Speed-of-Light Bounce: Concepts & Applications

• Speed-of-Light Bounce is a phenomenon where finite light-travel time causes observable reflection, timing minima, and geometric reorganization in diverse experimental and theoretical contexts.
• Round-trip optical measurements—exemplified by Fizeau’s method and modern chopper systems—demonstrate how extinction minima indirectly capture the light’s travel time.
• The concept extends to apparent superluminal illumination fronts, retarded-time imaging, and cosmological varying-speed-of-light models, linking laboratory techniques with theoretical astrophysics.

“Speed-of-Light Bounce” denotes a family of problems in which finite light-travel time, reflection, or passage through a light-speed threshold determines the observable phenomenon. In experimental optics, it is the round-trip measurement of $c$ by sending light to a distant reflector and inferring the flight time from periodic extinction rather than direct stopwatch timing (Semay et al., 2018). In string-gas cosmology, the phrase is used explicitly for a dilaton-driven varying-speed-of-light phase with an early superluminal regime, a crossover, and a late collapse of $c$ before matching to a branch with $c=1$ (Nayeri, 19 Mar 2026). Closely related literatures treat moving mirrors, relativistic reflection, apparent superluminal illumination fronts, retarded-time imaging, and time-resolved multi-bounce sensing as cases in which a “bounce” of light, or an effective crossing of the light line, organizes the kinematics and the data (Good, 2016, Maesumi, 2016, Nemiroff et al., 2015, Hornof et al., 2024).

1. Terminological range and scope

The phrase appears in multiple non-equivalent technical contexts. One context is literal round-trip optical propagation: light is emitted, reflected, and detected after a two-way journey, and the finite value of $c$ is inferred from the return condition (Semay et al., 2018). A second context is explicit varying-speed-of-light cosmology, where the quantity that “bounces” is the effective propagation speed itself rather than the scale factor (Nayeri, 19 Mar 2026). A third context consists of reflection and imaging problems in which a light-speed threshold, such as $v_r=c$ for an apparent spot or $w\to 1$ for a relativistic particle, marks a qualitative change in what is observed (Nemiroff, 2014, He et al., 4 Feb 2026).

Several nearby bodies of work are related only indirectly because the relevant “speed” is not the physical speed of light. In the ghost-condensate/Galileon nonsingular bounce, the notable speed quantity is the scalar perturbation sound speed $c_s$, which becomes imaginary near the bounce while super-horizon curvature perturbations remain essentially constant (Battarra et al., 2014). In the $\Lambda$CDM bounce scenario, the key scale is the sound Hubble radius $r_{sH}=c_s/H$, with the small sound speed of cold dark matter controlling scalar perturbations through an LQC bounce (Cai et al., 2014). In relativistic hydrodynamics, the sound speed can approach the speed of light from below, but the paper explicitly states that there is no “bounce” in the sense of a reflection or turning point at $c_s=c$ (Moore, 2024). This distinction is essential because the phrase “speed-of-light bounce” is otherwise easily conflated with sound-speed-controlled bounce cosmologies.

2. Round-trip optical bounce measurements

The classical experimental form of the topic is Fizeau’s 1849 measurement and its modern reproduction in Mons. The governing idea is to convert a microsecond-scale round-trip time into a null measurement: the returning beam is extinguished when a periodically interrupted optical element changes phase between transmission and return (Semay et al., 2018). In Fizeau’s original wheel with $c$0 notches rotating at $c$1, the first extinction satisfies

$c$2

so that

$c$3

For Fizeau’s $c$4, this becomes

$c$5

The Mons implementation preserved the same timing-by-extinction logic but replaced the cogwheel with an electric optical chopper, used a $c$6, $c$7 green laser, and observed the return with a CCD camera (Semay et al., 2018). The one-way baseline was reported as

$c$8

with the remote reflector realized as a $c$9 retroreflective panel. Because the outgoing and returning beams passed through diametrically opposite points of the chopper, the relevant analysis used the spacing between successive minima rather than a single first extinction: $c=1$0 With typical extinction frequencies

$c=1$1

the reported estimate was

$c=1$2

The apparatus was deliberately pedagogical rather than metrological. The chopper had $c=1$3 slots, diameter $c=1$4, and maximum rotation $c=1$5; the authors estimated extinction frequencies only to about $c=1$6, implying an expected overall error of order $c=1$7 (Semay et al., 2018). During the public exhibition, 30 acceptable measurements gave a mean

$c=1$8

with standard deviation

$c=1$9

and after removing two bad-weather outliers the same mean with reduced standard deviation

$c$0

The experimental significance lies in the fact that a two-way light “bounce” can be timed indirectly through transmission minima, making the round-trip structure itself the observable.

3. Apparent superluminal surface bounces

A second major usage concerns not a measured two-way flight time, but the motion of the locus where light first illuminates a surface. For a compact source that turns on near a planar wall, the first illuminated point lies at the closest approach, reached at

$c$1

and the illuminated front along the wall satisfies

$c$2

Its speed along the wall is

$c$3

which is formally infinite at first contact, remains $c$4 for all finite $c$5, and approaches $c$6 asymptotically (Nemiroff et al., 2015). In the generalized surface description,

$c$7

so for a nonradial surface direction, where $c$8, one obtains $c$9 (Nemiroff et al., 2015).

The associated “bounce-like” event in the observational sense is the virtual spot-pair creation described for sweeping beams across scattering surfaces. The relevant threshold is the point where the real spot’s radial velocity relative to the observer satisfies

$v_r=c$0

At that locus, observer time has a stationary point, and two apparent spots are seen to emerge and move away from the same location; one of them can appear to move opposite to the true sweep direction (Nemiroff, 2014). For a sphere illuminated by a sweeping beam, the virtual pair-creation location is

$v_r=c$1

while for a spot moving at constant real speed $v_r=c$2 along a wall it is

$v_r=c$3

The divergence of the apparent transverse speed at that point,

$v_r=c$4

is the kinematic reason the effect resembles a bounce or splitting event (Nemiroff, 2014).

These results do not violate relativity. The photons still travel locally at $v_r=c$5; what exceeds $v_r=c$6 is the geometric locus of equal arrival time on the surface, not matter or local causal propagation from one surface point to the next (Nemiroff et al., 2015).

4. Reflection, retarded-time imaging, and moving boundaries

For light reflected from uniformly moving mirrors, the finite signal speed and the mirror motion can be encoded geometrically through the effective surface of reflection (ESR). With a stationary source at $v_r=c$7 and a flat mirror moving as $v_r=c$8, the impact points of rays emitted at time $v_r=c$9 satisfy

$w\to 1$0

This is the focus-directrix definition of a conic: the ESR is a hyperbola for $w\to 1$1, a parabola for $w\to 1$2, and an ellipse for $w\to 1$3 (Maesumi, 2016). Applying Fermat’s principle to the ESR reproduces the relativistic reflection law,

$w\to 1$4

with equivalent forms

$w\to 1$5

In this framework, a “bounce” from a moving reflector is re-expressed as ordinary focal geometry of a static conic (Maesumi, 2016).

Retarded-time image formation leads to an allied phenomenon in relativistic visual optics. For a moving object photographed at one detection time, different surface points contribute photons emitted or reflected at different earlier times. The basic relation is

$w\to 1$6

This shift compensates apparent Lorentz contraction in a snapshot and yields the Terrell effect: for distant observation, a moving sphere appears approximately like a rotated sphere rather than a flattened ellipsoid (Hornof et al., 2024). The paper gives the apparent rotation angle as

$w\to 1$7

and implements the construction experimentally with fs-laser pulses and a gated intensified camera (Hornof et al., 2024). Here the observable “bounce” logic is again retarded-time completion of reflected paths, not simultaneity at one coordinate time.

A more extreme light-line limit occurs in moving-mirror quantum field theory. In the horizonless mirror trajectory

$w\to 1$8

one has

$w\to 1$9

For $c_s$0, the mirror asymptotically reaches the speed of light, produces thermal emission on the left,

$c_s$1

and finite total energy on the right,

$c_s$2

without an acceleration horizon (Good, 2016). The paper is explicit that this is not a literal finite-time bounce or turnaround; it is asymptotic lightlike drift of a reflecting boundary.

An engineered wave-physics analogue appears in synthetically moving spacetime gratings. In the constant-refractive-index model with

$c_s$3

the modulation speed is written

$c_s$4

Near the luminal regime, the “Chimera” Bloch branch rotates continuously, its group velocity vanishes at one critical point and changes sign, and a backward-propagating pulse can halt and re-emerge in the opposite direction when the grating is turned on (Pendry et al., 2021). This is a genuine bounce-like reversal, but the mechanism is mode conversion in a spacetime band structure rather than mirror reflection.

5. Cosmological and strong-gravity usages

The most literal cosmological use of the phrase appears in string-gas cosmology with a dilaton-controlled varying speed of light,

$c_s$5

On the analytic Hagedorn background, this gives

$c_s$6

so that $c_s$7 as $c_s$8 for $c_s$9, $\Lambda$0 at a crossover with

$\Lambda$1

and $\Lambda$2 as $\Lambda$3 (Nayeri, 19 Mar 2026). The self-dual point

$\Lambda$4

is identified as a T-duality anchor for matching to a late branch with $\Lambda$5. The reported horizon enhancement factors are $\Lambda$6 for $\Lambda$7 and $\Lambda$8 for $\Lambda$9, while the flatness parameter is numerically suppressed by $r_{sH}=c_s/H$0–$r_{sH}=c_s/H$1 over the late Hagedorn phase (Nayeri, 19 Mar 2026). In this literature, the “bounce” refers to the branchwise causal history of the effective propagation speed itself.

In regular black-bounce-Schwarzschild spacetimes, the light-speed limit organizes both weak- and strong-field lensing. For relativistic massive particles, the weak-field deflection is

$r_{sH}=c_s/H$2

which reduces smoothly to the photon case at $r_{sH}=c_s/H$3 (He et al., 2024). In the strong-deflection limit, the particle-sphere condition

$r_{sH}=c_s/H$4

reduces to the photon-sphere equation as $r_{sH}=c_s/H$5 or $r_{sH}=c_s/H$6, giving

$r_{sH}=c_s/H$7

for the lightlike case (He et al., 4 Feb 2026). The strong-deflection form

$r_{sH}=c_s/H$8

shows that the bounce parameter modifies the strong-deflection coefficients and photon-sphere radius even though the critical impact parameter remains $r_{sH}=c_s/H$9 in the model (He et al., 4 Feb 2026). Here the “bounce” belongs to the regularized black-hole geometry, while the “speed-of-light” limit is the null-geodesic endpoint of timelike particle lensing.

A different cosmological use concerns subluminality constraints rather than a varying $c_s=c$0. In beyond-Horndeski bouncing cosmology, one modified model is arranged so that

$c_s=c$1

during the bounce and $c_s=c$2 asymptotically, avoiding the earlier situation in which the exact bounce lay on the boundary of a superluminal region (Mironov et al., 2019). This is not a speed-of-light bounce in the VSL sense; it is a bounce made compatible with the metric light cone for perturbations.

6. Applications, misconceptions, and controversies

Time-resolved computational imaging has turned multi-bounce light into a direct inference channel. In single-photon lidar with multiplexed illumination, the paper “Shoot-Bounce-3D” models the two-bounce time of flight from laser spot $c_s=c$3 to visible scene point $c_s=c$4 as

$c_s=c$5

and uses learned decomposition of measured transient data into per-spot two-bounce structure to recover dense visible depth, occluded geometry, and specular surfaces from a single capture (Klinghoffer et al., 5 Dec 2025). In this setting, a speed-of-light bounce is literal optical transport: photons arrive later because they have traversed longer, multi-segment paths.

A central misconception across the literature is that any superluminal appearance implies superluminal signaling. The wall-illumination and sweeping-spot analyses are explicit that the moving quantity on the surface is a geometric front or virtual spot, not a material object or locally transmitted causal signal (Nemiroff et al., 2015, Nemiroff, 2014). A second misconception is that every abrupt change in optical speed is a reflection phenomenon. The photonic speed-bump literature instead describes a short slow-light section that changes group velocity from roughly $c_s=c$6 to $c_s=c$7 “without any reflection,” using impedance-matching tapers; the structure is a velocity transition, not a literal bounce from a mirror (Faggiani et al., 2016).

The principal controversy represented here concerns claims of observer-dependent $c_s=c$8 from round-trip ranging. A lunar laser ranging analysis reported a measured speed exceeding canonical $c_s=c$9 by about the observatory line-of-sight speed and interpreted this as a first-order violation of local Lorentz invariance (Gezari, 2009). The same technical record notes, however, that the conclusion depends on assigning the round-trip path length in the observatory frame as $c$00 while using a round-trip time that already reflects the moving reflector/observer geometry, so the claimed anomaly is tied to frame-dependent path assignment rather than the standard two-way invariance test (Gezari, 2009). This issue underlines a general point: in bounce-type measurements, the observable is always a combination of timing and geometry, and the interpretation depends on keeping those frame assignments consistent.

Taken together, these literatures do not define a single theory of “Speed-of-Light Bounce.” They define a recurrent structure. Light is emitted, reflected, swept, delayed, or driven toward a lightlike threshold; the observable is then fixed by how paths, phases, or effective propagation speeds reorganize around $c$01. In one domain that yields a public measurement of the speed of light by extinction minima (Semay et al., 2018). In another it yields superluminal illumination fronts that remain compatible with relativity (Nemiroff et al., 2015). In another it yields a dilaton-driven branchwise evolution of $c$02 in string-gas cosmology (Nayeri, 19 Mar 2026). The unifying content is the same: finite propagation speed converts geometry into timing, and timing into structure.