Watch the creator’s explanation
just7 · Eclipse — Official Tutorial · Graph editor at 1:06. This is the creator’s video, accompanying our independent written guide.
Loading the player connects to YouTube. It is a separate choice from optional website Analytics. You can read every step without loading the video.
No player is loaded. Video privacy details.
Start with the documented Eclipse Animator graph editor
The official Eclipse listing names a Graph Editor among its features. The creator covers graphs at 1:06 in the official tutorial. Use that chapter for the plugin’s interface and the current installed version for the controls available to you.
This lesson offers a small exercise for understanding a transition, rather than a guessed inventory of Eclipse tangent buttons, interpolation modes or curve colors. Our general examples do not establish that the plugin exposes a particular mathematical curve by name. Hands-on plugin testing has not been completed.
If you cannot already create two poses and return to their saved keyframes, complete the beginner animation lesson first. If you still need Eclipse, verify the asset through Get Plugin and use the Studio installation guide. Graph editing is easier to understand after the base movement has a clear start and finish.
What a graph can explain about a transition
Keyframe positions answer when important states occur. Their values describe those states. The interpolation between them determines the states you see at intermediate times. Two animations can share the same start pose, end pose and duration while looking different during the movement.
Roblox’s Curve Editor documentation gives background on plotting and adjusting animation values over time. It describes Roblox’s built-in editor, not Eclipse’s interface. Its buttons, rotation conversions and interpolation options must not be copied into plugin instructions without separate evidence.
Before interpreting any graph, identify what the selected track represents. A vertical change might describe an angle, position or another property; it is not automatically a speed reading. Likewise, an attractive-looking curve is not proof of an attractive movement. Always compare the graph with the object in the viewport.
Prepare one movement for the Eclipse Animator graph editor
Choose a short, readable change with two known endpoint poses. For this practice, use one selected part moving in a single clear direction or rotating through a modest range. Save a baseline copy before editing its transition. Leave the camera stationary and avoid adding layers or IK changes during the same comparison.
Choose a duration and use the actual project FPS to locate its endpoint. For example, a one-second interval spans 24 frame intervals at 24 FPS, 30 at 30 FPS or 60 at 60 FPS. The animation timing calculator helps you plan this consistently. These are arithmetic examples, not statements about Eclipse’s default settings.
Scrub the baseline at the start, one quarter, halfway, three quarters and the end. Describe the visible result in plain language: “Most of the change happens early,” or “The movement is similar throughout.” Record that observation before editing so that you have something more precise than “make it smoother.”
Follow the graph chapter and change one thing
Open the creator’s graph editor chapter at 1:06. Match the selected item and graph context to your own simple project. Identify the demonstrated control before attempting a change. If your version has a different layout, do not infer an unknown button from a similarly shaped icon.
Keep both endpoint times and poses unchanged for your first comparison. Adjust one supported transition control from the demonstrated workflow, then inspect the same five moments you checked in the baseline. You are trying to isolate the distribution of motion within the interval; moving a keyframe to a new time would also change timing and introduce another variable.
After one adjustment, preview at normal speed. Ask whether the movement starts, travels and arrives the way you intended. If you cannot describe a useful difference, compare with the baseline again before making another change. A collection of small unexplained edits is difficult to reproduce or reverse confidently.
A numerical easing example, separate from plugin settings
Imagine a value that goes from 0 to 1 over one second. For elapsed fraction t, a linear example is p = t. A smooth-start-and-finish mathematical example is p = 3t² − 2t³. The latter is often called smoothstep; it is an illustration here, not a claim that Eclipse has a preset with that name.
| Elapsed fraction of the interval | Linear progress | Smoothstep example |
|---|---|---|
| 0 | 0% | 0% |
| 0.25 | 25% | 15.625% |
| 0.5 | 50% | 50% |
| 0.75 | 75% | 84.375% |
| 1 | 100% | 100% |
Both examples agree at the endpoints and midpoint. Their quarter points differ, which is why checking only the first, middle and final pose can miss the effect of easing. This table describes a normalized mathematical value. It does not predict how a particular three-dimensional rotation or Eclipse channel will travel in world space.
Use the example to ask a better viewing question: should this motion leave gently, move quickly through the middle, or arrive sharply? There is no universal requirement to ease every movement the same way. An intentional mechanical motion and a relaxed gesture may need different transitions.
Troubleshoot the motion before adding more curves
The endpoint changed when you wanted different easing
Return to your baseline and compare the selected keyframe’s time and value. A pose change and a transition change are different operations. Do not compensate for an accidentally moved endpoint by adding several new keys around it; first establish which part of the intended animation changed.
The graph looks smooth but the motion still looks wrong
Check the selected channel, neighboring joints and other active motion. A smooth property curve does not ensure a believable full-body result. Keep the test small enough to distinguish an arm’s local movement from a torso or whole-rig movement affecting its visible path.
Playback seems faster after the edit
Verify the endpoint frame positions and playback context. Then compare how much of the change happens early or late in the interval. Concentrating motion into a smaller portion can make it feel faster even when the total duration has not changed. Use the same duration and viewing angle for a fair comparison.
Finish the graph exercise with a recoverable result
Keep the baseline and your preferred version. Record the selected property, endpoint times and intended difference in one sentence. Save the edited project, return to it and inspect the same comparison range. If you export, test that result separately through the export workflow.
Use the controls reference for contextual keyframe and transform actions. Continue to animation layers when you can explain this single transition and want to investigate a separate motion contribution. The tutorial directory links the other tasks without requiring you to combine every feature in one project.