Sixteen bits describe 65 536 amplitude levels (third link), and the decibel figure is twenty times the base-ten logarithm of that count: 96.329598612474 dB, on the first link. The second link is the rule everyone in the trade actually quotes — 6 dB per bit, so 6 × 16 = 96 — and the two figures differ by a third of a decibel. Data sheets show both conventions: 96 dB for the range of the quantiser, and 98.08 dB for a full-scale sine wave, where the formula is 6.02N + 1.76.
The logarithm is not decoration. A decibel is 20 log₁₀ of an amplitude ratio because the ear answers to ratios rather than differences, so doubling the resolution buys the same 6.02 dB wherever you start: 24-bit audio sits near 144.5 dB, which is why the extra eight bits are worth having on a recording desk and invisible in a phone. It is also why the number of bits, not the number of levels, is what the specification trades in — and why the CD's 16 bits at 44 100 samples a second were a capacity decision as much as a fidelity one: that pair of choices fills a seventy-four-minute disc to the brim.
Answer
96.329598612474
How to read it
20·log(65 536) on the workbench, log being base ten. The second link is the 6 dB-per-bit estimate and the third is 2¹⁶.
Source
Dynamic range of an N-bit linear converter is 20·log₁₀(2^N) = 6.0206N dB, so 16-bit audio is quoted at 96 dB by the 6 dB-per-bit rule and at 98.08 dB under the full-scale sine convention 6.02N + 1.76
Work it in the calculator
These links open the right calculator with this problem's numbers already typed in. The line above is what the screen should show — check it yourself.