A character moves from the left side of a screen to the right. If every frame moves the character by exactly the same distance, the motion looks steady. But what if the designer wants the character to begin gently, accelerate, slow down and stop smoothly?
That request can be translated into Mathematics.
Position is a function of time. Velocity describes how quickly the position changes. Acceleration describes how the velocity changes. By choosing an appropriate function, we can shape the movement rather than simply specify where it starts and ends.
This is a reason to study Additional Mathematics that many students may not expect: functions and calculus help explain the motion behind animation, interfaces and interactive graphics.
The examples below use hypothetical animation coordinates and simplified timing models. They are educational rather than professional game-engine instructions, and mathematical smoothness does not by itself guarantee comfort, accessibility, hardware performance or physical safety.
The useful skill is learning how algebraic conditions determine a curve and how derivatives reveal the movement it creates.
Start with two endpoints
Suppose a screen object begins at horizontal position:
x = 0 pixels.
It should finish at:
x = 120 pixels.
The motion should last:
T = 4 seconds.
These conditions tell us where the object begins and ends.
But they do not tell us what happens between those moments.
Many different functions satisfy:
x(0) = 0,
and:
x(4) = 120.
Some produce constant speed.
Some accelerate.
Some slow down.
Others overshoot and return.
This is an important modelling idea.
Two endpoint values alone do not uniquely determine a path through time.
We need more conditions if we want to control how the animation feels and behaves.
The simplest choice is linear motion
Assume position changes at a constant rate.
The linear model is:
x(t) = 30t,
for 0 ≤ t ≤ 4.
At t = 0:
x(0) = 0.
At t = 1:
x(1) = 30.
At t = 2:
x(2) = 60.
At t = 3:
x(3) = 90.
At t = 4:
x(4) = 120.
The object covers thirty pixels every second.
The graph of position against time is a straight line.
Its gradient is constant.
This model is mathematically simple and appropriate whenever steady movement is intended.
Differentiate the linear position
For:
x(t) = 30t,
differentiate:
v(t) = dx/dt = 30.
The velocity is thirty pixels per second.
Differentiate again:
a(t) = dv/dt = 0.
The acceleration is zero throughout the open interval of motion.
This does not mean the object never moves.
It means its velocity remains constant within that interval.
However, if the object is stationary before t = 0 and immediately moves at thirty pixels per second at t = 0, its velocity jumps at the start.
If it then stops instantly at t = 4, its velocity jumps again at the end.
Those changes may appear abrupt in some visual contexts.
A different function can make velocity start and finish at zero.
Define a normalised time variable
A convenient technique is to separate the total distance and duration from the shape of the motion.
Let:
u = t/T.
For our example:
u = t/4.
During the movement:
0 ≤ u ≤ 1.
Now define a shape function s(u) satisfying:
s(0) = 0,
and:
s(1) = 1.
The position is:
x(t) = Ls(u),
where L is the total displacement.
For this example L = 120.
This representation makes it easy to compare different easing curves.
Every curve starts at zero and ends at one in normalised coordinates.
The scale L determines the actual distance, while T determines the time duration.
What does “ease-in” mean mathematically?
A basic ease-in function is:
s(u) = u².
Then position becomes:
x(t) = L(t/T)².
For L = 120 and T = 4:
x(t) = 120(t/4)².
Simplify:
x(t) = 7.5t².
Differentiate:
v(t) = 15t.
Velocity starts at zero and increases linearly during the motion.
At t = 4:
v(4) = 60 pixels per second.
The object accelerates from rest.
But it does not finish at rest.
If the position is clamped at 120 pixels after four seconds, the velocity drops suddenly to zero at the endpoint.
Ease-in solves the starting-velocity problem but not the stopping-velocity problem.
What does “ease-out” mean?
A basic ease-out shape is:
s(u) = 2u – u².
It satisfies:
s(0) = 0,
s(1) = 1.
Differentiate with respect to u:
s'(u) = 2 – 2u.
The normalised velocity starts at two and decreases to zero.
For our actual position:
x(t) = 120[2(t/4) – (t/4)²].
Simplify:
x(t) = 60t – 7.5t².
The velocity is:
v(t) = 60 – 15t.
At t = 0, velocity is 60 pixels per second.
At t = 4, velocity is zero.
The object slows into its destination, but starts with an immediate non-zero velocity.
Different easing choices solve different endpoint requirements.
What does ease-in-out require?
Suppose we want a model that begins at rest, moves to the final position, then finishes at rest.
We need four conditions:
x(0) = 0.
x(T) = L.
x'(0) = 0.
x'(T) = 0.
The first two specify the endpoints.
The next two specify zero velocity at those endpoints.
A straight line cannot generally satisfy all four while travelling a positive distance.
A carefully chosen cubic polynomial can.
This is a useful mathematical design problem because it shows how several conditions determine coefficients.
Students can derive a function instead of copying one from an animation library.
Construct a cubic ease-in-out function
Use the normalised form:
s(u) = au³ + bu² + cu + d.
We want:
s(0) = 0,
s(1) = 1,
s'(0) = 0,
s'(1) = 0.
Begin with s(0) = 0.
This gives d = 0.
Differentiate:
s'(u) = 3au² + 2bu + c.
At u = 0, the derivative condition gives c = 0.
Now the function is:
s(u) = au³ + bu².
The remaining two conditions determine a and b.
Solve for the cubic coefficients
From s(1) = 1:
a + b = 1.
From s'(1) = 0:
3a + 2b = 0.
Use the first equation:
b = 1 – a.
Substitute into the second:
3a + 2(1 – a) = 0.
Simplify:
3a + 2 – 2a = 0.
Therefore:
a = -2.
Then:
b = 3.
The cubic easing function is:
s(u) = 3u² – 2u³.
It satisfies all four endpoint conditions.
This curve is often called a cubic smoothstep in computer graphics.
The name is less important than understanding how the conditions produce the coefficients.
Verify the endpoint conditions
At u = 0:
s(0) = 0.
At u = 1:
s(1) = 3 – 2 = 1.
Differentiate:
s'(u) = 6u – 6u².
Factorise:
s'(u) = 6u(1 – u).
At u = 0:
s'(0) = 0.
At u = 1:
s'(1) = 0.
Thus the cubic starts and finishes with zero derivative in normalised time.
The curve moves from zero to one and has stationary endpoints.
We have verified both position and velocity requirements.
This is exactly the sort of mathematical reasoning that makes a formula more meaningful than an unexplained preset.
Apply the cubic to our animation
Use:
L = 120 pixels,
T = 4 seconds,
u = t/4.
The position function is:
x(t) = 120[3(t/4)² – 2(t/4)³],
for 0 ≤ t ≤ 4.
At t = 0, x = 0.
At t = 4, x = 120.
Because the easing derivative vanishes at both ends, the velocity begins and finishes at zero.
The motion is not constant-speed.
It accelerates during the first half and decelerates during the second.
To see this precisely, differentiate the function with respect to time, remembering that u itself depends on t.
The chain rule is essential.
Differentiate with the chain rule
Write:
x(t) = Ls(u),
where:
u = t/T.
Then:
dx/dt = L(ds/du)(du/dt).
Since:
du/dt = 1/T,
we obtain:
v(t) = (L/T)s'(u).
For cubic smoothstep:
s'(u) = 6u – 6u².
Therefore:
v(t) = (L/T)(6u – 6u²).
For L = 120 and T = 4:
L/T = 30.
So:
v(t) = 30(6u – 6u²).
Factorise:
v(t) = 180u(1 – u) pixels per second.
This expression explains the motion’s changing speed.
Find the maximum speed
The velocity function on 0 ≤ u ≤ 1 is:
v(u) = 180u(1 – u).
Expand:
v(u) = 180u – 180u².
It is a downward-opening quadratic.
Complete the square or differentiate.
Its maximum occurs at:
u = 1/2.
Since u = t/4:
t = 2 seconds.
At u = 1/2:
v = 180(1/2)(1/2) = 45 pixels per second.
The maximum eased speed is 45 pixels per second.
The linear animation’s constant speed was 30 pixels per second.
The eased motion must travel the same 120 pixels in the same four seconds while spending time moving slowly near the endpoints.
Therefore it moves faster during the middle portion.
Calculate the position after one second
At t = 1:
u = 1/4.
The cubic position is:
x = 120[3(1/4)² – 2(1/4)³].
Calculate:
3(1/16) = 3/16.
And:
2(1/64) = 1/32.
Subtract:
3/16 – 1/32 = 6/32 – 1/32 = 5/32.
Therefore:
x = 120(5/32) = 18.75 pixels.
The linear animation would have reached 30 pixels.
The eased animation has moved less during the first second because it begins slowly.
This is a numerical way of describing the visual effect without depending solely on an animation preview.
Calculate the positions at the midpoint and third second
At t = 2:
u = 1/2.
Compute:
s(1/2) = 3(1/4) – 2(1/8).
So:
s(1/2) = 3/4 – 1/4 = 1/2.
Therefore:
x(2) = 60 pixels.
At t = 3:
u = 3/4.
Then:
s(3/4) = 3(9/16) – 2(27/64).
This equals:
27/16 – 54/64 = 108/64 – 54/64 = 54/64 = 27/32.
Therefore:
x(3) = 120(27/32) = 101.25 pixels.
The positions are symmetric around the midpoint in this particular cubic easing model.
Why the curve’s symmetry matters
For cubic smoothstep:
s(u) = 3u² – 2u³.
It satisfies:
s(1 – u) = 1 – s(u).
Therefore:
s(u) + s(1 – u) = 1.
For a movement of length L, the positions at complementary times add to L.
Our positions after one and three seconds are:
18.75 + 101.25 = 120.
The position at two seconds is 60.
This symmetry helps check calculations.
But not every easing curve must be symmetric.
A designer may want faster acceleration and slower deceleration or the reverse.
Mathematics allows those shapes to be specified through different conditions.
The useful skill is understanding what properties the chosen function actually has.
Acceleration explains the shape of the velocity
Differentiate velocity again.
We have:
v(t) = (L/T)(6u – 6u²).
Since du/dt = 1/T:
a(t) = (L/T²)(6 – 12u).
For L = 120 and T = 4:
L/T² = 120/16 = 7.5.
Therefore:
a(t) = 7.5(6 – 12u).
Simplify:
a(t) = 45 – 90u pixels per second squared.
At the beginning u = 0:
a = 45.
At the midpoint u = 1/2:
a = 0.
At the end u = 1:
a = -45.
The acceleration changes from positive to negative as the motion switches from speeding up to slowing down.
Zero endpoint velocity does not mean zero endpoint acceleration
Our cubic ease has:
v(0) = 0,
and:
v(4) = 0.
But the acceleration just inside the motion interval is:
a(0) = 45,
and:
a(4) = -45 pixels per second squared.
If the object is stationary before and after the four-second animation, the acceleration outside the active interval may be zero.
Clamping the position to a constant before and after creates an acceleration discontinuity at each boundary.
So cubic smoothstep gives continuous position and velocity at the joins, but not necessarily continuous acceleration.
This is an important distinction.
A smooth-looking position function can still have abrupt changes in its higher derivatives.
For physical mechanisms, these differences may matter considerably.
A smoother acceleration option: quintic smoothstep
A fifth-degree polynomial can meet more endpoint conditions.
One widely used normalised easing curve is:
s(u) = 6u⁵ – 15u⁴ + 10u³.
For 0 ≤ u ≤ 1, it satisfies:
s(0) = 0,
s(1) = 1,
s'(0) = s'(1) = 0,
and:
s”(0) = s”(1) = 0.
The extra conditions give zero acceleration at the active interval endpoints as well as zero velocity.
The function can therefore join stationary sections with continuity in position, velocity and acceleration under the simple clamping model.
This is a more advanced illustration of how additional mathematical conditions may require a higher-degree polynomial.
Differentiate the quintic
Take:
s(u) = 6u⁵ – 15u⁴ + 10u³.
Its first derivative is:
s'(u) = 30u⁴ – 60u³ + 30u².
Factorise:
s'(u) = 30u²(1 – u)².
This is non-negative throughout 0 ≤ u ≤ 1.
It is zero at both endpoints.
Therefore the position function moves forwards without reversing direction on the normalised interval.
Differentiate again:
s”(u) = 120u³ – 180u² + 60u.
Factorise:
s”(u) = 60u(1 – u)(1 – 2u).
At u = 0 and u = 1, the acceleration factor is zero.
The polynomial’s derivative structure verifies the promised endpoint behaviour.
Compare cubic and quintic curves
Both cubic and quintic smoothstep start at zero and finish at one.
Both have zero endpoint velocity.
Both are increasing throughout the active interval.
But their acceleration behaviour differs at the endpoints.
The cubic has non-zero one-sided acceleration at the ends of the active motion.
The quintic has zero acceleration there.
Neither automatically guarantees the most comfortable or appropriate motion for every interface or mechanical system.
A real animation may depend on frame timing, visual context, perceptual response and accessibility settings.
The mathematical lesson is that “smooth” can refer to different derivative conditions.
Students should specify which feature they want rather than assume every smooth curve has the same properties.
Another option: cosine-based ease-in-out
Consider:
s(u) = [1 – cos(πu)]/2,
for 0 ≤ u ≤ 1.
At u = 0:
cos 0 = 1,
so s(0) = 0.
At u = 1:
cos π = -1,
so s(1) = 1.
Differentiate:
s'(u) = (π/2)sin(πu).
At both endpoints, sine equals zero.
Therefore the motion also begins and finishes at zero velocity.
This easing function comes from trigonometry rather than an ordinary low-degree polynomial.
It shows why periodic functions can be useful even in an animation that is not intended to repeat.
The cosine easing speed
For:
x(t) = L[1 – cos(πt/T)]/2,
differentiate:
v(t) = (πL/2T)sin(πt/T).
For L = 120 and T = 4:
v(t) = 15π sin(πt/4).
At the midpoint t = 2:
sin(π/2) = 1.
So the maximum speed is:
15π pixels per second.
Numerically:
15π ≈ 47.12 pixels per second.
This is slightly greater than the cubic smoothstep maximum of 45 pixels per second for the same distance and duration.
Both have the same average speed:
120/4 = 30 pixels per second.
They distribute that movement differently over time.
Mean speed is fixed by distance and duration
For a movement from x = 0 to x = L completed in time T without changing direction, average speed is:
L/T.
Our animation has:
L = 120 pixels,
T = 4 seconds.
Therefore average speed is 30 pixels per second.
That is true for the linear, cubic, quintic and cosine easing functions when each completes the same monotonic move over the same duration.
But instantaneous speed can vary greatly.
The animation designer is changing the timing profile, not the total displacement.
The definite integral of velocity over the interval returns the net movement:
∫ from 0 to T of v(t) dt = L.
This is an important connection between animation and calculus.
Check the cubic movement through integration
For the cubic model:
v(t) = (L/T)(6u – 6u²),
where u = t/T.
Because dt = T du, the total displacement is:
∫ from 0 to T of v(t) dt = L∫ from 0 to 1 of (6u – 6u²) du.
An antiderivative is:
3u² – 2u³.
Evaluate at 1 and 0:
(3 – 2) – 0 = 1.
Therefore the total displacement is:
L × 1 = L.
For our example, it equals 120 pixels.
This verifies that the easing function’s changing velocity still produces exactly the intended net movement.
A-Math provides tools for checking a function at more than one level.
Sample a continuous function into frames
A mathematical curve provides a position at every real time in the interval.
A rendered animation is displayed through a sequence of frames.
If a four-second animation is sampled at sixty frames per second, the time step is:
Δt = 1/60 second.
There are 240 intervals of that duration across four seconds.
If both endpoints are explicitly included in a sequence of uniformly spaced timestamps, the samples can include 241 time points from t = 0 through t = 4.
The exact frame scheduling depends on the implementation.
At each sample time, the animation system can evaluate x(t).
The function defines the intended position; the rendering process determines how that position is displayed.
These are related but different layers.
Compare the frame movement near the midpoint
For the linear animation, velocity is constant at:
30 pixels per second.
Over a short interval of 1/60 second, the displacement is:
30/60 = 0.5 pixel.
For cubic ease near its midpoint, instantaneous velocity is:
45 pixels per second.
A local estimate of one-frame displacement there is:
45/60 = 0.75 pixel.
These are mathematical values.
The actual finite displacement for the easing curve changes slightly across the frame because velocity is not perfectly constant.
A display may render fractional-pixel positions through interpolation or other methods.
The number of visible screen pixels moved from frame to frame depends on the rendering implementation.
This is why a correct motion function and its visual appearance are not identical questions.
High frame rate is not the same as mathematical smoothness
Suppose a position function has a sudden jump at one time.
Sampling it more frequently does not remove the discontinuity.
It may simply show the jump at a more precisely located frame.
Conversely, a smooth mathematical function may still appear uneven if frames are delayed, dropped or displayed inconsistently.
A-Math can describe continuity, velocity and acceleration in the ideal function.
It does not by itself guarantee smooth performance on a particular device.
Students should distinguish the mathematical path from the hardware and software used to render it.
That is a useful habit whenever Mathematics is applied to technology.
A bouncing motion requires more care
Imagine a ball dropping, hitting the floor and bouncing.
A simple piecewise mathematical model might use one quadratic formula before impact and another after.
At the impact time, position should ordinarily remain continuous.
But vertical velocity can change direction abruptly in an idealised instantaneous-collision model.
A smoothstep easing curve does not automatically represent the physics of a bouncing ball.
To model a real bounce, we need assumptions about gravity, collision response and energy loss.
The mathematical functions must fit the intended process.
This is an important modelling lesson: visually pleasing motion and physically motivated motion are not necessarily the same.
The correct choice depends on the question.
Periodic motion uses trigonometric functions differently
Suppose an icon is meant to move gently up and down repeatedly.
An illustrative model is:
y(t) = 100 + 10sin(πt),
with y measured in screen-coordinate units and t in seconds.
The centre is y = 100.
The amplitude is 10.
The period is two seconds because sin(πt) repeats when t increases by 2.
At t = 0, y = 100.
At t = 0.5, y = 110.
At t = 1, y = 100.
At t = 1.5, y = 90.
At t = 2, y = 100.
The function describes repeating oscillation rather than one movement that stops permanently at an endpoint.
Differentiate the periodic animation
For:
y(t) = 100 + 10sin(πt),
differentiate:
y'(t) = 10π cos(πt).
The derivative describes vertical coordinate velocity.
At t = 0:
y'(0) = 10π.
At t = 0.5:
y'(0.5) = 0.
At t = 1:
y'(1) = -10π.
The direction of positive movement depends on the coordinate convention. Some screen systems count y increasing downwards, while Cartesian diagrams usually take upward as positive.
The mathematical sign is meaningful only relative to the defined axis.
A-Math teaches students to make that convention explicit before interpreting a derivative physically.
One-shot easing is not the same as periodic motion
A one-shot smoothstep function is intended for u from zero to one.
It begins at a chosen position and ends at another.
Outside the interval, an implementation may clamp the function to the appropriate endpoint value.
A sine wave repeats indefinitely under its mathematical definition.
That is useful for oscillation, but not necessarily for an object that should move once and stop.
Both use changing positions.
Their time behaviours are different.
A student should decide whether the task needs a finite transition or ongoing periodic movement before choosing a function.
The domain and boundary behaviour are part of the design.
This is another example of method selection guided by the actual target.
An easing curve can overshoot
Not every animation function must stay between zero and one.
Some effects intentionally move beyond the destination before returning.
For example, one might specify an intermediate position greater than L while keeping the same start and end.
Such a curve may require additional polynomial terms or a suitable oscillatory expression.
But overshoot can be undesirable when the position represents a constrained property.
A user-interface element might not be allowed to move outside a visible region.
A real mechanism might have physical travel limits.
A mathematical curve should be checked against these constraints.
The point is not that overshoot is inherently good or bad.
It is that the function’s range must fit the permitted movement.
A simple domain and range audit
For cubic smoothstep:
s(u) = 3u² – 2u³,
with 0 ≤ u ≤ 1,
we have:
s'(u) = 6u(1 – u) ≥ 0.
Therefore s is non-decreasing on the interval.
Since s(0) = 0 and s(1) = 1, every output lies between zero and one.
Thus:
0 ≤ s(u) ≤ 1.
For L > 0, the position x = Ls(u) stays between zero and L.
This is a useful guarantee under the model.
It means the cubic does not overshoot its destination over the active interval.
We derived that property from the derivative and endpoints, not from merely looking at a graph.
The same curve works for different distances and durations
Once the normalised function s(u) is chosen, we can reuse it with different L and T values.
For a distance of 300 pixels over five seconds:
x(t) = 300s(t/5).
For a distance of 80 pixels over two seconds:
x(t) = 80s(t/2).
The shape in normalised time is the same.
The real velocities differ because velocity scales with L/T.
The accelerations differ because acceleration scales with L/T².
This is an important mathematical insight.
Scaling distance and changing duration affect different derivatives in predictable ways.
The same easing pattern can therefore feel much faster or slower depending on its scale and duration.
Why acceleration grows when duration shrinks
For cubic smoothstep:
a(t) = (L/T²)(6 – 12u).
If we keep L fixed and halve T, the factor L/T² becomes four times as large.
So the model’s acceleration magnitudes become four times as large at corresponding normalised phases.
The velocity scale L/T doubles.
That is consistent with the requirement to cover the same distance in half the time while beginning and finishing at rest.
This scaling relationship is useful for understanding why speed and acceleration are not interchangeable.
It also shows why a motion rule comfortable in one timing context may behave differently when played much faster.
For physical machines, additional engineering limits matter.
A five-session study plan
Session one: linear movement
Define a 120-pixel movement over four seconds.
Write x(t) = 30t.
Calculate position and velocity at several times.
Explain why acceleration is zero in the active interval.
Session two: compare quadratic ease-in and ease-out
Use s(u) = u² and s(u) = 2u – u².
Differentiate both.
Identify which begins at rest and which ends at rest.
Session three: derive cubic smoothstep
Start with a general cubic polynomial.
Apply two endpoint-position and two endpoint-velocity conditions.
Solve for its coefficients and verify the result.
Session four: read the velocity and acceleration
For cubic smoothstep, calculate the position at one, two and three seconds.
Find the maximum speed and explain when acceleration changes sign.
Session five: compare alternative curves
Study the cosine ease-in-out and quintic smoothstep.
Identify which endpoint derivatives vanish and explain why that matters.
Then contrast one-shot easing with a periodic sine-wave animation.
The aim is understanding the mathematical trade-offs, not memorising a graphics preset.
Four errors worth diagnosing
Mistake one: using a position value as a velocity
The learner reports x(2) = 60 as though it meant sixty pixels per second.
The repair is to distinguish x(t) from x'(t) and label their units.
Mistake two: forgetting the inner time derivative
The learner differentiates s(t/T) without multiplying by 1/T.
The repair is to apply the chain rule carefully.
Mistake three: assuming zero endpoint velocity guarantees zero endpoint acceleration
The learner overlooks that cubic smoothstep has non-zero one-sided acceleration at its active boundaries.
The repair is to calculate the second derivative and examine the join with stationary sections.
Mistake four: treating a continuous formula as a display guarantee
The learner assumes mathematical continuity prevents dropped frames or pixel quantisation.
The repair is to distinguish the intended motion function from the rendering process.
These are different conceptual problems requiring different corrections.
What parents can ask during revision
Ask the student how a linear position function differs from an easing function.
Ask what the derivative of position describes.
Ask why cubic smoothstep can reach a higher peak speed than linear motion while covering the same total distance in the same time.
Ask how four endpoint conditions can determine a cubic polynomial.
Ask whether an object with zero velocity must also have zero acceleration.
Ask why a sine wave is appropriate for repeated motion but not automatically for a one-time move.
These questions connect familiar Mathematics with an engaging application.
A learner who can explain the curve’s behaviour has understood more than its formula.
Singapore’s Additional Mathematics context
The official 2027 Singapore-Cambridge Secondary Education Certificate G3 Additional Mathematics syllabus K341 includes polynomial functions, trigonometric functions, differentiation, rates of change and applications of calculus.
Animation easing and parametric screen-coordinate models are enrichment applications rather than separate examined G3 Additional Mathematics topics.
The SEAB 2027 G3 syllabus directory gives the current authoritative syllabus listing.
The educational purpose is to show how the assessed foundations connect with later digital-media, computing and interactive-design ideas.
How a three-student tutorial can help
One learner may differentiate a polynomial correctly but struggle when the input is t/T.
Another may understand the function values but confuse velocity with acceleration.
A third may calculate accurate endpoints without identifying the additional conditions needed for smooth starting and stopping.
These students need different explanations.
In a carefully managed three-student Additional Mathematics tutorial, a tutor can inspect the equation, the derivatives and the interpretation of the result.
The immutable Secondary 1 Mathematics small-group tutorial in Clementi illustrates why early fluency with equations, graphs and function notation matters before more advanced applications.
Students progressing well through school and independent practice may not need tuition. Targeted support should address an identifiable learning need.
Frequently asked questions
What is an easing curve?
It is a mathematical function used to control how a value changes over a finite transition, often shaping acceleration and deceleration rather than using constant speed.
Why can a cubic polynomial start and end at zero velocity?
Four coefficients can be determined using two endpoint-position and two endpoint-derivative conditions.
Does zero velocity imply zero acceleration?
No. An object can be instantaneously at rest while its velocity is beginning to change.
Is cubic smoothstep perfectly smooth in every sense?
No. It has continuous endpoint velocity when clamped to stationary sections, but its acceleration can change abruptly at the joins. Higher-degree curves can impose additional smoothness conditions.
Why is a normalised time variable useful?
It separates the duration T from a shape function defined over the standard interval from zero to one.
Can a higher frame rate repair an incorrect motion function?
Not necessarily. Sampling more frequently does not remove a discontinuity or make an unsuitable trajectory correct.
Is animation programming part of the SEC A-Math syllabus?
No. The application is an enrichment connection to functions, trigonometry and calculus taught in Additional Mathematics.
The lasting reason to understand mathematical motion
Additional Mathematics helps students see that an animation is not simply a sequence of positions.
A position function describes where an object should be.
Its derivative describes how quickly the position changes.
Its second derivative describes how that velocity changes.
Endpoint conditions determine which functions meet a desired movement pattern.
A carefully constructed cubic or trigonometric curve can therefore express an intention that would be difficult to describe through isolated coordinates alone.
The deeper skill is building a function from conditions and checking that its consequences match the intended behaviour.
That is a valuable reason to study Additional Mathematics.
For connected reading, explore Why Study Additional Mathematics | Read a Derivative Graph to Predict Function Behaviour, Why Study Additional Mathematics | Explore Circular Motion with Trigonometric Coordinates, Why Study Additional Mathematics | Understand How the Product Rule Tracks Two Changing Factors and the eduKateSG Additional Mathematics project library.