ya that's been around for a while, and it's OK. Students come in with questions about it - the Avastar ads scattered through the page confuse the information on how the system and mesh viewer rigs actually work. And don't mention that the Avastar rig is quite different than the viewer rigs. Avastar can be a useful tool for many, though.
I've done my own testing, with results that differ somewhat. After discussing how it could be expected to work with coders (I am not), I've been changing numbers in the xml files, then swapping the new Character folder into the viewer and seeing what crashes it. If no crash then which bits of the avatar changed and how. I demo'd a bit of that at the last OSCC; there's a video, somewhere. Answers have been all over the place. Quite a lot of the xml data does not work as expected by the if-then logic of a good coder.
Tool for working with the system avatar
- Ada Radius
- Posts: 435
- Joined: Sun Dec 23, 2012 6:20 pm
- Has thanked: 659 times
- Been thanked: 548 times
-
- Posts: 1
- Joined: Sat May 04, 2024 9:21 pm
- Has thanked: 0
- Been thanked: 0
Re: Tool for working with the system avatar
Your approach to creating replicas in Blender and enhancing documentation on the viewer Character folder seems like a solid first step. It's understandable that persuading software developers to implement the new avatar may be challenging, but having everyone use the same version of the viewer Character folder could help standardize the experience.Ada Radius wrote: ↑Fri Jun 02, 2023 6:40 pmTess Juel wrote: ↑Thu Jun 01, 2023 10:52 am
Seems great but I'm sure I'm not the only one who are a little bit confused about the scope of the project so can you explain a little bit in simple layman terms?
A few questions off the top of my head:
- Is this intended to be a replacement for the sadly outdated existing system avatar?
Eventually yes. First we need replicas in Blender and far more documentation on what is in the viewer Character folder and how that data relates to the code. The original model was developed in versions of Poser and Maya that no longer exist. Documentation has been thin and some of what is available is wrong.
- How do we persuade the software developers to implement the new avatar? Viewer developers in particular are likely to be very reluctant to add functionalities not suppoerted by Second Life.
We don't. Everyone has to use the same version of the viewer Character folder to see everyone else the same way. BUT groups of people might want to use shared viewer data for their purposes, and the tools I'm working on (with much help from Kayaker Magic and Owl Eyes) could be used to do so. For example: 1) educators who need modesty standards built into the viewer avatar without attachments 2)SciFi and fantasy role play, e.g. changing mesh morphs to make an Appearance slider that makes horns on the head. Also, if you still have problems with the software, you can contact Cogniteq, they specialize in developing various software based on Flutter and therefore can definitely help you solve this issue.
- What kind of changes are within the scope of the project? A better and more detailed mesh? A more advanced rigging? Changes to the skeleton? More shape modifiers? More basic shapes beyond the existing male and female ones?
The xml data and code will support changes to the avatar mesh, once we have an ap that will convert Blender data, dae, or another 3D model format back to the llm format. So yes, changes to the shape modifiers are possible, but not yet. Kayaker Magic demo'd that very thing at OSS - making horns on an avatar head by changing a few numbers in the llm data. He also converted the llm data to dae, which I used to make mesh morph sliders in Blender. He also converted the llm data to a readable format that I used, in a spreadsheet, to figure out the system avatar vertex weighting. But we don't have an ap that will do all that in reverse. Kayaker or Owl Eyes have been working on it, off and on.
The current code and xml data will support some improvements to the rigging - I demo'd that at the last OSCC by making everyone's neck a meter long in my viewer. But it won't support the kinds of changes needed to make a game industry standard rig. Part of the challenges are the SL Appearance sliders. Unlike other game engines.
- I don't know if Kronos has established a standard for virtual reality avatars yet but if they haven't, they certainly will. How will this relate to that?
I don't know - I'd have to see it.
The potential for groups to use shared viewer data for specific purposes, such as educators needing modesty standards or SciFi and fantasy role-playing communities, highlights the versatility and importance of your project. The possibility of implementing changes to the avatar mesh and rigging, as well as exploring improvements to shape modifiers, shows the depth of your commitment to enhancing the avatar system.
Regarding virtual reality avatars, it's essential to stay informed about any emerging standards and consider how your work may align with or complement them in the future.
Overall, your dedication to addressing these issues and collaborating with others to develop solutions is inspiring. Keep up the excellent work, and I look forward to seeing how your project progresses.