Size control lab
Same Sicily silhouette, same physical scale limits (0.3–2.2). No answer to chase: compare coarse travel and small corrections. The main game does not use a wheel.
Linear slider
16px of a 240px track at this value: +25.3%. At the upper limit, travel clamps.
Keyboard arrows adjust; Shift + wheel arrows = 0.2%. Slider Home/End reach limits. Nudge buttons are shared across all three.
Log slider
16px of a 240px track at this value: +14.2%. At the upper limit, travel clamps.
Keyboard arrows adjust; Shift + wheel arrows = 0.2%. Slider Home/End reach limits. Nudge buttons are shared across all three.
Rotary wheel
Two turns cover the range. Clockwise grows; the center is a dead zone. Crossing 12 o’clock never wraps the value.
Keyboard arrows adjust; Shift + wheel arrows = 0.2%. Slider Home/End reach limits. Nudge buttons are shared across all three.
Try a finger-sized change, then a tiny correction
Reset at 0.3 and drag about 16px. Linear adds about 42.2%; log adds 14.2%. At 1.8, the same travel adds 7.0% linear or 14.2% log. Log improves relative precision at small values, but is coarser at large values. Both benefit from fine nudges; this is not evidence that log alone fixes every size task.
Log mapping: s(t) = 0.3 × (2.2/0.3)t; inverse t = ln(s/0.3) / ln(2.2/0.3). Rotary movement uses unwrapped, incremental log-domain deltas, with a center dead zone and no value wrap.
References: D3 log scales · JUCE slider styles · JUCE skew/custom mappings. These document techniques, not a user study or a claim that other games use log sizing.