What is the Signal-to-Noise Ratio (SNR)?
The signal-to-noise ratio compares the level of a desired signal to the level of background noise. It is a fundamental measure of quality in electronics, audio, telecommunications, and imaging. A higher SNR means a cleaner, clearer signal; a low SNR means noise begins to obscure the information you care about. SNR is most often expressed in decibels (dB) because signal and noise levels can span many orders of magnitude.
How to use this calculator
Enter the signal power (\(P_{\text{signal}}\)) and the noise power (\(P_{\text{noise}}\)) in the same units — milliwatts, watts, or any consistent power unit. The calculator divides the two to get the linear power ratio, then converts it to decibels. Both inputs must be positive numbers and the noise power cannot be zero.
The formula explained
The calculation uses $$\text{SNR}_{\text{dB}} = 10 \cdot \log_{10}\!\left(\frac{P_{\text{signal}}}{P_{\text{noise}}}\right)$$ Because decibels are a logarithmic, ratio-based unit, every 10 dB represents a tenfold increase in power. Note the factor of 10 applies to power ratios; if you are working with amplitude or voltage instead, you would use a factor of 20.
Worked example
Suppose the signal power is 100 mW and the noise power is 1 mW. The power ratio is \(100/1 = 100\). Then $$\text{SNR}_{\text{dB}} = 10 \cdot \log_{10}(100) = 10 \cdot 2 = 20\ \text{dB}$$ This indicates a strong, clean signal.
Typical SNR Ranges by Application
What counts as a "good" SNR depends on the application, the modulation scheme and the equipment. The figures below are general rules of thumb, not universal limits. Several come from vendor guidance or theory rather than a single standard.
| Application | Typical SNR | Notes |
|---|---|---|
| Wi-Fi / wireless link (vendor guidance) | < 10 dB poor; 10–15 dB marginal; 15–25 dB fair; 25+ dB good; 40+ dB excellent | Required SNR rises with the modulation and coding rate. Higher-order modulations such as 256-QAM need far more SNR than BPSK or QPSK. These bands are common vendor rules of thumb, not an IEEE-mandated threshold. |
| CD audio (16-bit PCM) | ≈ 96 dB (theoretical 98 dB for a full-scale sine) | Quantization limit: \(20\log_{10}(2^{16}) \approx 96.3\) dB. |
| Vinyl record | ≈ 55–70 dB | Depends on the pressing, cartridge and playback chain. |
| Consumer digital audio equipment | ≈ 90+ dB | Specifications are often A-weighted, which makes figures look better than unweighted ones. |
| 24-bit audio converter | Theoretical ≈ 146 dB; real devices roughly 110–125 dB | Analog circuit noise, not quantization, is the practical limit. |
| Digital imaging (ISO 12232 convention) | Amplitude SNR 10:1 (20 dB) "acceptable"; 40:1 (≈ 32 dB) "excellent" | Used in the noise-based ISO speed definition. The ratios are amplitude ratios, so dB = 20·log10. |
| Analog video | Roughly 40+ dB looks clean; below about 30 dB noise ("snow") is clearly visible | Approximate. Values are usually weighted and depend on the broadcast standard. |
ADC ideal SNR
For an ideal N-bit analog-to-digital converter with a full-scale sine input, the quantization-limited SNR is
$$\text{SNR}_{\text{dB}} \approx 6.02\,N + 1.76$$
| Resolution | Ideal SNR |
|---|---|
| 8-bit | ≈ 49.9 dB |
| 10-bit | ≈ 62.0 dB |
| 12-bit | ≈ 74.0 dB |
| 14-bit | ≈ 86.0 dB |
| 16-bit | ≈ 98.1 dB |
Real converters fall short of these values because of thermal noise, clock jitter and distortion. The usual way to express this is the effective number of bits (ENOB).
dB to Power Ratio Conversion Table
Power ratio: \(P_s/P_n = 10^{\text{SNR}_{\text{dB}}/10}\). Amplitude (voltage) ratio: \(V_s/V_n = 10^{\text{SNR}_{\text{dB}}/20}\), which is the square root of the power ratio. That is why voltage uses 20·log10 while power uses 10·log10.
| SNR (dB) | Power ratio (Ps/Pn) | Amplitude ratio (Vs/Vn) |
|---|---|---|
| −20 | 0.01 | 0.1 |
| −10 | 0.1 | ≈ 0.316 |
| −3 | ≈ 0.501 (about ½) | ≈ 0.708 |
| 0 | 1 | 1 |
| 3 | ≈ 1.995 (about 2) | ≈ 1.413 |
| 6 | ≈ 3.98 (about 4) | ≈ 1.995 (about 2) |
| 10 | 10 | ≈ 3.162 |
| 20 | 100 | 10 |
| 30 | 1,000 | ≈ 31.62 |
| 40 | 10,000 | 100 |
| 60 | 1,000,000 | 1,000 |
Handy shortcuts: every +10 dB multiplies the power ratio by 10, and every +3 dB roughly doubles it. A negative SNR means the noise power exceeds the signal power.
SNR Comparison Across Example Scenarios
Each row uses \(\text{SNR}_{\text{dB}} = 10\log_{10}(P_s/P_n)\). Both powers must be in the same unit, so 1 W is converted to 1000 mW.
| Scenario | Signal power | Noise power | Power ratio | SNR (dB) |
|---|---|---|---|---|
| A: Strong signal | 100 mW | 1 mW | 100 | 20 |
| B: Very strong signal | 1 W (1000 mW) | 1 mW | 1,000 | 30 |
| C: Equal powers | 10 mW | 10 mW | 1 | 0 |
| D: Noise-dominated | 1 mW | 10 mW | 0.1 | -10 |
| E: Same ratio as A | 50 mW | 0.5 mW | 100 | 20 |
Worked check (scenario D)
$$\text{SNR}_{\text{dB}} = 10\log_{10}\!\left(\frac{1}{10}\right) = 10 \times (-1) = -10\ \text{dB}$$
Scenarios A and E have different absolute powers but the same 100:1 ratio, so they have the same SNR. SNR depends only on the ratio, not on the absolute power level. Equal powers always give 0 dB, and any noise-dominated signal gives a negative value.
FAQ
What is a good SNR? It depends on the application. For audio, 60+ dB is excellent; for many digital communications, even 10–20 dB can be acceptable with error correction.
Can SNR be negative? Yes. If the noise power exceeds the signal power, the ratio is below 1 and the dB value becomes negative, meaning the signal is buried in noise.
Do I need to use watts? No — any consistent power unit works because SNR is a ratio. Just make sure both inputs use the same unit.