Skip to content

Setting Up Axis Acceleration

[BOTH] Finding the maximum safe acceleration for a rig axis requires a methodical procedure using Flair's velocity and acceleration graphs. This is used when commissioning a new rig, reconfiguring after a hardware change, or when optimising performance for a demanding move. The procedure is described by @peter.

In plain language, this procedure deliberately increases how hard one motor accelerates until its protection system stops it, then reduces the setting to leave a safety margin. A trip is that protective stop. A flat top on the speed graph means the motor reached full speed long enough for the test to be useful. This is a commissioning procedure, not a routine way to make one shot faster.

Do not perform this test without commissioning authority

The axis is intentionally driven to its trip point. Clear the full travel area, keep the E-stop in hand, test only one axis at a time, and do not alter an unfamiliar rig's acceleration limits without approval from the person responsible for that system.

Sufficient travel required

This procedure requires enough physical travel on the axis to reach top speed and hold it for a constant-speed section. If the axis travel is too short, you cannot reach max speed and the measurements will not be valid.

Where these values actually live

The position, velocity, and acceleration limits this procedure tunes are three separate fields per axis, all under Setups > Axis Setup > [axis] > Limits. On a correctly commissioned rig these should already be set — this procedure is for commissioning or reconfiguration, not something to casually adjust on a rig that's already working. If Flair blocks a move as over-velocity or over-acceleration on hardware you didn't commission yourself, don't assume the fix is to just raise the limit — it's possible the stored value is simply wrong for that hardware (e.g. after connecting to something new Flair hasn't seen before), but it's also possible the limit is correct and the move genuinely needs to be slower.


What "Fairings" Actually Control

A fairing (also called an "ease") is the ramp portion of a move — the section spent accelerating up to speed or decelerating back down — around a waypoint. On a simple two-point move this is straightforward: the axis ramps up, holds a constant-speed section, then ramps down. This is exactly the shape this procedure tests: how hard that ramp (the fairing) can be pushed before the axis trips.

If you build a multi-waypoint test move instead of a simple two-point move, be aware that fairings only take effect at a waypoint where the path actually changes direction. Flair joins waypoints with a smooth spline, so a waypoint that is just "on the way" — motion continues through it in the same direction — gets no fairing effect at all, and adjusting its fairing percentage will make no visible difference to the curve. Only a genuine reversal point behaves like the fairing described above.

At a genuine change of direction, the Prevent Curve Reversal checkbox controls what happens next:

  • Checked (default): the axis comes to a complete stop at the waypoint, then accelerates away in the new direction. This is the behaviour this procedure assumes.
  • Unchecked ("allow curve reversal"): Flair instead fits a continuous spline through the waypoint, which overshoots past the point before curving back — the axis never fully stops. This can look smoother on camera but is a different motion profile than a hard reversal, and is not what this test procedure is measuring.

Fairings near zero — pre-roll and post-roll

If fairings are reduced toward 0% so the recorded frame range is pure constant speed with no visible ramp, the axis physically cannot be at full speed exactly at frame 0. Flair compensates by running the axis backwards before frame 0 to build up to speed — a pre-roll — and, symmetrically, continuing past the final frame to slow back down — a post-roll — so the programmed frame range is at true constant speed throughout. This is a second reason the "sufficient travel" warning above matters: the axis needs physical travel beyond the nominal frame range to perform the pre-roll and post-roll. During a run, Flair's status line will read "pre-rolling" and "post-rolling" while this happens — this is expected behaviour, not a fault.


Procedure

Step 1 — Initial setup

  1. Set the camera/move frame rate to something high — 200 fps is recommended.
  2. Set the acceleration figure to something average (not high — you are starting cautiously).
  3. Set Runtime Factor to 1.0.
  4. Set Target Tracking (TT) off.
  5. Set axis fairings to approximately 40%.

Step 2 — Confirm max speed is reached

  1. Run the move.
  2. Open the axis graph — turn on Speed graph, turn off Position graph.
  3. Use the cursor to check whether the axis reached and held a constant top speed (a flat top on the speed curve).
  4. If max speed was not reached, reduce the number of frames (or run the move faster) until Flair finds the maximum speed for you with fewest frames.

Let Flair find max speed

The easiest way to confirm top speed is reached is to use too few frames and let Flair report the maximum speed achievable. You should see a flat plateau at the top of the speed graph.

Step 3 — Read peak acceleration

  1. Turn off the Velocity graph and turn on the Acceleration graph.
  2. Use the cursor to read the peak acceleration value at the ramp-up.

Step 4 — Increase acceleration until the axis trips

  1. Reduce the fairings gradually.
  2. As fairings decrease, Flair will eventually increase the "fastest" frame count — complaining that acceleration is too high.
  3. When this happens, increase the acceleration limit in the axis setup.
  4. Run the move again and read the new peak acceleration from the graph.
  5. Continue reducing fairings and increasing acceleration until the axis trips on the ramp-up (not at top speed).

Trip on the ramp-up, not at speed

The goal is to find the acceleration limit at which the axis trips during the acceleration phase, not while running at speed. This tells you the true mechanical acceleration ceiling.

Step 5 — Back off and set Runtime Factor

  1. Back off the acceleration limit by approximately 5% below the value at which the axis trips.
  2. Run the move several times to confirm it is repeatable and not a fluke.
  3. Set Runtime Factor to 0.95 as a safety margin for production use.

Bolt Fast-Stop Post-Roll

If a fast faired stop on the Bolt causes a small post-roll (a slight overrun after stopping), try one of:

  • Add a second keyframe identical to the end point, held for at least 1/10th the move length (e.g. 100 frames hold at 1000 fps).
  • Use the Smoother Stop option in Miscellaneous Setups.
  • Use the Firm Stop option.

The double-keyframe approach is the most reliable. @Niko confirms both "add a hold" and "firm stop" options work well in practice.


Summary of Settings

Setting Starting value Purpose
Move FPS 200 High rate reveals limits quickly
Acceleration Average Conservative starting point
Runtime Factor 1.0 → 0.95 final 1.0 for testing; 0.95 for production
Fairings ~40% Starting fairing; reduce during testing
TT Off Isolate axis behaviour


Source: Setting up Acceleration on an Axis — talk.mrmoco.com