Also known as: effective number of bits
The effective number of bits (ENOB) is a measure of how many bits of real resolution a converter delivers, derived from its measured SINAD rather than its nominal bit count.1 A converter may be sold as “12-bit,” but once its own noise, distortion, and nonlinearity are folded in, it might behave like a perfect 10-bit part. ENOB puts a single, honest number on that gap by asking: how many ideal bits would give the signal quality this converter actually achieves?
How it works
For an ideal N-bit uniform quantizer, a full-scale sine wave has a signal-to-quantization-noise ratio of 6.02·N + 1.76 dB. ENOB simply inverts that relationship using the converter’s measured SINAD — the ratio of signal power to the sum of all noise and distortion power:
ENOB = (SINAD − 1.76 dB) / 6.02
Because SINAD counts thermal noise, harmonic distortion, clock jitter, and quantization error together, ENOB captures everything that degrades the converter below its ideal, not just the step size. A 14-bit ADC with a SINAD of 74 dB has an ENOB of about 12 — its two “missing” bits are consumed by real impairments.
Two consequences matter:
- ENOB is not a fixed number. It falls as the input frequency rises (aperture jitter and settling errors grow with frequency) and can change with amplitude and sample rate, so it is quoted at specific test conditions.
- The 6 dB-per-bit rule works both ways. Every 6 dB of SINAD lost to real-world effects costs one effective bit; that is why datasheets show ENOB rolling off across a converter’s usable band.
In practice
ENOB is the number to compare when the nominal bit counts mislead. A 16-bit audio codec running near DC may deliver close to its full resolution, while a fast 12-bit RF converter sampling at hundreds of megahertz might show an ENOB of 9 or 10 at the top of its band. Techniques like dither and oversampling can raise the in-band ENOB above the nominal count by moving quantization noise out of the band of interest, which is why a well-designed system sometimes measures more effective bits than its converter nominally has.
Relevance to SDR
For an SDR, ENOB — not the advertised bit depth — sets the real dynamic range and therefore how well a weak signal survives next to a strong one across the captured band. A receiver claiming 16 bits but delivering an ENOB of 11 has roughly 30 dB less usable range than the headline suggests, which shows up directly as a higher dBFS noise floor and worse blocking performance. When comparing SDR front ends for wideband trunking work, ENOB at the operating frequency is the honest figure of merit.
GopherTrunk decodes whatever samples the device delivers, so a capture’s usable weak-signal margin is bounded by the source’s ENOB long before the software sees the data — another reason the decode chain is only as good as the front end feeding it.
Sources
-
Effective number of bits — Wikipedia, on deriving real converter resolution from measured SINAD. ↩