The following notes are primarily taken from the Content Creation User Group meeting, held on Thursday, February 15th, 2018 at 13:00 SLT. For the purposes of Animesh testing, the meetings have relocated to the Animesh4 region on Aditi, the beta grid – look for the seating area towards the middle of the region. The meeting is chaired by Vir Linden, and agenda notes, etc, are usually available on the Content Creation User Group wiki page.
Medhue Simoni streamed the meeting, and his video is appended to the end of this update. Timestamps in the text will open the video in a separate browser tab at the relevant point in the meeting. As always, these notes cover the core points of discussion, and my thanks to Medhue for recording it.
Bakes on Mesh
Extending the current avatar baking service to allow wearable textures (skins, tattoos, clothing) to be applied directly to mesh bodies as well as system avatars. This involves server-side changes, including updating the baking service to support 1024×1024 textures, and may in time lead to a reduction in the complexity of mesh avatar bodies and heads. The project is in two phases:
- The current work to update the baking service to support 1024×1024 textures on avatar meshes.
- An intended follow-on project to actually support baking textures onto avatar mesh surfaces (and potentially other mesh objects as well). This has yet to fully defined in terms of implementation and when it might be slotted into SL development time frames.
This work does not include normal or specular map support, as these are not part of the existing baking service.
- A reminder that this will not initially include Animesh as the Baking Service requires an avatar shape, which Animesh currently does not support.
- There is a lot of confusion about this project and what it will do, so an overview document has been suggested.
- One example of some confusion: Anchor linden mentions being able to “upload 1024 textures” – this is actually a reference to the Baking Service supporting 1024×1024 textures, not to users uploading 1024 textures (because they are already supported).
- [5:21] The target date for completing this first phase of this work is the end of March 2018.
- [12:09-22:15] Alpha layer and Alpha masking discussion:
- [12:09-13:55] A means to hide the system avatar has yet to be determined, but Anchor hopes to focus on this in the course of the next week.
- [13:06-15:44] One suggestion for this is to have a new “avatar” alpha mask which would allow specific parts of the system avatar to be hidden in a similar manner to how it is currently handled (e.g. to allow the body to be hidden when using a mesh body, but the head exposed, for those not using a mesh head).
- [13:55-15:44] These updates should allow elements of a mesh body to be hidden much as can be done with the system avatar, possibly eliminating the need to break the avatar into multiple faces which can be individually masked. However, this has yet to be tested.
- [19:05-22:15] Repeat that alpha layering via the Baking Service (e.g. make-up / tattoo layers), and full alpha masking (e.g. hiding an entire body / body part) should work with mesh avatars exactly the same was as for system avatars at present (again, as per above, without the need for the mesh itself to have multiple faces which can be individually masked).
- [22:56-24:46] Clothing layers when applied to a mesh body should hopefully work as per the system avatar (e.g. “jacket” layers are applied over “shirt” layer, which are applied over “underwear” layers, with the last worn item in each layer being uppermost and the most visible.
- [25:56-27:37] Discussion on wearable layers for specific mesh models, branding and creator / user understanding.
[3:55-5:03] This is the code-name for the project to re-evaluate object and avatar rendering costs. It also now falls under Graham Linden’s leadership as a part of the overall rendering system work.
[32:02-34:06] Oz explains the function of the project, and the fact that originally, the weightings (land impact / avatar complexity) applied to avatars and in-world objects, were somewhat biased to allow for the simulator having to handle a lot of texture, etc., transfers to the viewer (so for example, avatar complexity include texture counts, LI doesn’t). As all of this data now goes via the CDNs, rather than the simulators, this bias can be removed, and the actual costs of rendering complex objects made (hopefully) a lot more accurate and thus promote better content creation.
[34:48-35:48] There are currently no details of how LI / avatar complexity values may change, as the Lab is still gathering data at this point in time in order to make informed decisions on how best to revised the calculations.
[35:51-37:04] For Animesh this means the project will be released with an initial land impact calculation assigned to it for objects. However, as with the rest of SL, these may be subject to change as a result of the work on Project Arctan. See this Animesh forum comment from Vir for more.
- [39:59-43:30] In brief: the value for Animesh will primarily be based on tri count based on the high detail LOD version of the mesh, with Medium,, Low and Lowest having less of an impact in pushing up the overall LI. This will be added to a basic cost component for driving an avatar skeleton (uniform across all Animesh objects). Again, no precise values are available as yet, as LL are still performance testing.
[37:07-39:40] Land Impact: obviously, Land Impact changes are potentially disruptive. To ease this, the Lab plan to add code to the simulator and the viewer to calculate the new LI values but not enforce them. Instead the data on LI using the new calculations will be gathered with data on the existing values, and the Lab will then assess how many parcels they might push over their LI capacity (and thus force object returns) and by how much, were the new values to be enforced. Depending on the overall impact, they will look at increasing region land capacity to compensate to some degree.
So, for example, if the new calculations so parcels will on average go over their LI limit by 10% under the new calculations, region LI might be increased by 15% to compensate. Then, for those who still exceed their limit, there will be a period of grace when then can consolidate and bring their LI use within the limit of the revised calculations before the latter are enforced.
Overall, the Lab is extremely aware of the risks in altering LI values, and will be approaching this work carefully to try to avoid it having a major negative impact with object returns, etc.
- Some of the Animesh test regions on Aditi (animesh1 through Animesh4, Animesh Adult and Animesh XL, have become more sandbox than testing areas (including being used for non-Animesh items). To prevent this, the regions are to start being cleaned-up, auto-return made a little more aggressive and warnings that they are only for Animesh are to be added to the About Land descriptions.
- [50:52-51:50] Worn Animesh Limit: there shouldn’t be an LI cap on worn Animesh, although the limit of only wearing one Animesh at a time remains in place pending performance testing results.
Rendering Improvements and Fixes
[2:10-3:50] The Lab recently split work on the viewer’s rendering pipe into a separate viewer development branch (Project Render, with a project viewer – version 18.104.22.1682446, dated February 9th at the time of writing – currently available). Graham Linden, a long-time Linden, will be working on rendering updates more-or-less full-time going forward. He’ll be looking at things like increasing rendering stability, fixing bugs, and potentially other aspects of the rendering system.