Math & Units

dBFS (Decibels Relative to Full Scale)

/dee-bee-eff-ess/
Used throughout digital signal chains, this unit reports an amplitude in decibels referenced to the full-scale level of a data converter rather than to an absolute power. Zero dBFS marks the largest code an ADC or DAC can represent, so every real signal is expressed as a negative number below it; a sine wave occupying half the converter range reads -6 dBFS. Unlike dBm, which is anchored to 1 mW, dBFS is a relative scale that only maps to absolute power once the converter's full-scale voltage and input impedance are known. It is the natural language for digitizer headroom, quantization noise, and the noise floor inside software-defined radios and digital receivers.
Category: Math & Units
Reference: 0 dBFS = converter full scale
Typical range: 0 to -120 dBFS

How dBFS Anchors Digital Signal Levels

In an analog signal chain, levels are stated in absolute units like dBm because every node sits in a known impedance and 1 mW is a fixed amount of power. Once a signal is digitized, that absolute anchor disappears; the only meaningful reference becomes the largest number the converter can produce. dBFS captures exactly that. The full-scale value is the peak amplitude that drives the converter to its maximum positive and minimum negative codes, and 0 dBFS is defined as that level. Because no encoded sample can exceed full scale, dBFS readings are normally negative, and a value pinned at 0 dBFS is a warning that the signal is clipping.

There is an important convention to watch. A full-scale sine wave swings between the converter's positive and negative full-scale codes, so its RMS value is 3.01 dB below the peak. Some instruments define 0 dBFS as a full-scale sine RMS, while others define it as the peak. A square wave that toggles between the two full-scale codes therefore reads +3.01 dBFS on a sine-referenced meter even though it never exceeds the code range. Engineers must confirm whether a given dBFS figure is peak-referenced or sine-RMS-referenced before comparing measurements across instruments, because the two differ by exactly 3.01 dB.

dBFS also sets the budget for converter dynamic range. The quantization noise floor of an ideal converter sits at a fixed level in dBFS, so the usable span between a near-full-scale carrier and that floor is what defines the spurious-free dynamic range and SNR of the digitizer. Receiver designers set analog gain so the strongest expected signal lands a few dB below 0 dBFS, maximizing the distance above the noise floor without risking clip.

Core dBFS Relationships

Level in dBFS (peak-referenced):
LdBFS = 20 × log10(Vpeak / VFS)

Full-scale sine versus square wave:
FS sine RMS = VFS / √2 ≈ −3.01 dBFS (peak ref); FS square = 0 dBFS RMS

Ideal quantization SNR (full-scale sine):
SNR = 6.02N + 1.76 dB   (carrier at 0 dBFS, N-bit converter)

Mapping dBFS to absolute power:
PdBm = LdBFS + PFS,dBm

Where VFS = full-scale peak voltage, N = converter resolution in bits, and PFS,dBm = the converter full-scale power for its rated input impedance. Example: a tone at −10 dBFS into a part whose full scale is +4 dBm sits at −6 dBm.

dBFS Compared with Other Decibel References

UnitReference (0 point)Absolute or relativeSign of typical valuesWhere used
dBFSConverter full scaleRelative to device FSNegative (0 = clip)ADC/DAC, SDR, digitizers
dBm1 mWAbsolute powerPositive or negativeRF chains, link budgets
dBcCarrier powerRelative to carrierNegativeSpurs, phase noise, harmonics
dBV1 V RMSAbsolute voltagePositive or negativeAudio, instrumentation
dBHz1 Hz bandwidthBandwidth ratioPositiveNoise density, C/N0
Common Questions

Frequently Asked Questions

Why are dBFS values almost always negative?

Because 0 dBFS is defined as the largest amplitude the converter can encode, any signal not driving it to the full code range falls below 0 dBFS and is therefore negative. A sine using half the voltage range reads −6 dBFS; a low-level tone near one tenth of full scale sits near −20 dBFS. A reading of exactly 0 dBFS means the signal touches the maximum code, and going higher clips. Designers leave headroom, often −1 to −3 dBFS for a peak sine and −12 to −20 dBFS RMS for noise-like waveforms, so crest-factor peaks do not clip.

How do you convert between dBFS and dBm at an ADC input?

You must know the converter's full-scale input power, which depends on its full-scale peak voltage and input impedance. For a part with 1.4 V peak-to-peak differential full scale into 100 ohm, full-scale sine power is about +4 dBm, so a tone at −10 dBFS is roughly −6 dBm at the pins. The general rule is P(dBm) = level(dBFS) + FS(dBm). Because every converter has a different full-scale voltage and reference impedance, dBFS values are not portable between parts until anchored to that device's full-scale power.

What is the theoretical relationship between dBFS, ENOB, and SNR?

For an ideal N-bit converter driven by a full-scale sine, quantization SNR is 6.02N + 1.76 dB with the carrier at 0 dBFS. Effective number of bits is ENOB = (SINAD − 1.76) / 6.02, where SINAD is taken with a near-full-scale tone, usually −1 dBFS to avoid clipping. When the input backs off from full scale, the signal drops in dBFS while the noise floor stays roughly fixed, so in-band SNR degrades nearly decibel for decibel below full scale.

Digital Receiver Front Ends

Tame Your dBFS Budget

From low-noise downconverters to wideband digitizer front ends, our integrated assemblies set the right gain so your strongest signal lands just below full scale. Talk to our engineering team about your dynamic-range targets.

Get in Touch