Home/ Tool index/ Metronome Frame-Scrubber
Launch ready Session 3 · Video Interactive

Metronome Frame-Scrubber

Video systems must maintain relationships across time. This authored teaching simulation changes how many earlier drawings its code uses as references; it does not measure a real model's memory or guarantee that wider context locks motion together.

tools/metronome-frame-scrubber/

Live preview · launch for the interactive version

§ A · What it makes visible

one designed comparison · not a model benchmark
Fig. 01

Simulated reference span

Set how many earlier authored frames this browser simulation uses when drawing the current one.

Fig. 02

Onion-skin trail

Past frames ghost behind the current one, like an animator's light-box.

Fig. 03

Drift vs. coherence

In this designed animation, a short span adds more variation and a longer span adds less. That result is built into the code.

§ B · How to investigate it

run it like an experiment, not a toy
01 · Predict

Before you widen

Predict what the simulation will display when it uses only one earlier frame.

prediction: more or less visible variation?
02 · Change one thing

Move the window

Change only the simulation's reference span; keep playing.

1 frame → 24 frames
03 · Compare evidence

Read the trail

How do the onion-skin ghosts line up at each setting?

longer authored span → more aligned ghosts here
04 · Name it

Bound the claim

Name what changed in the simulation, then separate that observation from a claim about a real architecture.

observed here ≠ universal mechanism

§ C · Debrief questions

after the investigation
What changed when this simulation used a shorter reference span?
Which relationship does the light-box analogy make visible?
What does this authored animation not demonstrate about a real video model?
Which model documentation or controlled test would you need before making an architecture claim?

§ D · Related

pairs well with · use in context