Audio
Loudness of one chunk, linear 0..1 relative to full scale.
Both values use the plain sample-value convention: a full-scale square wave reads 0 dBFS peak and RMS, a full-scale sine reads 0 dBFS peak and about -3 dBFS RMS (not the AES17 convention, which adds 3 dB to RMS). Conversational speech close to a phone typically peaks around -20 to -10 dBFS.
Use peakDbfs or rmsDbfs for meters in decibels, or feed a LevelNormalizer for a 0..1 value that behaves the same across devices.