Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


OptimoMaximo last won the day on February 15

OptimoMaximo had the most liked content!

Community Reputation

953 Excellent


About OptimoMaximo

  • Rank
    Maya MahaDeva

Recent Profile Visitors

1,360 profile views
  1. Small correction: Y axis up: Unity, stingray and Maya default settings (but there is the option to switch to Z up in Maya) Zaxis up: Unreal engine, Godot, SL and 3dsmax (and I never heard of a switch up axis in 3dsmax) The collada exporter in Autodesk softwares is a module of the Fbx Plugin, where you can set the export up axis and that by default is Y axis up (an interoperability standard between apps, as in video production the z axis is the scene depth, not the height). So you either find this option under the collada exporter, or you will first need to do an export with fbx setting the up axis as Z in order to make this settings stick in the plug in and forget about it when using the collada exporter. It's a one time setting as it will be stored until you change it again.
  2. Assuming that you're ticking the checkbox for weights import in the third tab and not where you set the LoDs, if that is coming in as static the problem sits in how many vertex group there are on the mesh in total. From the introduction of Bento, you need to remove the unused influences, meaning that bones with zero weights need to be deleted from the vertex groups list, as opposed to before Bento introduction wanted mesh uploaded as.
  3. Exactly what @Klytyna and I have been saying all along in the main feedback thread time ago, and why BoM should have been dropped in favor of a materials layering system. It isn't nice when a piece of fabric has the same specularity and normal mapping as the underlying skin, annihilating fabric details like folds and wrinkles (and making the fabric as shiny as many plasticy skins I've seen around)
  4. Sure. It's ten years that i work on real games on real game engines, thing that Sansar can't be called as, and learned when to bother the programmer and when i can get away with another solution. After these ten years i also learned to code these simpler solutions so to have my own set of choices without bothering the programmers when code is necessary. It's not a matter of "code only" or "animation only", it's more about finding a functional balance where you can, in this example, determine in which direction the door should swing, and therefore decide which animation clip to play, based on the clicker's position in relation to the clicked interactive object orientation. The doors haven't shoved anyone, but the character stood behind it and had to turn around to walk through. If the avatar position would have been more central, it would have been shoved. And during a gameplay where the characters are being chased, that's the most likely situation a player would find theirselves in while rushing to reach the exit or an indoors hideout. Yeah, sure, 1 thousand times the game was played, with repeated users, perhaps. Sansar has a concurrent user base of around 15 people and probably the total active users doesn't even reach that number you're claiming, but oh well. With this said I'm out, i can see the OP raging for a simple question i made that did not deserve such reaction. Keep up that professional looking and working work.
  5. Indeed there's nothing you can do on your side to fix it. It's an animation problem and it needs to be sorted from animation export. Having used avastar 1 or 2 doesn't make a difference, that creator exported a custom joint position avatar to a bvh file including joint positions. The solution on their side is to export disabling joint translation.
  6. Most likely the animations used in the bed have joint positions in use. This happens if the animations include face animations and have been exported in one single file using joint position to a bvh file, which suffers of this bug. Even though only some joints move in position, the whole body gets joint position data, deforming the avatar
  7. That's what instancing is used for in real game engines. Then it IS 4 lines of code. I haven't argued for a second, mine was just a simple question, "have you thought about it?" and that was it. From the looks of it, who's arguing is not me after being somehow triggered.
  8. Four more lines of code are trivially a resource hog, besides if a LOT of testers told us that a shoving door ruins the gameplay, there's nothing exciting about it.
  9. Much easier, not better. Starting from the fact that the whole control would be like 4 lines of code (dot product is a vector math library function call that returns a scalar value and a flow control if-else to determine which animation to play) and this wuold not add such strain on resources, not more than what in use already, under a gameplay point of view it makes it easier to move from a room to another, as in MANY other stealth and/or zombie fighting games where chasing/being chased is a main part of the game. But then if you prefer animating doors and use mixamo's zoombies animation packs, it's up to you (not to mention the baseball bat hitting animation clip...)
  10. Aside from the looks of the shown content, the door that shoves you aside is really bad for a game-play like this. Have you thought to code the door using a dot product between the avatar's forward vector and the door's so to reverse the opening swing direction?
  11. Visual studio code supports lsl but it has no syntax check or test run feature, but it highlights all library functions correctly and does autocompletion, along with handling brackets , quotation marks and parenthesis. I script lsl very very seldom in comparison to how much MEL and python scripting i do, which are greatly supported (therefore my choice)
  12. There are too many shaded areas with this approach. I think LL should work on a proper method to use the wearable assets in compliance with the permission system, instead of leaving holes that may lead to easy to achieve improper practices. While your proof of concept with the eyedropper would be fine, as it retains resolution and may also be charged as an actual upload for permanent storage, it defeats the purpose of bakes on mesh system as a whole by using it as a medium to compose a new static texture as opposed to its intended dynamic nature to quickly add and remove wearables. I understand that you can't wait to make your animeshes with this degree of customization, really. But BOM is already limited and some functionality doesn't properly work yet for serious production other than for legacy avatar uv mapping. The lack of alpha support for all channels makes it unusable already for custom uv mapped avatars, figure if LL can afford to extend it now to use for animesh at its current state.
  13. You can do both, if you play smart with nodes. Most PBR offline renderers don't have a slot for an AO map, but basically what it does is to tone down the reflectivity or the glossiness in the AO darker areas. So if you have a roughness input slot, you can layer the inverted AO map and set it as add or screen blend mode over the roughness map, therefore reducing that area's amount of reflected light, which in turn doesn't allow much light bounces and resulting in a darker texture. I don't know cycles well enough though, so i can't tell whether that is the best approach. For example in VRay the method mentioned above doesn't work as well as layering the AO map over the reflection map input as multiply without inversion. On the other hand, resist the temptation of layering the AO on top of the color map, as that would result in a non physical behavior with shadowed areas darker in color but with reflection highlights.
  14. I think it's the usual issue with the detailed objects you are trying to upload: there is a limit to the number of triangles each material can have, around 21k-ish, that once exceeded, makes the uploader silently split the section in one or more material sections. As result, the lower LoDs generated by the uploader straight from the original file mismatch in such regard and you get the error.
  15. Ah, but would the neighbours appreciate a squarking Lorikeet Well, that's the joy of versioning or, perhaps, scripting on demand behavior with a menu 😉 since it's a learning project, anything goes... So your neighbors would enjoy your lori's crystaline voice when you're online and feel like to have the sounds on, switching it off before logging out... Just an idea
  • Create New...