Methodology

Timers use target timestamps rather than naive one-second decrement loops. Random selection uses browser cryptographic randomness where supported. Unit converters use centralized definitions.

Device and human-test results have browser and hardware limitations. Physical measurement tools require calibration before claiming real-world accuracy.