v7.5 Basefount Miarmy
EDIT in Admin Edit in WP
Downloads
Download demos and trial versions, product documentation, free presets and more.
You may download a free Demo after logging in.
Assemble My Army - Better Crowd Simulation for Maya
Miarmy is the innovative, revolutionary and easy to use toolkit for Autodesk Maya crowd simulation, behavioral animation, creature physics and rendering, making your work more intelligent.
Miarmy (read My Army) is a Human Logic Engine based Maya plugin for crowd simulation, AI & behavioral animation, creature physical simulation and rendering.
- Build human fuzzy logic network without any programming or node connecting
- Support standard production pipeline, reference, HumanIK and Motion Builder
- Integrated with PhysX: ragdoll, RBD emitters, force field, cloth, wind, fluid
- Create Stunning crowd VFX using Maya particles, field, fluid and transform
- Support all renderers
New Features in Miarmy v7
Universal Rig Support
Standard Bone Tree Builder
In the early versions, albeit we support variety types of rig, there still are some limitations such as we need a non-break full-bone-chain joint tree on the rig, so you might have to modify the rig. From now in the new version, we provided a bone tree builder which allow us to rebuild a standard joint tree cling on the original rig even if it contains overall scale, bone scales or break joints. Then we can use the new generated bone tree to generate Original Agent (OA) and bring it into the crowd pipeline without modifying the original rig.
After converting the rig bones and use it generating the Original Agent (OA), we continually need to handle the geometries of character rig, we provided a full set of auxiliary tools (Skinning, UVs and materials checkers) which able to correctly transfer rig geometries to OA. This makes sure we can smoothly finish the OA building and bring the assets into next stages such as real time display and render.
- UVs checker
- Models checker
- Skin weight transfer
- Bone checker
- Face material splitter
OpenGL Rewrite
Viewport 2.0 full support
Utilizing the latest OpenGL, we rewrote all display part of the Miarmy nodes and make it better compatible with viewport 2.0. It is more stable and faster than before, also more intuitive on agent selection.
Nowadays, the previs quality requirement is much higher, for meet the needs of the customers, we rewrite the real time display feature with the newer instance and GPU features of OpenGL, made it able to display tens of thousands agents in a single scene fast and smooth.
Render Rewrite
The procedurals of each popular renderer (Arnold, Redshift, V-Ray, 3DS Max V-Ray and PRMan) has been rewritten. Comparing with previous version, it’s faster because we utilized better renderer API and organize the data better and remove the bounding box in the new version render API, and it’s more stable (especially the random shaders part) and more intuitive (automatical export render contents)
We redesigned and developed the textures and shaders variation pipeline and tools, now it’s easier and more comfortable to work with. We can control the shader and texture randomize with weights and variables easily and add/remove variant directly in GUI. Everything can be customized now, for all renderers.
With above material pipeline, when rendering the scene, all render procedurals’ shaders are binding to the shaders in Maya scene. It reads the materials directly in scene when rendering rather than export the materials info outside Maya. This optimized the material pipeline and made everything more stable, get rid of any trash nodes creations. Additionally, the Alembic export uses the same workflow.
Better User Experience
In the newer version, we can render the agents just like the elements of Maya, get rid of the older exporting and render workflow, most of the render setup and works can be automatically setup. We need to setup renderer correctly first.
We updated and created more than 70 icons and node splitters (4K hi-res screen supported), optimized more than 20 UI and window widgets, changed all progress bar and some prompt dialog and file browser version. Made the software more comfortable to work with.
Scene Inspector is able to easily search and list scene elements like roads, zones, spots ect, allow users easily allocate and find the stuffs in scene and control them in a higher level.
We can use the keywords to search and list the target decision nodes in a heavy scene which contains a lot of logic decision nodes. This facilitates the agent brain debug and logic parameters tweak.
We can easily check whether the actions and their original agent bone structure and name match
We can use this tool to find the bones which turned on some specific attributes. This allow us inspect and debug our OA much easier when the OA has a large number of bones.
Top Features
Powerful tool to bring your simulation to life
Easy and smooth rig pipeline
- Support any bone structure or 3rd party rigging system (HIK, Rocketbox Lib, Advanced Rigging, TSM-Game, all of them)
- No any extra setup for bone
- All skin method available
- Bone scale animation pipeline support
Flexible placement
- Arbitrary placement method
- Inverse placement infrastructure
- Hierarchical placement and solver space
- Script able placement
Full functional animation engine
- Flexible transition
- Action group
- Advanced blend system
- Stable ik and rotation lock
Human logic engine
- No program or connections needed
- Tree structure Auto priority arrangement
- More than 200 channels
- Expression support direct control
Diversity perception contents
- Feel each other by sound, vision, collide or even tell\hear
- Tons of perception contents, road, spot, geo, uv, etc
- Repel system limit agent movement
Useful behavior control
- Advanced Aim system
- Smooth uv pin
- Traffic system
- Action sync system
Maya VFX tools integration
- Feel maya particle, field, fluid, etc.
- Support muscle, ncloth
- Maya script fully control
- Maya transform based solver spade
Physx engine
- Industrial fastest rigid body simulation speed
- Realistic and fastest cloth simulation
- Plenty of physical features
- Scale support in physical simulation
Rendering
- Variety cache method and support layer cache
- Mesh Drive Maya procedural for all render eras
- DSO for VRay Arnold and Renderman
Description
System Requirements
Notice: Miarmy only supports 64 Bit OS
Miarmy 7.5
Windows
- Autodesk Maya 2018
- Autodesk Maya 2019
- Autodesk Maya 2020
- Autodesk Maya 2022
- Autodesk Maya 2023
Linux
- Autodesk Maya 2020
- Autodesk Maya 2022
- Autodesk Maya 2023
Change Log
What’s new in version 7.5?
USD (Universal Scene Description) Support
USD is getting more and more popular in recent production across DCC and look dev software. Miarmy 7.5 supports procedural skeletal mesh and instancing USD exporter which is able to export USD files directly from agents fast. Then you can import the USD to Clarisse iFX, Katana, Houdini Solaris or UE4 etc to render it without setup other render procedurals.
Arnold Fur Procedual Plugin (Yeti and XGen Support)
In the new version, we added a set of new tools which allow us to automatically transfer the XGen and Yeti fur onto the agent geometries and procedurally render fur on each agents.
Action based Cloth, Facial & Deformable Objects Pipeline Tools
We introduced a new tool which can bake the geometries deformation data along with each action from the animated rig and apply the deformation to each agent for different actions in render time.
FBX and Unreal Engine Exporter
Now we can fast export the simulation result into the Unreal Engine.
Unity Plugin for Crowd Importing
As same as the Unreal Engine, we can also export and import simulation result into Unity Game Engine. A new unity plugin has been introduced too to apply simulation result into Unity.
Arnold Facets Shaders Supports
The geometries which assigned facial shaders are supported in Arnold Render now and this allows us to render combined geometries with multiple shaders.
Automatic Clone Render Optimizations
Now we can automatically render most parts of agents as clone to accelerate the render speed.
Cloth Simulation Optimized
The cloth has been refactoried and now the simulation has been optimized especially the reset speed with large scene. Now we can feel more comfortable to simulate thousands of clothes in large scene.
What’s new in version 7?
Universal Rig Support
Standard Bone Tree Builder
In the early versions, albeit we support variety types of rig, there still are some limitations such as we need a non-break full-bone-chain joint tree on the rig, so you might have to modify the rig. From now in the new version, we provided a bone tree builder which allow us to rebuild a standard joint tree cling on the original rig even if it contains overall scale, bone scales or break joints. Then we can use the new generated bone tree to generate Original Agent (OA) and bring it into the crowd pipeline without modifying the original rig.
After converting the rig bones and use it generating the Original Agent (OA), we continually need to handle the geometries of character rig, we provided a full set of auxiliary tools (Skinning, UVs and materials checkers) which able to correctly transfer rig geometries to OA. This makes sure we can smoothly finish the OA building and bring the assets into next stages such as real time display and render.
- UVs checker
- Models checker
- Skin weight transfer
- Bone checker
- Face material splitter
OpenGL Rewrite
Viewport 2.0 full support
Utilizing the latest OpenGL, we rewrote all display part of the Miarmy nodes and make it better compatible with viewport 2.0. It is more stable and faster than before, also more intuitive on agent selection.
Nowadays, the previs quality requirement is much higher, for meet the needs of the customers, we rewrite the real time display feature with the newer instance and GPU features of OpenGL, made it able to display tens of thousands agents in a single scene fast and smooth.
Render Rewrite
The procedurals of each popular renderer (Arnold, Redshift, V-Ray, 3DS Max V-Ray and PRMan) has been rewritten. Comparing with previous version, it’s faster because we utilized better renderer API and organize the data better and remove the bounding box in the new version render API, and it’s more stable (especially the random shaders part) and more intuitive (automatical export render contents)
We redesigned and developed the textures and shaders variation pipeline and tools, now it’s easier and more comfortable to work with. We can control the shader and texture randomize with weights and variables easily and add/remove variant directly in GUI. Everything can be customized now, for all renderers.
With above material pipeline, when rendering the scene, all render procedurals’ shaders are binding to the shaders in Maya scene. It reads the materials directly in scene when rendering rather than export the materials info outside Maya. This optimized the material pipeline and made everything more stable, get rid of any trash nodes creations. Additionally, the Alembic export uses the same workflow.
Better User Experience
In the newer version, we can render the agents just like the elements of Maya, get rid of the older exporting and render workflow, most of the render setup and works can be automatically setup. We need to setup renderer correctly first.
We updated and created more than 70 icons and node splitters (4K hi-res screen supported), optimized more than 20 UI and window widgets, changed all progress bar and some prompt dialog and file browser version. Made the software more comfortable to work with.
Scene Inspector is able to easily search and list scene elements like roads, zones, spots ect, allow users easily allocate and find the stuffs in scene and control them in a higher level.
We can use the keywords to search and list the target decision nodes in a heavy scene which contains a lot of logic decision nodes. This facilitates the agent brain debug and logic parameters tweak.
We can easily check whether the actions and their original agent bone structure and name match
We can use this tool to find the bones which turned on some specific attributes. This allow us inspect and debug our OA much easier when the OA has a large number of bones.
Upgrades
- Maya 2019 Support
- Better auto avoid geometry (direct logic)
- Random left/right when face to wall 10/-10
- Smarter self clone for fish simulation (random amplitude, time, phase, drag, ect)
- Smarter self clone with factor, affected by number of nearby agents
- Drive shell can save agent scale
- 3d creature flag of direct logic now is controlled by logic sentences.
- Both agent match and character cache support action-based blendshape
- Collision speed input language
- Add python command evaluate support after dynamics
- Renew tool upgrade for car/quadruped non flipping creature
- Continue baking after bake some frames
- Check if geometries collapse when maya version changed.
- Geo visibility control by action upgraded
Bug Fixes:
- Repeated name
- Assign shader by SG node instead of shader nodes in some cases
- Add diffuse reflection option in Arnold render
- Manually setup agent variable and render with run-time var mode and a bug.
- Lock/unlock cache translate feature
- Switch the clear trash shader codes with Maya build-in codes
- Bake feet locator issue, attribute type id not setup correctly
- Trash global node deal with
- A bug when using Mesh Drive with geometry with SG nodes but shader
- Turn off problematic parallel compute (Advanced Skeleton) before creating OA and action
- Attach terrain supports disable left/right or front/back
- Real time display mark only non-cloth object, UI changed
- Trim action bug when self-cycling.
- Auto scale CCT radius more than 0.3 if less than that (Or it crashes Maya)
- Layer cache terrain adapt added (previous it only support regular agent cache)
- Fix all dialog UI to the newer version (Maya issue)
- Crash when setup disable dynamics on the second bone from root bone.
- Agent precision error when distance from origin more than 100M
- Fix character cache reading
- Offset redundant when layer cache drive
- Fix agent match error when using dynamics
- Crash when enable agent clone avoid collision while simulation clone (without logic)
- Agent match no need scale compensate
- Redshift memory leak cause crash
- Renderer disordered geometry (need turn off real time display)
- Direct logic in 3d space the priority issue of spot/road/zone
- Bug fix for the 2-sided road
- Major bug of UV reading, change all MFnMesh to itMeshPolygon
- Fixed the brain data value in Maya 2018 (different)
- Remove redundant warning when drive shell
- Timer and Ping-Pong, better time distribution and randomization
What’s new in version 6.5?
Layer Cache Based Rigid Body Simulations
With Layer Cache Based Dynamics simulation, all layer cached agents can enable dynamics and create a shell interactive with the simulating agents. With this technique, we can simulation big number of dynamical agents like 10K-30K or more.
New & Easier Formation Control
With a single direct logic language, we can easily control the formation now in 2D and 3D space. Agent can move forward to the target placer automatically. Also the formation control can work with other logic and arrange the priority, like avoid spot or zone. It will be much easier to simulate sardin fish ball than before.
Self Clone Instancing Simulation and Rendering
For simulating birds or fishes, we added a new self clone toolkit into our system, it is a new system allows us to add some “droid” or “drone” agents around current agent. It works like interpolation, they will move along with the main agent smoothly and can auto avoid each other and perform some minor behaviors, it is going to support more features in future. Also in the background, It supports transformation change, render, cache etc
Dynamic Bones Lock on Geometry
Now we can lock the dynamic segments onto the geometry, then we can deform the geometry by cloth simulation or skin-bone keyframes, achieve more controllable result.
New OpenGL Viewport (Beta)
We’ve been rewriting entire viewport part in last 3 months. Using the latest OpenGL 3.3 and GPU computing, it gets faster and stable, and compatibles with all platforms (win, linux & macOS) and viewport 2. Additionally, the new viewport is more user friendly too. Now beta and more features will be released in Miarmy 7 (September).
New Brain Share UI and Decision Outliner
This feature was born from a real production, with a scene contains more than 40 agents’ types, we can use the latest brain share UI and outliner to help us precisely share and check and edit the decision logic. This can tremendously increase our productivity.
New RenderMan Plugin
In the new version of Miarmy, we’ve re-wrote the prman render procedural and exporter and now it supports pixar renderman and 3deligth better. It support native shader and better randomize shaders, we don’t need to compile and deal with shader anymore. (Coming Soon in A Week)
Improvements
Bug Fixes
- Prman Precision caused jitters on rendered object
- fixed some engine issue of depth parse
- agent reset with cloth reset
- timer not random enough
- rate control conflict with auto sphere control
- brain share some trash nodes cause it not work
Minor Upgrades
- cloth force field
- crossroad re-created
- drive shell support agent scale
- spot distance with geometry (control from a feature of 6.2)
- add dead flag when playing final action
Renderer Support Upgrade
- Arnold 3.0, 3.1 Support
- V-Ray 3.6 Support
- Redshift 2.6 Support
- Prman 21, 22
What’s new in version 6.2.67?
Layer Cache Based Rigid Body Simulations
With Layer Cache Based Dynamics simulation, all layer cached agents can enable dynamics and create a shell interactive with the stimulating agents. With this technique, we can simulate a large number of dynamical agents like 10K-30K or more.
New & Easier Formation Control
With a single direct logic language, we can easily control the formation now in 2D and 3D space. Also, the formation control can automatically work with other logic and arrange the priority, like avoid spot or zone.
Dynamic Bones Lock on Geometry
Now we can lock the dynamic segments onto the geometry, then we can deform the geometry by cloth simulation or skin-bone keyframes, achieve a more controllable result.
Improvements
- Bug Fixes
- Prman Precision caused jitter on a rendered object
- fixed some engine issue of depth parse
- agent reset with cloth reset
- timer not random enough
- rate control conflict with auto sphere control
- brain share some trash nodes cause it not work
Minor Upgrades
- cloth force field
- crossroad reborn
- drive shell support agent scale
- spot distance with geometry (control from a feature of 6.2)
- add a dead flag when playing final action
- Arnold 3.0 Support
- V-Ray 3.6 Support
What’s new in version 6.2?
Easy Cache Editing Solution
After simulation and caching, we can easily fix some small bugs of simulation by grouping and editing tools. And in the new version, we updated it and make it support layer cache and allow the agents inside it are able to adapt terrain.
Exit Trigger
We can use logic to control which action the agent will select to transit to next action in transition map along with exit trigger.
Flocking System
Agents now are able to mimic the orientation of a nearby group of agents and achieve the flocking behavior, this can be highly frequently used in animals simulation like birds, fishes or sheep.
Auto Fight System
2 armies can auto fight each other now, charge, meet, aim and attack can be automatically calculated by our direct logic engine.
Geometry Defined Spot
Now we can define the spot by geometry. The target point won’t be spotted again, it can be random points distributed inside the geometry, we can use it sculpt the shape of agent flock.
New Action Proxy Method
New Action Proxy tool has been involved into the new version, we can easily check the action and tweak the transition of actions much easier than before.
Colorize and Randomize Real Time Display
Before rendering, we can use the real time display tool to preview agents geometries and see is everything all right. And in this new version, we introduced randomize geometry and color on this display geometries.
Redshift Fully Support
Redshift renderer now can read the shader of each geometry and render everything correct.
New Transition Map and Brain Viewer
We bring a new QT based Transition Map and Brain View (in Maya 2018), it makes easier to debug your agents.
Upgrades:
- Redshift Upgrade and Support 2.5
- Maya 2018 Support for Mac, Windows, and Linux
- Arnold 2.0 2.1 Crypto Support
- Footmap Baking upgraded to selection agents
- Prman batch export tool
- Maya 2018 Real Time Display Culling
- Viewport 2.0 Brighness error in Maya 2018
Bug Fixes:
- Clone Cache does not work when multiple placements with Logic
- Inverse Placement crash when agent number exceed 32000
- Clone Crash without real agent placement
- Mesh Drive error when geometry no shader
- Negtive Frame error with Sine and Cosine Logic Function
- 3DS Max Catalog File Reading
What’s new in version 6.0?
- Clone Simulation
- Road, Path Memory, Road resetting
- Any Road
- Direct Logic – Maya Field Control
- Road Placement
- Simple Crossroad
- Fuzzy control in Direct Logic
- Sphere control + Direct Logic Auto Priority
- Render farm for all renderers
- Instance Rendering for All renderer (Arnold, V-Ray, Redshift)
- Better formation control
- Arnold 5 supports.
- Better and faster Geometry Adaption control
- New Brain Share Much Simpler than Before
- Perfect Layer Cache (Fixed all bug, make it much faster and more stable)
Easy Cache Editing Solution
After simulation and caching, we can easily fix some small bugs in simulation by grouping and editing tools. And in the new version, we support layer cache and allow the agents inside being able to adapt terrain.
Flocking System
Agents are now able to mimic the orientation of nearby group of agents and achieve the flocking behavior, this can be used very frequently in animals simulation like birds, fishes or sheep.
Auto Fight System
2 armies can auto fight each other now.
Charge, meet, aim and attack can be automatically calculated by our direct logic engine.
Geometry Defined Spot
Now we can define the spot by geometry. The target point doesn’t have to be spot again. Instead it can be random points distributed inside the geometry. We can use it to sculpt the shape of agent flock.
New Action Proxy Method
New Action Proxy tool has been updated in the new version, we can now check the action and tweak the transition of actions much easier than before.
New Transition Map and Brain Viewer
We bring a new QT based Transition Map and Brain View (in Maya 2018), making it easier to debug your agents.
Redshift Full Support
Redshift renderer can now read the shader of each geometry and render everything correctly.
Exit Trigger
We can use logic to control the action agents will select to transit to next action in transition map along with exit trigger.
Colorize and Randomize Real Time Display
Before rendering, we can use the real time display tool to preview and verify agents geometries. And in this new version, we introduced randomize geometry and color on the display geometries.
What’s new in version 5.5?
Clone Agents
We introduced the Clone feature in this version of Miarmy. The clone is an intelligence instance. It not only the instance but also it supports logic simulations. Clone allows users to build a large number of simple crowds with an intuitive method. Especially in stadium, square or terrain, we can fast deploy a huge number of actors for simulation and rendering.
For dealing with 200K agents in the scene, we can get 20X faster in simulation and 100X faster for rendering (currently supports Arnold & V-Ray)
Direct Logic Engine Upgrade
In the new version of Miarmy, we finished the Direct Logic Engine and also expanded it to 3d space. With the new engine, we can skip most of the logic setup, and directly build most of the logic with a simple setup. For example, if you want to your actors walk along the road, just tell them to “follow the road”. Auto avoids and auto fight logic also added into the engine.
Faster Placement & Viewport
We improved the speed of viewport and the placement infrastructure and many related nodes. Now it can be 2-3 times faster than before verison for displaying the agents and clone agents.
Redshift Renderer Supports
Redshift Render is based on GPU and very popular nowadays, we added this renderer procedural support in this version so that we can render a big number of agents in the scene.
Minor New Features
- Expression can be execute after dynamics
- PP attribute in output language
- Pipeline scripts for auto action creation and build OA (for too many different
- Cache Agent Color
- Blendshape Copy to OA
- Relative speed input language
- Auto fight, meet/fight && front near stop
Improvements
- Depth first priority full tree in logic tree, more intuitive logic setup method
- 3 levels of LOD
- block some agents’ RBD target (RBD target can ignore some agents randomly)
- (Direct Logic) Road rotation speed
- Lock name and proportions in place node (agent type can swap order)
- Arnold Opaque fix
- RBD collide debug nodes
- Arnold render catalog file
- Auto avoid (block push left/right)
- Noise id based on agent id instead of scene index
Bug Fixes
- Drive Shell bug fix with inverse placement
- (Direct Logic) zone indexing distance feel not correct
- Mesh Drive 3 missing type geom stop copying
- Bounding box small than Arnold render requirement cause render stop
- Blendshape drive in referencing mode.
What’s new in version 5.2?
- New Blendshape Pipeline – Action Attribute and Blendshape
- New Sphere Range – Ellipse Range
- New Kinematic Primitive Pipeline – Deforming Kinematic Primitive
- New Layer Cache and Post-Cache Editting Tools – Edit Cache After Sims
- New Independent Innertia Logic – Inertia Control Logic
- New Action Blend Tools – Action Blending
- Mark Mesh as Delete Zone – Agent Delete Mesh
- New Pipeline of Mark Dead Agents
- Before Simulation – Mark Dead for Placers
- After Simulation – Agent Hide List
See full release notes here
Special New Features in v4.7
Buoyancy Mesh
In this new version of Miarmy, we can make some specific geometries as buoyancy mesh. Any agent enabled dynamics can interactive with them and get the result of floating on the water. In addition, we added drag force, draught force and offset for making the result more natural.
Ragdoll Renewal Stand Up
In traditional dynamics simulation, the agent cannot stand up after enabling dynamics and turning into a ragdoll. Even adding servo force which was only able to make agents moving in-place after dynamics. From this version, we have tool which is able to pull agents back from dynamics to action cycle seamlessly. Then, agents can continually join the main logical simulation again.
Arbitrary Directional Stacking
We’ve been keeping in upgrading the CCT in recent releases. This time, we make the CCT stack effect into 3D vertical space. In this new version, even on the wall or ceiling, we can achieve the natural stacking result by adding CCT up vector auto adaption.
Geometry Real Time Display
Miarmy new real-time display feature available, with GPU skin-weight computing and latest nVidia driver, we can display more than 100000 geometries without stuttering.
Faster IK System
Thanks for Autodesk new API from Maya 2015, we made the IK faster. Additionally, we can switch the IK on/off by camera distance, this balances the IK precise result and the efficiency, which can tremendously improve the productively.
Geometry Force (Attract & Repel)
From Miarmy 4.7, we added some geometry force feature, which allow geometries interactive with dynamical agents. It can attract or repel dynamics agents, using this, we can simulate some magnetic field attraction and repelling. Funny features.
Arnold aiUserData
In this new version, we added Arnold aiUserData feature, with this tool, we can randomize more on the agents, for adding every one of them unique color. Also, we can use this add color ID for agents, which allow us adjust specific agent in post-production software.
Small improvements
- dynamic simulation can return speed
- indexing for kinematic primitive like _KINEPRIM_[??]
- API for render character cache
- ?? ~ ?? syntax can be recognized in the language input
- ik lock preview slots increased
- batch mode for exporting v-ray contents
- contact/pin restricted outside the mesh, for avoiding the jitter caused by colliding
- character cache workflow supports randomize geometries
- V-Ray render culling
- foot generate locator with agent type id info
- substep attribute can be keyable
Special New Features in v4.5
Jump System
Miarmy 4.5 introduced the new jump system. It allows agents to jump among the trees in the forest. We can easily randomize the actions of climb and jump. Also, it allows us to determine which trees the agents can jump on or not.
Smooth UV/Curve Pin
Smooth UV/Curve Pin can control the “landing time” of agents when they work with CCT. This is a very important development for Ant-Man Film. It can make the UV/Curve pin more natural and smooth.
Geometry Logic
With geometry logic tool, we’re able to make the agent interactive with Maya models more precisely. We can have our agents align the normal, move along the surface of geometry, achieve the roll/pitch results.
2 Sides Road
We can create the road by more natural way. Each road can be defined and controlled by 2 sides, and we can easily create the road with arbitrary shapes/width for making them fit all different conditions.
Cloth Optimization & Acceleration
Cloth system has been optimized in this version, now it can collide with terrain ground, much faster when reset, and also faster to display.
Viewport 2.0 Full Support
Miarmy supports viewport 2.0 now, also, the speed of viewport is 1.5X faster than older viewport 1.0. Now from Miarmy 4.0, our viewport has already get about 10X speed up of the display.
FAQs
Tutorials
Miarmy Quick Start Tutorials #gettingstarted
This series of 5 tutorials will get you up and running quickly with Miarmy, the crowd simulation tool for Maya. The series covers character rig to OA, animation to action, export, simulation and render.
More...'%3e%3cg id='Final-Copy-2_2_' transform='translate(1275.000000, 200.000000)'%3e%3cpath class='st0' d='M7.4,12.8h6.8l3.1-11.6H7.4C4.2,1.2,1.6,3.8,1.6,7S4.2,12.8,7.4,12.8z'/%3e%3c/g%3e%3c/g%3e%3c/g%3e%3cg id='final---dec.11-2020'%3e%3cg id='_x30_208-our-toggle' transform='translate(-1275.000000, -200.000000)'%3e%3cg id='Final-Copy-2' transform='translate(1275.000000, 200.000000)'%3e%3cpath class='st1' d='M22.6,0H7.4c-3.9,0-7,3.1-7,7s3.1,7,7,7h15.2c3.9,0,7-3.1,7-7S26.4,0,22.6,0z M1.6,7c0-3.2,2.6-5.8,5.8-5.8 h9.9l-3.1,11.6H7.4C4.2,12.8,1.6,10.2,1.6,7z'/%3e%3cpath id='x' class='st2' d='M24.6,4c0.2,0.2,0.2,0.6,0,0.8l0,0L22.5,7l2.2,2.2c0.2,0.2,0.2,0.6,0,0.8c-0.2,0.2-0.6,0.2-0.8,0 l0,0l-2.2-2.2L19.5,10c-0.2,0.2-0.6,0.2-0.8,0c-0.2-0.2-0.2-0.6,0-0.8l0,0L20.8,7l-2.2-2.2c-0.2-0.2-0.2-0.6,0-0.8 c0.2-0.2,0.6-0.2,0.8,0l0,0l2.2,2.2L23.8,4C24,3.8,24.4,3.8,24.6,4z'/%3e%3cpath id='y' class='st3' d='M12.7,4.1c0.2,0.2,0.3,0.6,0.1,0.8l0,0L8.6,9.8C8.5,9.9,8.4,10,8.3,10c-0.2,0.1-0.5,0.1-0.7-0.1l0,0 L5.4,7.7c-0.2-0.2-0.2-0.6,0-0.8c0.2-0.2,0.6-0.2,0.8,0l0,0L8,8.6l3.8-4.5C12,3.9,12.4,3.9,12.7,4.1z'/%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
Your Privacy Choices Notice at Collection