WATT, power and loudness

WATT is one of the most abused numbers in audio marketing. Electrical input power matters, but it is only one part of an electroacoustic system. To estimate useful acoustic output we must also understand sensitivity, efficiency, excursion, thermal limits, amplifier voltage/current capability, DSP protection and the measurement method behind a claimed maximum SPL.

1. What does WATT mean?

The WATT is the SI unit of power: one joule per second. For a purely resistive load, P=U²/R=I²R. A loudspeaker is a frequency-dependent complex load, so instantaneous and average power depend on voltage, current and phase. A single “WATT” number therefore does not describe the acoustic output of the system.

2. Sensitivity and efficiency

Sensitivity expresses acoustic output for a defined electrical stimulus and distance, commonly dB SPL at 1 m for 2.83 V or 1 W. These are not identical for every nominal impedance. At 8 Ω, 2.83 V corresponds to about 1 W; at 4 Ω it is about 2 W, creating a 3 dB trap if specifications are compared blindly.

Electroacoustic efficiency η₀ is the fraction of electrical input converted to acoustic power. Direct-radiating moving-coil drivers are commonly around 0.5-3%, while horn-loaded systems can be substantially higher. A Thiele-Small estimate of reference efficiency in half space is often written approximately as:

Symbols: η₀ is dimensionless reference efficiency, f_s is in hertz, V_as,L in litres and Q_es electrical quality factor. The coefficient 9.64×10^-10 is valid specifically for those units; using cubic metres requires a different coefficient.

This illustrates Hofmann’s “iron law”: deep low-frequency extension, small enclosure volume and high efficiency cannot all be maximised independently.

3. Why theoretical +3 dB does not always happen

Doubling electrical power gives +3.01 dB only while the loudspeaker remains linear. Voice-coil heating raises resistance, suspension and motor parameters become excursion-dependent, the amplifier may hit voltage or current limits, and DSP limiters may intervene. At high drive levels, acoustic output often grows by less than the small-signal prediction.

4. Peak, RMS, “program” and AES power ratings

Power ratings are meaningful only with a test standard, bandwidth, crest factor and duration. Professional driver testing such as AES2 uses specified noise, filtering and duration rather than an undefined “peak power” number. Marketing “peak” ratings may simply reflect a theoretical instantaneous voltage limit that says little about sustained operation.

5. Crest factor - music is not a sine wave

Music has transient peaks above its average level. Crest factor is the ratio between peak and RMS values. A system sized only for average power can clip on transients; a system sized only from a giant peak WATT number may have inadequate thermal capability. Useful design separates peak headroom from long-term energy.

6. An active loudspeaker is an amplifier-DSP-transducer system

In an active box the power-amplifier rating cannot be evaluated independently. DSP crossover, EQ, excursion control, voltage limiting and thermal limiting define how much of that amplifier capability reaches each driver and for how long. A lower stated amplifier WATT rating can outperform a larger one when the transducers and processing are better matched.

7. Maximum SPL - but how was it measured?

A max-SPL specification needs test distance, bandwidth, duration, distortion criterion, weighting and whether the value is calculated or measured. “Peak SPL” may be a short burst; continuous broadband output is a different question. Compare like with like.

8. Why a lower-WATT system can be louder

If loudspeaker A is 6 dB more sensitive than loudspeaker B, it needs only one quarter of the electrical power for the same small-signal SPL. Efficiency, horn loading, directivity and bandwidth can outweigh dramatic amplifier-power numbers.

9. Low frequencies are especially expensive

For a piston radiator, maintaining the same far-field pressure as frequency falls requires rapidly increasing volume acceleration and excursion. Roughly, for constant pressure in the piston region, excursion scales as 1/f². One octave lower frequency therefore requires about four times the excursion, all else equal. This is why subwoofer design is constrained by displacement as much as by WATT.

10. What should we read in a datasheet?

• frequency response with stated tolerance;
• horizontal and vertical directivity;
• measured/defined maximum SPL;
• distortion or test criterion where available;
• driver size and excursion capability;
• limiter and DSP architecture;
• power test standard and thermal conditions.

The amplifier WATT rating is one line among many, not the verdict.

11. Amplifier output limits: voltage and current

At higher load impedance, maximum output is often voltage-rail limited. At low impedance, transistor current, power supply and SOA protection become dominant. In simplified form:

Symbols: P_V-limit is power estimated from the voltage limit, P_I-limit from the current limit and R_L the simplified resistive load. Real loudspeaker impedance varies with frequency, so these are approximate boundaries.

A “2 Ω capable” amplifier therefore needs more than a large voltage swing; it needs sustained current and thermal margin into reactive loads.

12. What does “2 × 1000 W” really mean?

It may mean two channels measured simultaneously, one channel at a time, short burst, continuous sine, a particular load or a particular THD threshold. Multi-channel amplifiers can also be limited by a shared power supply. Without the test conditions, the number is incomplete.

13. Mains power and acoustic output

An active loudspeaker that claims several thousand W of internal amplifier power does not draw that amount continuously from the mains during normal music reproduction. Class-D efficiency, signal crest factor, duty cycle and power-supply storage matter. Conversely, low average mains consumption does not mean weak transient capability.

14. Sensitivity versus maximum SPL - small signal versus large signal

Sensitivity is usually a small-signal metric. Maximum SPL is a large-signal system metric. Between them lie thermal compression, motor nonlinearity, suspension limits, DSP limiting and amplifier constraints. Extrapolating max SPL by simply adding 10log(P) to sensitivity can be useful as a first estimate but not as a substitute for measurement.

15. Power compression is frequency-dependent

Voice-coil resistance rises with temperature:

Symbols: R(T) is coil resistance at temperature T, R₀ resistance at reference temperature T₀ and α_Cu the approximate temperature coefficient of copper. The rising resistance of a hot coil contributes to power compression.

A hot coil can increase resistance dramatically, reducing current and acoustic output. At 150-200 °C, a 6 Ω DC resistance can rise toward 10-11 Ω depending on actual temperature and materials, producing several dB of compression. Mechanical and magnetic compression mechanisms add further frequency dependence.

16. A fair real-world comparison of two active loudspeakers

Place them in the same acoustic environment, same position and orientation, match input conditions, measure broadband and band-limited SPL at useful distances, examine limiting behaviour and distortion, and listen at matched level. Then the relevant question becomes “which system delivers the required coverage, headroom and quality?” rather than “which has more WATT?”