An animation that runs in 300 milliseconds flashes past every intermediate year unseen

What it is

An animation's informational value presumes "the reader can finish reading the intermediate states": a 300-millisecond animation of yearly data evolution flashes the intermediate years past unseen—only start and end are perceived, and side-by-side static images would serve better. When duration is too short, the animation degenerates from "information carrier" to "fancy transition"—or worse, creates the illusion of "having seen the process" when the reader saw only the endpoints. Duration is therefore an information parameter, not an aesthetic one, and its floor must be set by the slowest target readers' processing rate.

Why it happens

The duration floor's mechanism is the ratio of cognitive processing rate to information density: the time a state needs depends on its information load (data points, text, objects to compare) and the reader's processing rate (shaped by age, cognitive load, assistive technology). For screen-reader users, "reading" runs at listening speed—a narrated summary plus key points takes seconds to tens of seconds; visual scanners can process a single frame in a few hundred milliseconds, yet cumulative understanding across consecutive states still takes seconds. The workable design rule: compute a total-duration floor as "states × minimum recognition time per state," and make the interval between state changes configurable (extending into the same group's pause and stepping—speed control is more fundamental than precisely tuning a default duration). A second mechanism is judging "when the animation carries information": if intermediate states have no independent value (transitions masking loading), a short animation harms nothing; if the intermediate states are the content (the data's evolution), insufficient duration is information truncation.

Where it stops holding

Duration relates to attention with a ceiling as well: overly long animations are equally harmful (readers are forced to wait or leave early), so the practical constraint on total length is the reader's patience window for the task, past which skip and speed-up must be offered. The duration floor differs across channels: the visual channel can run at a few hundred milliseconds per state, while the auditory channel (narrated summaries) is bounded by speech rate and noticeably slower—when a visual animation and audio summary run together, total duration should align to the slower one. The boundary is also gradual: moving from "enough" to "insufficient" duration is continuous; in practice, calibrate with user testing (can participants restate the intermediate states?) rather than picking milliseconds from intuition, testing with the slowest target reader group included.

Applying it

• Validate information-bearing animations against the duration floor: states × minimum recognition time per state sets the animation's minimum total duration.
• Offer playback-speed control (0.5x/1x/2x), handing rate decisions to the reader as the fallback for any fixed duration.
• When audio summaries run alongside visual animation, align total duration to the auditory rate.
• Verification: have target readers watch the animation and restate the intermediate states; if most can name only the endpoints, the duration is insufficient—lengthen it or switch to static frames.

Related

• Same group: U11.03.1 Auto-playing animations must be pausable and support frame stepping · U11.03.2 States existing only during playback need static equivalents · U11.03.3 Rapid flashing and high-contrast motion can trigger discomfort or seizures · U11.03.4 Transition animations must respect the system's reduced-motion preference
• Nearby: U11.03.1 Auto-playing animations must be pausable and support frame stepping · U9.03.1 Annotations point to where readers should look
• Search terms: animation duration · reading speed · animation timing