Wednesday, July 1, 2009

The misleading story of the guitar amp RMS "watt"

One of the greatest misconceptions among users of guitar amps is the notion that the output power of the amp-- usually (mistakenly!) referred to as "wattage"-- is somehow directly correlated to the perceived loudness of the amplifier.

In reality, the amount of voltage and current across the speaker terminals is only one of several factors that help determine how loud the amp will sound to the listener.

Amplifier manufacturers usually market amplifiers with ratings in "RMS watts." RMS stands for "root mean square" and is a statistical measure to calculate the mean, or average, amount of voltage or current delivered into a given load. First of all, it should be pointed out that there is technically no such thing as an "RMS watt." RMS is a way to measure voltage or current, not power. But the notion of an "RMS watt" prevails somehow as the standard by which amplifiers' loudness capacities are judged.

However, eschewing semantics for a moment, one might wonder: Why isn't an amp's power rating the best means of mentally estimating its capability to be loud? And what other factors should I consider?

First, it helps to understand a bit about how the human ear works. The human ear can detect a staggeringly wide range of sound pressure levels. The level of sound pressure that causes permanent hearing damage is more than a million times greater than the sound pressure produced by the faintest sound a human can hear. For this reason, the decibel is a logarithmic unit of measure. An increase of 3dB is equivalent to a doubling of sound pressure in Pascals (Pa), the SI unit of measure for pressure.

What does this mean? Well, for our purposes, it means that all else being equal, a 100 watt Marshall head is not "twice as loud" as a 50w Marshall head, as many seem to think-- it is only a scant three decibels louder.

Further complicating matters is that power ratings are taken with respect to a given "%THD," or percentage total harmonic distortion. If an amplifier is rated 100w with 1% THD, that means it develops 100w across the speaker terminals, with one percent of that power coming in the form of distortion products (i.e. stuff not present in the source input). If it is 100w with 5% THD, then 5% of the output comes in the form of harmonic distortion.

Complicating matters significantly (and impeaching the usefulness of that measurement for our purposes) is that guitarists almost never use their amplifiers with distortion content that low. Even 10% THD or more can sound "clean" in a guitar amp-- with the soft clipping of a tube amp, the onset of distortion into significant percentages manifests itself as a gentle compression or 'fattening' before true clipping occurs.

Beyond this, guitarists often like a little (or a lot) of audible distortion in their sound, so they deliberately induce audible clipping-- likely 50% THD or more. So while a typical 1% THD power rating might be relevant in a hi fi amplifier, it is almost never terribly relevant to actual use in a guitar amplifier.

Besides output power, here is a list of a few factors that can affect a listener's perception of how loud a guitar amplifier is, in no particular order:

  • Number of speakers
  • Efficiency/sensitivity of speakers
  • %THD tolerable to the user/listener
  • Onset of core-saturation in the output transformer
  • Ability of power supply to keep voltage up when amp is pushed
  • voicing/frequency content of amplifier

Let's take them in order, in brief.

First, the number of speakers is critical. Addition of a second, equal speaker, provided the amp will drive the additional load, will usually net an extra 3dB in loudness, roughly equivalent to doubling amplifier power. This is the largest source of the "more watts means louder" myth. A 50w Marshall is often paired with a single 4x12 cabinet while a 100w Marshall is often paired with two 4x12s. No wonder it seems "twice as loud!" A 22w Deluxe Reverb has a single 12" speaker whereas an 85w Twin Reverb has two 12" speakers (and usually very efficient ones, at that). For this reason, 'physical size of the cabinet' is often a more accurate predictor of loudness than the RMS power rating.

Secondly, the efficiency/sensitivity of the speakers matters a lot. Not all speakers are created equally. In a loudspeaker, only a fraction of the current input is turned into sound energy; the vast majority is dissipated as wasted heat energy. Some speakers are more efficient than others. For example, a JBL D120F has approximately 102dB efficiency at 1W white noise @ 1m distance. A less efficient speaker might only have, say, 96dB efficiency at the same power and distance. Consequently, a 25w amp through the JBL could easily sound louder than a 50 or even 60w amp through the less efficient speaker, especially when set up for clean operation.

Thirdly, the farther you are pushing an amplifier past the %THD used for its power rating, the more unpredictably it will behave. Amps do not stop getting louder once they hit 1% THD. They will continue to get louder as well as more distorted, and the nature of those distortion products will influence how loud an amp seems.

How much louder it will get past that point usually has to do with one of the following factors:

Output transformers are varying sizes. Sometimes they are small and will exhibit a behavior called "core saturation." When this happens, the transformer reaches its limits and cannot allow any more current to pass through. It will distort audio passed through it. In other words, it can serve as a "bottleneck" for the power trying to get from the tubes to the speakers. Consequently, a bigger output transformer will often allow an amp to keep getting louder as it is driven into clipping.

As another factor, the power supply feeds the amp the power it needs to amplify sound. If a power transformer, rectifier, or other components within the power supply are limiting the amount of power that can actually be used for amplifying, that can put a ceiling on how loud an amp will get. So a small power transformer might also limit an amp's ability to keep getting louder past the point of rated % THD.

Finally, the voicing of an amplifier can affect how loud it sounds. The Fletcher-Munson curve indicates that certain frequencies in the speech-range appear louder to humans, as our ears are more sensitive to them. An amplifier with a lot of those mid-band frequencies will conceivably seem louder to a human than an amp containing a lot of spectral content outside of that range.

So, in other words, largely ignore the numbers. When gauging whether or not an amplifier will be loud enough (or too loud) for your needs, focus more on the number and size of the speakers-- with speaker efficiency as a secondary factor-- and then the size of the transformers. Bigger output transformers almost always equal the ability to have more headroom and give up more volume even as the tubes saturate. Big power transformers, large filter cap values, and solid-state rectifiers signal a power supply that will keep up and won't mush out before the output tubes flat-line. All of these things are better clues than the rated output as far as determining how loud the amp will be when you crank it up and bounce it off a chair. Of course, you could always just judge by the physical size of the cabinet.


  1. monkeys fly freely while undergarmets made of spider webs are being worn by Michael Jackson's son Blanket. Penguins eat ramen while whistling to elavator music with McSheamus....


  3. I just realized one of my legs is shorter than the other.

    ...and they say I strut around like I'm all that...