Signal to Noise Ratio

When you tune across a crowded HF band and struggle to copy a weak DX station through the hash and hiss, what you are experiencing is poor signal-to-noise ratio. The station is there — photons are arriving at your antenna — but the noise is so close in level to the signal that your brain cannot extract the intelligence. Signal-to-noise ratio, universally abbreviated SNR, is the single most important figure for describing how well a receiver can recover a signal. Every other receiver specification — noise figure, noise floor, dynamic range — ultimately exists to tell you something about SNR under specific conditions.

Unlike raw signal strength, SNR is a relative measurement. A strong signal is worthless if the noise is equally strong. A weak signal can be perfectly readable if the noise is even weaker. This lesson explains what SNR means, how to calculate it, what SNR is required for each operating mode, how the S-meter relates (and does not relate) to it, and what you can do to improve it in your station.

What is SNR? The Fundamental Definition

Signal-to-noise ratio is the ratio of the power of the desired signal to the power of the noise measured at the same point in the receiver chain. It answers the question: how many times stronger is the signal than the noise? Expressing this ratio as a plain number is unwieldy because useful SNR values range from less than 0.001 (signal buried deep in noise) to more than 100,000 (crystal-clear signal far above the noise floor). Logarithms compress this range into manageable numbers, which is why SNR is almost always stated in decibels.

The decibel formula for SNR is:

This is one of the great conveniences of the dBm scale: subtraction of dBm values gives you SNR directly. If your signal is at −90 dBm and your noise floor is at −120 dBm, the SNR is simply −90 − (−120) = 30 dB. No logarithms to calculate by hand.

Interpreting SNR values

A few reference points help anchor your intuition:

• 0 dB SNR: Signal power equals noise power. The signal is right at the noise floor. Detection is barely possible and intelligibility is essentially zero for voice.
• 3 dB SNR: Signal power is twice the noise power. Still very poor — a trained CW operator might detect a few characters but would not copy a message reliably.
• 10 dB SNR: Signal power is 10 times the noise power. Marginal SSB voice copy; good CW copy for an experienced operator.
• 20 dB SNR: Signal power is 100 times the noise power. Comfortable copy on SSB voice; solid copy on all CW speeds.
• 30 dB SNR: Signal power is 1000 times the noise power. Excellent audio quality. Even AM sounds good at this level.

Negative SNR values — where the signal is below the noise floor — are not only possible but are routinely exploited by digital modes. A signal at −10 dB SNR has one-tenth the power of the surrounding noise, yet FT8 can decode it with high reliability. This remarkable capability arises from coherent signal processing techniques covered in the digital modes section below.

Worked example: calculating SNR from two dBm readings

The power ratio form of SNR

Sometimes you need to convert dB SNR back to a plain power ratio, for example when calculating required signal levels for a specific performance target. The formula is:

The factor of 10 in the power formula versus 20 in the voltage formula reflects the fact that power scales as voltage squared. A 20 dB SNR means a power ratio of 10² = 100 and a voltage ratio of 10¹ = 10 simultaneously — both are correct descriptions of the same signal condition.

High SNR (left): the signal peak rises clearly above the noise floor. Low SNR (right): the signal is barely distinguishable from the noise. The SNR in dB is the gap between signal level and noise floor.

View Larger

SNR Thresholds by Operating Mode

Different operating modes have radically different SNR requirements. The key factor is how efficiently each mode uses the available information in a received signal. Traditional voice modes like SSB require relatively high SNR because the human ear must directly perceive and understand the audio. CW (Morse code) is far more efficient because the brain only needs to detect whether a tone is present or absent, not decode complex speech. Digital modes using coherent detection and error correction can operate at SNR values that would make a voice signal completely inaudible.

Understanding these thresholds matters enormously for practical operating decisions. If you are calling CQ on 40m SSB with 100 watts into a dipole and receiving no responses, the problem might be that stations hearing you are at 8 dB SNR — just below the intelligibility threshold — and simply cannot copy your callsign. Switching to FT8 would let you work those same stations even if their signal at your location is 15 dB below your noise floor.

*FM has a characteristic called the capture threshold. Below a certain carrier-to-noise ratio (typically 10–12 dB), an FM receiver produces very poor audio because the demodulator locks on to the noise rather than the signal. Above the threshold, the receiver captures the signal and rejects noise very effectively, producing audio that sounds nearly perfect even in relatively noisy conditions. This is very different from SSB, which degrades gradually as SNR falls.

Why digital modes can work far below 0 dB SNR

The seemingly magical ability of FT8 to decode signals at −10 dB SNR, or WSPR at −28 dB SNR, comes from a principle called processing gain. When you know exactly what you are looking for — the precise frequency, the exact timing, and the modulation scheme — you can integrate the received signal over many symbol periods and average down the noise.

Noise is random. When you add many noise samples together, they partially cancel because some are positive and some are negative. The root-mean-square noise grows only as the square root of the number of samples added, while a coherent signal grows linearly. This means every time you quadruple the integration time, you gain 6 dB of SNR improvement. FT8 uses 15-second transmission windows; WSPR uses 110.6 seconds. The WSPR integration advantage over a 1-second CW character is approximately 10 × log10(110.6) ≈ 20 dB — explaining most of the difference in their operating thresholds.

S-Meter Calibration and SNR

The S-meter on your transceiver is one of the most commonly misunderstood instruments in the ham shack. It reads signal strength — but not SNR. Understanding the difference, and knowing what S-meter readings actually represent in dBm, lets you extract much more useful information from your radio.

The IARU S-meter standard

The International Amateur Radio Union (IARU) defined a standard for S-meter calibration that most modern transceivers follow, at least approximately. The reference point is:

From this single reference, you can calculate the absolute signal level corresponding to any S-meter reading:

S-meter reading versus SNR

An S9 signal does not mean good SNR. It means the signal is at −73 dBm. Whether you can copy that signal depends entirely on your noise floor. Consider these scenarios on 40m at night, both with an S9 signal arriving at the antenna:

• Rural location, quiet night: Noise floor −115 dBm. SNR = −73 − (−115) = 42 dB. Crystal-clear audio.
• Urban location, summer evening: Noise floor −85 dBm due to electrical interference and atmospheric noise. SNR = −73 − (−85) = 12 dB. Barely readable voice.
• Near a lightning storm: Noise floor −70 dBm from static crashes. SNR = −73 − (−70) = −3 dB. Essentially unreadable.

This is why experienced operators never assess signal quality from the S-meter alone. An S7 signal in a quiet rural environment may be far more readable than an S9 signal at a noisy urban location.

Why S-meters disagree between radios

Despite the IARU standard, S-meter calibration varies widely between radios. Manufacturers often inflate S-meter readings to make signals appear stronger (customers feel good about their purchase). An S9 on one transceiver may be only S7 on another, and −73 dBm may not be S9 on either one. For anything more than a rough guide, do not rely on S-meter readings for absolute comparisons between stations or between radios. Use a calibrated signal generator and a known attenuator to verify your radio's S-meter if accuracy matters.

SINAD — Signal Plus Noise Plus Distortion to Noise Plus Distortion

SINAD (pronounced "sigh-nad") is a variant SNR measurement used specifically for FM voice receivers. It accounts not only for noise but also for distortion in the received audio — both of which degrade intelligibility. The full name reveals the formula:

The reason SINAD is used rather than simple SNR for FM testing is that FM receivers produce distortion products when the carrier-to-noise ratio is low — particularly near the FM capture threshold. A receiver might have seemingly adequate SNR on a pure-noise basis but still produce highly distorted audio due to FM demodulator non-linearities. SINAD captures both effects simultaneously.

The 12 dB SINAD sensitivity standard

The standard sensitivity specification for FM voice receivers — including the receivers in most 2m and 70cm transceivers — is the input level required to produce 12 dB SINAD. This corresponds to roughly 40% audio distortion, which sounds terrible by high-fidelity standards but represents the threshold of usable communications. It is the point where an operator can just barely understand transmitted speech.

Typical specifications for modern VHF/UHF FM transceivers:

• 2m FM sensitivity (12 dB SINAD): 0.18–0.35 μV (approximately −122 to −116 dBm)
• 70cm FM sensitivity (12 dB SINAD): 0.25–0.50 μV (approximately −120 to −113 dBm)
• Commercial/public-safety FM: often specified at 0.25 μV for 12 dB SINAD

SINAD versus SNR: which should you use?

Use SNR when discussing HF receivers, which are primarily noise-limited (not distortion-limited) and where the main modes are SSB, CW, and digital. Use SINAD when discussing VHF/UHF FM receivers, where the capture threshold and FM distortion behavior are important. Many VHF/UHF transceiver data sheets specify SINAD exclusively because FM distortion is a more relevant performance limiter than simple noise at those frequencies.

Improving SNR — What Works and What Does Not

Every operator wants to copy weaker signals, which ultimately means improving SNR. But not every approach works equally well, and some popular "improvements" have no effect or even make things worse. Understanding why requires understanding what is limiting your SNR in a given situation.

1. Narrow filters — the most effective tool

The noise power in a receiver is proportional to bandwidth. Halve the bandwidth and you halve the noise power — a 3 dB SNR improvement. This relationship makes narrow filters the single most powerful SNR improvement tool available in a conventional receiver.

In practice: switching from a 2.4 kHz SSB filter to a 500 Hz CW filter when copying CW reduces noise power by 10 × log10(2400/500) = 6.8 dB. This is a major improvement. Switching to a 50 Hz digital mode filter gives another 10 dB over the 500 Hz CW filter. This is why CW operators can copy signals that would be completely inaudible on SSB even though they are using the same signal amplitude — they are using a much narrower effective bandwidth.

2. Better antenna — more signal, same noise

An antenna with more gain receives more signal power from the desired direction. If the noise arriving at your receiver comes primarily from the antenna (external atmospheric and man-made noise) rather than from the receiver itself, a higher-gain antenna improves SNR because it picks up more signal without picking up proportionally more noise from other directions. A 3-element Yagi with 7 dBd gain over a dipole will improve SNR by approximately 7 dB on signals in its main lobe.

However, this only works if the antenna is pointed at the signal and the noise is coming from other directions. If your primary noise source is broadband atmospheric noise equally distributed around the horizon, a horizontal Yagi will improve SNR significantly compared to an omnidirectional antenna. If your primary noise is from a power line immediately behind your house, pointing the Yagi away from it helps. An omnidirectional vertical with a noise-blanker may be more practical in many installations.

3. Low-noise preamplifier — only helps when receiver-noise limited

A low-noise amplifier (LNA) placed at the antenna reduces the effective system noise figure by improving the signal level before the lossy coaxial feedline. This only helps SNR when the limiting noise source is the receiver itself (including coax losses), not external atmospheric or man-made noise. On VHF and UHF weak-signal work, where the sky temperature is very low and man-made noise is minimal, a good LNA can make a dramatic difference. On 40m at night with high atmospheric noise, adding an LNA is largely pointless because the noise floor is set by external noise, not by the receiver electronics.

4. More transmit power — helps the other station's SNR

Doubling your transmit power (+3 dB) increases the signal level at the other station's receiver by 3 dB, improving their SNR by 3 dB. This is straightforward and effective. Note that it has no effect on your receive SNR — only on what you transmit.

From S9 to S9+40 dB is a 40 dB increase in signal level, requiring a 10,000-fold increase in transmit power (from 5 W to 50,000 W). This illustrates why transmit power has diminishing returns. Going from 100 W to 1500 W (+12 dB) is an impressive gain, but it only helps stations who are close to the copy threshold. For stations much below threshold, even 1500 W at your end leaves them unable to copy. Mode changes to digital are usually far more productive.

5. Digital modes — processing gain

As described above, digital modes like FT8 and WSPR achieve effective SNR improvements of 20–30 dB over voice modes through coherent detection and time integration. This is not magic — it comes at the cost of throughput (FT8 conveys about 10 characters of meaningful information per 15-second transmission) and latency. But for making contacts at the edge of band conditions, no hardware improvement comes close to the gain from switching to a digital mode.

Ham Operating Context: RST Reporting and SNR

The Readability-Strength-Tone (RST) system is the standard reporting format in ham radio. The R (readability) component most directly reflects SNR, while S (signal strength) corresponds to the S-meter reading discussed earlier. Understanding what RST reports actually mean helps you calibrate reports you give and receive.

The Readability scale

The key insight: an R5S9 report (often given as "59") means the signal is strong AND perfectly readable — both signal level and SNR are excellent. An R3S9 report would be unusual but meaningful: the signal is physically strong (S9) but barely readable, suggesting severe interference, heavy fading, or very high local noise at the receiving station. A R5S3 report is more common in real conditions: the signal is weak but the band is quiet and the signal is perfectly copy-able.

59 vs. 55 vs. 33: practical implications

In a casual QSO, a 59 report is nearly always given regardless of actual conditions — it is considered polite. In contest and DX work, accurate reports matter. A genuine 33 report tells the DX station that their signal is barely audible and conditions are poor. They may respond by asking you to QSY, switching to CW, or repeating more slowly. An accurate 55 report — readable but not strong — tells them the band is open but marginal.

For DX operating, the key threshold is whether SNR is sufficient to confidently exchange callsigns. The DX station is typically running high power with a good antenna; the limit is almost always at the DX end, where the pileup of callers creates significant interference that raises the effective noise floor. The SNR required to work a DX station in a pileup is much higher than the SNR for a quiet two-way QSO because you must be distinguishable from many other calling stations.