dBFS (Decibels Full Scale)

In digital audio, there is an absolute ceiling: the maximum value that the bit depth can represent. For 16-bit audio, this is 32,767 (positive) or -32,768 (negative). For 24-bit, it is 8,388,607. The dBFS scale maps this maximum to 0 dBFS, with all lower levels expressed as negative numbers. A signal at -6 dBFS is half the voltage (or one quarter the power) of full scale. A signal at -20 dBFS is at one-tenth the voltage of maximum. Professional practice targets recording levels between -18 dBFS and -12 dBFS for average program material, leaving headroom for peaks. The relationship between dBFS and acoustic SPL depends entirely on the system's gain structure. There is no fixed conversion. A microphone preamp at one gain setting might produce 0 dBFS at 120 dB SPL; at a different gain, the same acoustic level might register at -20 dBFS. Calibration establishes this relationship for a specific measurement chain. For measurement purposes, SonaVyx works in the digital domain and displays levels in dBFS when showing raw capture levels. When calibrated against a known SPL reference (such as a 94 dB calibrator), the dBFS readings are offset to display true SPL values. Without calibration, relative measurements like frequency response shape and RT60 remain accurate, but absolute SPL readings are uncalibrated. Digital meters should always show peak levels in dBFS to warn of clipping, while average levels are better represented by RMS calculations. Modern DAWs use floating-point processing internally, allowing inter-stage levels above 0 dBFS without clipping, but the final output stage is always fixed-point and must not exceed 0 dBFS.