FPV Rates Explained: Actual, Betaflight and Simulator Models

What do FPV rates control?

A rate model converts stick position into a rotational speed, often expressed in degrees per second. Center sensitivity, maximum rate and expo shape the response. Two profiles with similar-looking numbers can feel different when they belong to different models.

Three quantities are doing the work, and it helps to name them separately:

• Centre sensitivity — how much rotation you get for small stick movements near the middle. This is what “twitchy” refers to, and it is the value beginners usually need less of.
• Maximum rate — the rotation speed at full deflection. This determines how fast a flip or a roll can be, not how controllable it is.
• Expo — a curve that reduces response near centre while preserving the extremes. It lets you keep a high maximum rate without making small corrections violent.

Changing one of these and calling the result “rates” is the source of most setup confusion, because the same complaint (“too sensitive”) can be fixed by lowering centre sensitivity, adding expo, or lowering the maximum — with three different outcomes at the extremes.

Betaflight’s documentation explains the relationship between RCRate, Rate and Expo, and describes Actual, Quick, Betaflight, Raceflight and KISS models. It also records a useful piece of history: Actual rates arrived in 4.2 and became the default in 4.3, so a profile published after that point almost certainly assumes the Actual model unless it says otherwise. Read the official Rate Calculator guide for implementation details.

Why do the same numbers produce different results?

Because the models are different equations wearing similar labels. A “rate” of one value in an Actual model describes an angular velocity directly; in another model the same slider position might be a multiplier on a curve with an expo term folded in.

VelociDrone’s manual is explicit about the scale of this problem: it documents five rates types — Betaflight 4.2, QuickRates, Actual, KISS and FL1 — inside one simulator. Switching the type changes what every number on the screen means, without changing the numbers themselves.

So a profile is only a specification when it carries four things: the simulator, the rates type, the applicable version, and the flying style it was tuned for. A screenshot with numbers and no model is not a setting, it is a rumour.

What ranges do pilots actually use?

Betaflight’s own documentation gives typical maximum turn rates by discipline, and they are far apart:

• Racing: a fairly linear curve, with a maximum around 550–650 deg/s.
• Freestyle: a soft centre region combined with a high maximum, around 850–1200 deg/s.
• Cinematic: a flatter centre region, which makes small framing corrections less abrupt.

The gap between the racing and freestyle ranges explains a lot of bad advice. A freestyle pilot recommending “about a thousand” to a racer is not wrong — they are describing a different task. If you fly gates, a high maximum mostly buys faster recovery and more overshoot; if you fly lines for footage, a linear racing curve feels nervous.

The same page notes that it takes time to get used to a change in rates. That is worth quoting to anyone who evaluates a new profile in one flight: the first ten minutes of a new curve are not a fair test.

Why should simulator profiles stay separate?

A simulator may expose its own rate controls or emulate a real flight-controller model. The menu names, units and defaults are not automatically interchangeable. The settings database therefore keeps the simulator, flight style, skill level and rates type together.

There is a second, less obvious reason. Simulator implementations vary in how they smooth and sample input. Two simulators both claiming “Betaflight rates” can still feel different because the input pipeline differs, so a value that is comfortable in one may be sharp in the other.

Before changing numbers, complete controller setup and fly the default profile long enough to describe what feels wrong. The beginner guide recommends changing one variable at a time.

How should a beginner choose a starting point?

Keep the simulator default while learning the axes and Acro recovery. Then make one small change for one task, such as smoother center movement or a different maximum rotation. Record the simulator, model, version and reason for the change so the result can be repeated.

A sequence that avoids most dead ends:

1. Fly the default for two sessions. You need a baseline you can describe.
2. Diagnose in words. “Turns start too abruptly” is centre sensitivity or expo. “I cannot flip fast enough” is maximum rate. “It drifts in slow flight” is probably not rates at all.
3. Change one value by a small amount. Save it — several simulators require an explicit save, and an unsaved change looks exactly like a change that did nothing.
4. Fly the same course, same number of attempts. Compare judged control, not speed.
5. Revert if it did not help. A profile you cannot undo is a profile you cannot learn from.

What are the common traps?

• Copying a racer’s profile into a cinematic session. High maximum rates help gate entries and hurt smooth framing work.
• Tuning around a calibration fault. A drifting axis cannot be fixed with expo.
• Changing rates and camera tilt together. Both alter perceived speed, and the combination is unreadable.
• Believing a “universal” profile. Without the model and version, the numbers are meaningless.
• Ignoring the bank or profile slot. Simulators with multiple banks will happily let you tune a slot you are not flying.

Where do rates fit in the bigger picture?

Last, usually. Input mapping comes first, then camera tilt, then rates. A pilot who cannot hold a height has no way to judge whether a rate change helped.

If you are comparing simulators, read Liftoff vs VelociDrone with this in mind — the deeper settings documentation in one of them is only an advantage if you intend to use it.