Why 100 Watts Is Not Twice as Loud as 50 Watts

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

No. Doubling the power adds three decibels, and three decibels is a change most people describe as slightly louder rather than louder. The step that listeners usually call twice as loud is about ten decibels, and ten decibels is ten times the power, so the honest answer to the question is that you would need something near a 500 watt amplifier to make a 50 watt one sound twice as loud, driving the same speaker in the same room.

That is only half of it, though, and the second half is the part that decides what you should actually buy. Loudness is not a property of the amplifier. It is the sum of two terms: how efficiently the speaker converts watts into sound, and the logarithm of how many watts you send it. The first term is a constant called sensitivity, it is published as decibels at one watt at one metre, and it swamps any amplifier decision you are likely to make.

The two terms, and which one you can actually move

Write it out. Sound pressure level at one metre equals the sensitivity plus ten times the logarithm of the power in watts. Everything about the question falls out of that single line.

Put one watt in and the logarithm term is zero, so the speaker produces exactly its published sensitivity. Double to two watts and the term is ten times the logarithm of two, which is three decibels. Go to four watts and it is six. Ten watts gives ten. A hundred watts gives twenty. Each doubling is worth the same three decibels no matter where you start, which is why going from 50 to 100 watts and going from 500 to 1000 watts buy you the identical, modest improvement.

Now put the other term next to it. Consumer speakers commonly sit somewhere between 84 and 94 decibels at one watt at one metre, and a six decibel spread inside that range is ordinary. Six decibels is four doublings of power. So choosing an 87 decibel speaker over a 93 decibel one commits you to buying four times the amplifier forever, in exchange for nothing. Work it the other way and the arithmetic is brutal: a 93 decibel speaker on 10 watts hits 103 decibels at a metre, and an 87 decibel speaker needs just under 40 watts to arrive at the same 103. The drawing below is that sentence as a graph.

The two lines in that drawing are parallel because the power term is added to the sensitivity rather than multiplied by it. Parallel lines never meet. No amplifier you can buy closes a sensitivity gap; it only slides you along your own line. That is the single most useful thing to know before spending money on either component.

Why a 50 watt amplifier running at one watt still runs out

If three watts is enough for comfortable listening, why does anyone sell a 300 watt amplifier? Because music is not a steady tone, and the rating is not about loudness at all. It is about peaks.

The ratio between the loudest instant and the average level is the crest factor. Recorded music typically runs ten to twenty decibels of it, more on a sparse acoustic recording and less on something heavily compressed. Ten decibels of crest factor means the peaks demand ten times the average power. Twenty decibels means a hundred times. So an amplifier sitting at a one watt average needs ten watts of instantaneous headroom to pass a modest peak cleanly, and a hundred watts to pass a large one. The rating is sized for the loudest tenth of a second in the track, and the average draw during the rest of it is a tiny fraction of that.

This is why amplifier power still matters while being almost irrelevant to loudness. Run out of headroom and the peak does not get quieter; it gets clipped, the waveform tops flatten, and you hear harsh edges and a hardening on transients. The fix is headroom, not volume. And headroom is exactly what sensitivity buys you for free: every three decibels of speaker sensitivity halves the watts your peaks need.

What our own audio products publish, and what they leave out

Here is the honest position across the site, and it is not flattering. Not one published product on ThePickpedia gives a sensitivity figure in decibels at one watt at one metre. The number that governs loudness is absent from every listing we carry.

What one of them does publish is the power side, properly specified. A Juke Audio Juke-6 six zone amplifier states 40 watts per channel into 8 ohms across up to twelve channels, and separately states impedance stability down to 2 ohms per channel. Both halves of that are meaningful, and the second half is rarer than the first. Forty watts into 8 ohms is about 17.9 volts across the load drawing 2.24 amps. Hold the voltage and drop the load to 2 ohms and the same swing asks for 8.94 amps and would deliver 160 watts, which is why a current limit stated as an impedance floor is the useful way to publish this. For distributed audio the 40 watts is ample, because the arithmetic above says 40 watts is 16 decibels above the one watt figure, and a ceiling speaker only has to fill a room rather than win an argument. Our own record flags that the listing states no manufacturer warranty, which is a real gap on an installed component, and that the 40 watts suits distributed audio rather than a main listening room. Both of those are fair. Full details in our Juke Audio Juke-6 review.

Then the silence, and one listing is worth singling out because it is total. An Andover Audio Spinbase 2 turntable speaker base publishes no power rating, no driver details and no dimensions at all, on a product a turntable has to physically sit on top of. Its amplification is named Dual Direct Drive and never explained. Our own record lists every one of those as a drawback, which is the right call, and it means there is no way to place this product anywhere on the graph above. See our Andover Audio Spinbase 2 review.

At the small end the question changes shape rather than going away. A Luzsco Bluetooth shower speaker publishes no power and no sensitivity, and for a sealed single driver mini enclosure both figures would be close to meaningless anyway: its output is limited by cone excursion and cabinet volume long before any amplifier runs out, and the published figure that actually constrains it is the IPX4 rating, which our record is careful to state is splash resistant rather than waterproof. Details in the Luzsco shower speaker review. The same applies to a pair of Defunc True Basic open fit earbuds, where the listing publishes 22 hours of total playtime and the fit, and the honest driver of perceived loudness and bass is that an open fit does not seal the ear canal, so low frequencies leak out and no amount of power replaces the seal. Our record says exactly that, in its own words, as a drawback. See the Defunc True Basic review. In both cases the right reading is that sensitivity is the wrong question for a sealed portable, and the sealing surface is the real one.

One thing we have said on this site that needs a number attached

Our speaker selector switches guide compares its picks across "zone count, power handling, impedance protection, and wiring compatibility", and it does explain the load problem rather than leaving it hanging. It says plainly that "each added pair lowers the total load your amp sees, which can overheat it", and it is honest enough to add that "the impedance protection can slightly reduce output to attached speakers, so demanding listeners may notice less headroom at high volume". That is correct on both counts and we should not pretend otherwise.

What is missing is the size of the word slightly, and this page can supply it, because the cost of that protection is measured in decibels. The cheap way to protect an amplifier from a low load is a resistor in series with each zone. Put a 4 ohm resistor in front of an 8 ohm speaker and the speaker receives the fraction 8 over 12, squared, of the power, which is 0.444, which is 3.5 decibels down. A 2 ohm resistor costs 1.9 decibels. Translate that through the graph above: 3.5 decibels is most of the way to halving your available power, so the switch that lets you add a second room has quietly taken a step and a half of amplifier doubling away from both rooms. It also raises the source impedance the driver sees, which loosens the electrical damping on the woofer and is heard as a softer, less controlled bass rather than as lower volume.

That is also the clearest way to read the Juke-6 against a selector switch. The switch solves the problem by throwing power away. The Juke-6 solves it by giving each zone its own amplifier channel and publishing an impedance floor instead, so nothing is dissipated in a resistor. We do not carry a speaker selector switch as a published product, only that guide, so there is no listing to check this against on our own site.

The honest limit

Nobody in this part of the market publishes sensitivity, so the number this whole page turns on is not available to you on any product we list. If you are buying passive speakers from a maker who does publish it, the figure is worth more than any amplifier comparison you could run. If you are not, treat the watts as a headroom specification and stop reading them as a loudness specification.

And then do the free thing first, because it is worth more than the upgrade. Sound level falls with distance, about six decibels per doubling in the open, so halving your listening distance is worth the same as four times the amplifier and costs nothing. Putting a speaker against a wall couples it to the boundary and typically returns a few decibels in the bass for free; putting it into a corner couples it to two or three surfaces and returns more, at the cost of some muddiness you may or may not mind. Lifting a speaker off a soft surface and onto something rigid stops the cabinet losing energy into the furniture. Pointing it at your head rather than past it is worth several decibels at the top end on its own, because dispersion narrows with frequency.

Taken together, placement and distance routinely move the level at your ear by more than the difference between any two amplifiers you were choosing between. Do all of that, listen, and only then decide whether you have a power problem. If peaks still harden up when you turn it up, that is a headroom problem and more watts is the right answer. If it is simply not loud enough at any setting without hardening, the speaker is the component you are fighting, and no amplifier will move its line.