Visual Neurobiology & Physiology

Animal vision: how fast flicker becomes steady light

How quickly can a light flash before it looks steady? The threshold differs among animals and depends on light and measurement method. Critical flicker fusion frequency, or CFF, describes this limit. Vision is not a sequence of camera frames.

Peregrine falcon
≥129 Hz
7.75 ms · 1000/f
Human
60 Hz
16.67 ms · 1000/f
Leatherback turtle
15 Hz
66.67 ms · 1000/f

One rotation, sampled at different frequencies

Change the frequency to compare the positions retained by the model. The animation runs ten times slower: one model second takes ten screen seconds. The dot stays lit. This is a sampling diagram, without vision calibration or reproduction of rapid flicker.

Model paused
Sampling Controls
Model frequency 60 Hz
Period 1000/f: 16.7 ms
Time scale: 1 s → 10 s
Use: Sampling diagram

The same movement, different intervals

Choose a published value to change the mathematical sampling interval. CFF is an illustrative parameter here; the result does not reproduce a species’ visual experience.

Peregrine falcon Model: 129 Hz
Mathematical samples in one model second 2.15×
0.0 s (start) 0.5 s 1.0 s (1000 ms)

Peregrine falcon

Falco peregrinus
≥129 Hz
Published value

The selected published value is ≥129 Hz. Conditions and methods differ among studies; this number does not establish how many images the animal sees each second.

Period 1000/f: 7.75 ms
Conditions: Behavioural test
Interval relative to 60 Hz: 0.47×
Potier et al. (2020), abstract

Published values and their conditions

Ten references come from Healy and colleagues’ table (2013); the falcon value comes from Potier and colleagues (2020). Light and methods differ. Comparing rows is not a controlled experiment between species.

On a phone, scroll the table sideways for conditions, values and sources.
SpeciesScientific nameConditionsPublished valuePeriod 1000/fRatio to 60 Hz
Peregrine falconFalco peregrinusBehavioural test · behavioural test≥129 Hz7.75 ms2.15×
Rock pigeonColumba liviaHigh light · ERG100 Hz10.00 ms1.67×
Domestic chickenGallus gallus domesticusHigh light · behavioural test87 Hz11.49 ms1.45×
DogCanis lupus familiarisHigh light · behavioural test80 Hz12.50 ms1.33×
GoldfishCarassius auratusHigh light · ERG67.2 Hz14.88 ms1.12×
HumanHomo sapiensHigh light · behavioural test60 Hz16.67 ms1.00×
CatFelis catusLow light · ERG55 Hz18.18 ms0.92×
Loggerhead turtleCaretta carettaHigh light · ERG40 Hz25.00 ms0.67×
Tiger salamanderAmbystoma tigrinumLow light · ERG30 Hz33.33 ms0.50×
Leatherback turtleDermochelys coriaceaHigh light · ERG15 Hz66.67 ms0.25×
European eelAnguilla anguillaLow light · behavioural test14 Hz71.43 ms0.23×

What these values can tell us

The limits of the comparison matter as much as the difference between the numbers.

Flicker and motion are different

A screen’s refresh rate, its luminance flicker and a film’s frame rate describe different properties. A dog’s CFF value alone cannot establish whether every 60 Hz television produces smooth motion for it.

A distance calculation, without a perceived speed

In a chosen 4 m/s example, an object travels about 6.67 cm in 1/60 s and 3.10 cm in 1/129 s. These distances follow from speed × time. They do not imply that the object appears to move more slowly to an animal.

Light changes the threshold

Lisney’s chicken study measured thresholds at different light intensities. That is why the table retains the compilation’s lighting conditions and 60 Hz remains only a calculation reference.

Experimental Data & Visual Physics

Flicker fusion is the frequency at which an alternating light can no longer be distinguished from steady light under the test conditions. The threshold depends on illumination, stimulus, method and individual. Electroretinography (ERG) measures the retina’s electrical response; behavioural tests measure stimulus discrimination.

The human 60 Hz reference comes from Healy’s compilation, rather than being a universal limit. The falcon study demonstrated at least 129 Hz; the animation uses 129 Hz as its selected value. Differences between these values do not measure how slowly the world appears to move.

The period 1000/f and ratio to 60 Hz are calculations about a periodic signal. They are not retinal frame durations, reaction times or perceived speeds. The model stops at that analogy; the screen and browser constrain its rendering.

Healy’s comparative analysis found associations between temporal resolution, body mass and metabolism while accounting for light and relatedness. This is not a strict rule for every animal pair. Niven and colleagues’ experiments in four fly species show that photoreceptor performance and information-transmission costs differ among species and with illumination.