Skip to content

Using Push Moco to Create a Move

CONFIRMED ANY RIG FLAIR 7

Not robot-specific

Push Moco is a general Flair 7 feature — it works the same way on any rig it's enabled for. The source videos for this guide were demonstrated on a Cinebot Mini (including track), but nothing here is Cinebot-specific.


What Push Moco Is

Push Moco lets you physically push/drive the rig by hand to teach keyframe positions, rather than typing in coordinates. The Push Moco button toggles a green (pushable) / red (locked) state — never try to push the rig while the light is red, this can trip the motors off. Turn it green, push it, turn it red again when you're not actively moving it.

You'll also see this same mode called teach mode or free drive in Flair's own interface and documentation — all three names (Push Moco, teach mode, free drive) refer to the same feature.

Check your payload before you enable it. If the payload doesn't match what's actually mounted, the rig can start moving on its own the moment Push Moco goes green — see the full danger warning in Cinebot Mini — Unboxing, Assembly & First Startup. With the payload set correctly, enabling it just makes the rig pushable by hand — it won't drift or move by itself.

On a Cinebot Mini running Flair Lite, Push Moco was demonstrated being toggled with a dedicated button on the physical hand controller itself, rather than a screen-only control — press it to go green (pushable), press again to go red (locked). See the Flair Lite (Cinebot Nano) — Interface Notes section below for more Lite-specific detail.


Basic Workflow — Creating a Move

  1. Enable sound feedback (optional but recommended): Tools > Sounds, enable application sounds for Waypoint modifications. This lets you store/delete waypoints by feel without watching the screen. (On Flair Lite this was demonstrated under Setups > Sounds instead — select Waypoint Modification, enable application sounds, Apply. Same feature, different menu.)
  2. Engage the robot first, using the main engage/disengage button — this turns every axis's motor on together in one sequence; you cannot engage a single axis individually from this control (on some rigs, e.g. Cinebot Mini, engaging takes a few seconds while it measures and releases brakes — this is normal, not a fault). Then put the rig into free mode and position it at your first frame.
  3. For focus: either pull focus manually and store it with each keyframe, or turn off any external hand-controller focus control and set Flair to Focus Follows Sensor (autofocus assist) so Flair focuses for you as you move — useful while roughing out a move, though see the note on Focus Follows Target below for the more accurate option once you're happy with a move.
  4. Press Store to save your first keyframe. Move to the next position, store again. Repeat for as many keyframes as you need (minimum two).
  5. Each new keyframe adds a default block of frames (e.g. 100 frames) between it and the previous one, at your configured frame rate.
  6. To run it: Forward Run. If Flair warns that certain axes (e.g. track, zoom, turntable) "are not tripped or turned on," you can dismiss it and run anyway, or — better — lock any axis you're not using (small lock icon per axis) so Flair ignores it entirely instead of warning every time. In plain terms: "tripped" means something has caused that axis to stop (e.g. a limit or fault), while "not turned on" just means you never engaged/powered that axis in the first place — the warning covers both cases with one message. A locked axis is not just silenced for the warning — Flair will not move it at all, during manual pushing or during a run, even if its stored value would otherwise put it out of range; unlock it before you need that axis to move again.
  7. Go To drives the rig from wherever it currently is to the move's start (or wherever you've scrubbed the playhead to) — but not along the programmed path. It takes the shortest route it can find, ignoring every keyframe in between. Shoot (sometimes labelled Play) is what actually runs the programmed move for real, once Go To has finished and the status shows ready to shoot.

Go To takes the shortest path, not your programmed path

Go To ignores every keyframe you've stored and drives straight to the target position by the shortest route it can find. If your rig is a long way from the move's start, or the direct line crosses an obstacle (or the rig itself), Go To can cause a collision on its own — separate from anything wrong with the move itself. If you're not sure what a Go To will do, slow it down first: Setups > Miscellaneous Setups > Go-To Speed, turn it down, Apply — then speed it back up once you trust the move. Flair also shows a "ready to shoot fast" warning before running a move that can move quickly at normal speed — don't dismiss it without checking what's around the rig first.

Record vs Rehearse: these run identically except for one thing — Record additionally turns the camera on automatically when you hit Shoot (if your rig has a camera on/off trigger cable connected) and off again at the end of the move. Everything else about how the move runs is the same.


Adjusting Timing and Speed

  • Typing a new frame-count directly into one keyframe's gap only changes that gap.
  • To shorten a gap and ripple the change through the rest of the move (keeping every other gap the same), type the number of frames to remove followed by a minus sign, e.g. 50-. This reduces that one gap by 50 frames without disturbing the others.
  • The same shortcut works in reverse: type a number followed by a plus sign, e.g. 40+, to lengthen the gap before the selected keyframe by that many frames — every later keyframe ripples forward with it, and no other gap changes.
  • To retime the whole move, use the Scale tool (top of the move panel) — change total frames, total seconds, or percentage, and every gap scales proportionally. Useful for "make this move 3 seconds instead of 6" without manually recalculating every keyframe.
  • Rounding caution: scaling very small gaps down can round to zero frames (there's no such thing as half a frame), which silently collapses that segment. Don't scale a move so far down that individual gaps get too small.
  • Making one gap very short relative to its neighbours can cause a large overshoot — Flair always tries to join keyframes with a continuous curve, and if it has to reach the next point in a very short time, it can overshoot well past it to keep the curve smooth. If a move suddenly does something violent, check whether a nearby gap has been made too tight. For example: if two keyframes are genuinely close together but still carry the default ~100-frame spacing, cutting that one gap down to something like 20 frames removes the overshoot without touching any other gap.
  • Prefer working in seconds instead of raw frame counts? Flair can display keyframe timing as time instead — see Flair Display Time Instead Of Frames.

Focus: Sensor vs Target

Two different "auto" focus behaviours are available while running a move:

  • Focus Follows Sensor — uses a distance sensor (e.g. LiDAR/focus assist) live during the move. Reactive, but not always fast enough to keep up with quick moves.
  • Focus Follows Target — uses the target distances you've already stored at each keyframe (via focus assist or manual entry) to calculate focus precisely along the path, without needing a live sensor read. More accurate for a move that's already been keyframed with correct focus data.

Switch between them by clicking the focus mode indicator and choosing the mode.


Adjusting an Existing Keyframe

  • Drag the timeline to any point along the move to preview that frame.
  • If you nudge the rig at that frame (e.g. reframing, or adjusting focus), you must re-store that keyframe or the change is lost — moving the playhead to a frame does not itself save anything.
  • You can store an entire keyframe (all axes) with the main Store button, or store just one axis's change by right-clicking that axis's individual field — useful when you only want to correct, say, lift, without touching everything else at that frame.
  • Mouse button conventions (three-button mouse): left-click selects, right-click stores a position (built in so the same controls work on touchscreens, which have no right-click), middle-click stops/locks an axis.
  • Browsing mode: click the browse/slider control to scrub continuously along the whole path (rather than jumping keyframe to keyframe) — useful for checking the move flows correctly end-to-end before committing.
  • Inserting a blank keyframe (rather than storing one) creates an empty slot with no position data in it — Flair skips it completely when running the move until you actually store a real position into it (e.g. right-click that keyframe > Store Current Axis Positions).
  • Deleting a keyframe: select it, then Delete — Flair asks you to confirm. The keyframe before it becomes the new end of that segment and the timeline reflows around the gap.

Track-Specific Notes (if your rig has one)

  • The track has a physical brake — release it before you can push the rig along the rail (small button near the track drive).
  • Watch the soft limit vs the physical buffer/magnetic limit: you can physically push the rig further than the software will allow it to run. If you store a keyframe near or past the soft limit, the move won't run until you either move that keyframe back or increase the configured limit.
  • Curves between keyframes are continuous — don't plan a point that would require the camera to pass through the arm or another obstacle on the way.
  • Coming out of teach/free mode does not automatically re-enable the track on every rig configuration — if the track won't move when you expect it to, check it's turned back on.
  • Watch the focus/target proximity warning: if two keyframes are measured a good distance from the subject but an in-between point on the curve passes very close to it, Flair may warn (or block running) because the calculated focus distance goes below the lens's minimum focus.

Kinematic Limits & Common Mistakes

The robot's kinematic solver (the maths that converts your camera path into individual axis moves) can fail or produce an unexpected result in a handful of predictable situations. None of these are bugs — they're geometry the arm genuinely can't (or can only awkwardly) solve — but they're easy to hit by accident while pushing the rig around freehand.

  • Don't fully extend the arm. Near full reach, Flair may not error outright — it can silently keep the move under budget by not actually reproducing the same motion you pushed it through, especially if you asked for something similar but slightly higher or further. If a stored move looks like it "worked" near full extension, don't assume it reproduced your exact path — check it in Rig View, and be more cautious the closer to full reach you're working.
  • Don't move the camera up and over while in Roll Level or Roll Relative. Both modes try to keep the camera level, and neither can pass through straight-up/straight-down (the ±90° tilt singularity) — going from underslung to overslung requires an instant flip that isn't physically possible. Flair may demand you reduce speed to 0 fps, or simply refuse to run cleanly. See Roll Control Modes — Complete Guide for the full picture; the two practical fixes are switching to Roll Up for that section of the move, or reprogramming with an added midpoint in Roll Relative so the camera doesn't have to flip in one step.
  • Don't let the target path pass very close to (or through) the camera itself when orbiting/moving around a subject with only two keyframes. Flair joins two keyframes with a straight line — if that line's closest approach to the subject gets very close to the camera's own position, the calculated pointing angle becomes extreme and the move looks (and runs) badly, getting worse the closer the path comes to passing through the camera. Fix it by adding an intermediate waypoint, or by switching the move type to Camera Orbital (a curved path around the target) instead of a straight interpolation.
  • Camera Orbital with only two keyframes can't be forced past 180°. Flair always takes the shortest route between two orbital keyframes — if you intended to travel more than 180° around the subject, it will go the other, shorter way instead, which may be the opposite direction from what you planned. Add an intermediate waypoint (e.g. roughly halfway around) so Flair has no shorter option than the direction you actually want.
  • Double-check Push Moco is off before using the on-screen directional controls to move around a target — it's easy to leave it enabled from a previous step and not notice, which can fight with the on-screen controls.

Flair Lite (Cinebot Nano) — Interface Notes

The core Push Moco workflow above is the same on Flair Lite — keyframes, Store, Forward Run/Rehearse, Go To, and the frame-timing shortcuts all work the same way. A few menu names and behaviours are worth knowing if you're on a Cinebot Nano (or a Mini/Max running Flair Lite instead of full Flair 7):

  • Estimate Target is reached via Job > Estimate Target rather than Move Effects > Estimate Target — pick which keyframes/waypoints to include, then Estimate. Same underlying feature as Estimate Target.
  • No physical feedback if you push past a reach limit. You won't feel resistance on the control when a keyframe goes beyond what the rig can run — the only warning is on screen: the arm turns red, and the move won't run while that keyframe stays out of range.
  • Widely-spaced keyframes without an intermediate point can plan an unreachable path. Even when neither keyframe alone is invalid, Flair may need to route through a position the rig can't physically reach (e.g. through itself) to get between two points that are far apart — the move loses its green "OK to run" check even though nothing looks wrong at either keyframe. Fix it by adding an extra keyframe partway between the two, giving Flair a valid point to route through, rather than trying to widen the limits.

Saving

File > Save As to name and save a job. Right-clicking the save button instead makes an incremental save (appends 001, 002, …) — useful for keeping a history of small changes without overwriting your last good version. Left-clicking the save/read icon reloads the last saved version, discarding unsaved changes.


YouTube 9B - Using Push Moco to create a move incl track Flair 7

▶ Track brake release, soft limits, curve-continuity/obstacle planning, incremental save

YouTube 11 - Adjusting moves & keyframes using Push Moco and the Flair 7 interface Flair 7

▶ Target measurement entry, the "N-" ripple-timing shortcut, the Scale tool, overshoot from over-tight timing


Source: MRMC video tutorials "9A - Using Push Moco to create a Move," "9B - Using Push Moco to create a move incl track," "11 - Adjusting moves & keyframes using Push Moco and the Flair 7 interface," "6 - Starting Flair & Overview," and "Cinebots Kinematic Do Nots" (Cinebot Mini/Max & Flair 7 Training playlist). The Flair Lite (Cinebot Nano) section is additionally sourced from "Flair Lite - PushMoco & Waypoint Timings" and "Flair Lite - Keyframes & Timeline" (Flair Lite & Nano Tutorials playlist).