheet as a visualization interface
- Status
  - Phase 1: license, process, TSC
  - Phase 2: successfully completed
    - Useful assets published, under a useful license
    - 2D and 3D, multiple contributing orgs
    - To do: download stats
    - Observation: production-scale test assets
  - Phase 3 = ?
    - Wish list: production materials in MaterialX - copy pasta, warts and all
    - None of the current assets were built for a production, they were built for testing
    - Moana Island is an example it can be done.
    - Materials
      - Lots of simple materials in DPEL, but nothing you would see on a typical lighter’s screen, would be great to copy paste one of those out of a production, test their software
      - Materials are also hard to move out of house, there can be custom plugins, data in disk files. It’s going to take some work to do this. But it will promote more development that will help in the long run.
    - Animation
      - Noa model is rigged, but doesn’t contain animation
      - Rigging is the last great “black art” that’s facility specific
  - If you have something you wish was in DPEL, or wishing something was in DPEL, there’s a wishlist in the Wiki, post to the Slack and we’ll add it.
  - Michael: if someone can get something cleared to share, having an animated version of an asset isn’t that much bigger of an ask. Something that’s animated for 10s of seconds would be amazing, since it’s a whole other of performance. It’s hard since you have that data in front of you but you can’t share it. Don’t necessarily need the geometric complexity, but would be great to have assets with non trivial animation. Eric: are you something baked out to frames? Michael: imaging every point on every mesh moving for every frame. Example of Finding Nemo data set animating live at SIGGRAPH Real time Live
- Adopted Project Review Criteria
  - Adopted Stage review criteria
  - Don’t have a strong feeling about it either way (moving to Adopted stage)
  - We should probably get our website and download stats sorted out first
- Michael: we were definitely concerned that we would get a couple of assets and that would be it, have been gratified to see steady contributions. Was great to talk about Weta idea, similar to what Animal Logic did. Using internal tooling to generate a custom asset for the community. If we could get 2-3 of those a year, would make it worth while.
  - Michael: concern about bitrot, but so far hasn’t been an issue. The ALab asset continues to be relevant and useful. Contributors have continued to make fixes, for instance follow the evolution of USD.
  - Kimball: can be hard to judge against the CII criteria.
  - Larry: proposal to move to Adopted stage. Carol: seconded
  - JF: we just have quorum, 14 out of 25
  - All for, no opposes or abstains, motion passes.
  - Congratulations! Moved on to Accepted Stage
- Any questions?
  - Kimball: very happy this exists
  - Scott (chat): What’s the story with the Blender short movies? What would need to happen to get some of the shorts added to the DPEL? Michael: We are in conversation with them. Eric: Yes though again, not exactly feature production.
- David: on the question of how to obtain movie data. For Moana Island, the permission was obtained because Nick Cannon made a case internally, and the studio owned the whole IP. But how to talk to other studios? Michael: most studios don’t have the rights to the IP they would share. Larry: there’s another problem is that it’s one thing to prepare a production scale asset, but extracting an asset from a production is really hard, it has connections to the rest of the show and the facility itself. Michael: Disney showed that it is possible, they managed to do it. David: taking the easiest adjacent case, at Sony / Pixar / Dreamworks animation, who own the assets, is there someone we can approach? Michael: Disney also owns Star Wars… Kimball: the Weta environment scene would use generic assets we build for our own purposes that we can give away. Our target is to do OpenPBR material since that’s a pending release thing, the long term idea is to generate baked animation for a weta character. And release a layer for USD animation once that releases. Also release a best practice guide, and present a way of working with USD, how to have multiple departments contributing to USD entities / assets. Different layers we could contribute over time.
- Michael: so we just need to convince the CTO of a company! Kimball: well, a CTO who also attends these meetings! The show asset is always going to be really difficult, would have to be from a studio that owns the content and willing to “go to battle for it”.
- Carol: a lot of studios put out short films, could be easier target? Maybe a Pixar short film that wouldn’t be as big battle as a feature length film?
- Bill Ballew: would a secondary / background character with animation still be interesting? Might be easier to get clearance for that. We’ve looked at contributing, we got to a lot of props that don’t distinguish the look of our films, but is that valuable? Are background characters worth it? Kimball: it’s usually about whether it can be conveniently converted into something that others can interact with. Bill: was assuming baked geometry. Michael: to get an animated USD shot from Pixar would be really useful, and it would be good for a bunch of uses and testing importing USD data. If we can get cached geometry and materials, that would still be good. Until there’s an OpenRigging format. David: the Moana Island is a recognizable asset from a movie, so that had value. Would be nice to have others, get the legal folks to understand the process.
- David: if anyone needs help making those pitches and you think I can help, don’t hesitate to reach out, I can talk to producers, business people… We also have the LF Legal team that can engage with company legal teams if needed.