Welcome to the Machinima FAQ! Here we’ll attempt to regularly answer frequently asked questions with Omniverse Machinima and the current status of the build.
How do I bring in my own character into Machinima?
Connector Workflow: Users can use an Omniverse Connector to export a USD skeletal mesh. UE4 and Maya currently have full USD skeletal mesh support on export. Install the Connectors from the Omniverse Launcher and simply export your asset. Specific tutorials on this in relation to Machinima will be provided soon.
Import Workflow: Users can import FBX files directly into Omniverse Machinima by right-clicking on the content browser and choosing the import option. This allows for skeletal mesh assets that already exist in FBX format to be brought into Machinima. This workflow can also be used to import animations and supports basic material creation.
How do I use pose-estimation on my own character in Machinima?
- Machinima ships with 5 retargeting scripts. They are located here: (application)\exts\wrnch.engine.wrx_decoder\wrnch\engine\wrx_decoder\ and are labeled “wrx_xform_*.json. Users can generate skeletons with the same joint names as one of the existing rigs, or they can create a new *.json script mapped with their joint names.
How do I bring my own animations into Machinima?
- Connector Workflow: Users can use an Omniverse Connector to export a USD animation clip. UE4 and Maya currently have animation support on export. Install the Connectors from the Omniverse Launcher and simply export your animation asset. Specific tutorials on this in relation to Machinima will be provided soon.
- Import Workflow: Users can import FBX files directly into Omniverse Machinima by right-clicking on the content browser and choosing the import option. This allows for animation assets that already exist in FBX format to be brought into Machinima.
What type of effects can I do in Machinima?
- Currently, the main effect that is available is Omniverse FLOW. It is a real-time simulation of smoke and fire. A traditional particle system will be available at a later time.
How do I add audio to my video?
- For now, Machinima does not export audio with its movie capture function. Users will need to use a third party application to composite the video from Machinima together with their own audio tracks.
- In the sequencer, when a camera is added and a keyframe is dragged past the length of the clip, the keyframe will not be displayed until the clip is lengthened to include the keyframe.
- When using multiple constraints on a camera,there may be slowdowns in the stage or playback. The path tracer may also get stuck at 1 spp. However, capturing video will render at the correct samples per pixel setting.
Sequencer - Tracks
- If users add an animation to a track without setting its target (the asset it’s animating on), a target will have to be manually assigned. The track doesn’t set this automatically.
Time Samples - Elongating Clips
- If users animate an object with the time sample editor and then duplicate its clip, they will notice an increasingly expanding clip size due to time samples reading the entire scene. To avoid this, animate on an Xform and then apply that Xform to the object as an animation in the sequencer. (VIDEO)
- Currently, if users run out of VRAM, they may crash. Systems with low VRAM GPUs should be extra careful and pay attention to their VRAM usage in the task manager.
- We do see some rare crashes browsing the content browser tree. We hope to have updates on this soon.
- A2F Cache needs to be rehooked occasionally, some times you’ll see your A2F cache doesn’t play, you need to simply click on the a2fmesh in the stage and click on the clip and “Use selected prim as target”
- When rendering video, users may experience a “cropped” view of the scene. Please ensure the viewport resolution is the same as your render resolution and re-enable reshade. They must match in the current version.
- To achieve perfect looping of duplicated animations, users must ensure that the start and end keyframes of the animation are identical.