Notice: Undefined variable: raticle_ordinal in /usr/home/dlamkins/public_html/lamkins-guitar/struct-links.phpi on line 25
http://lamkins-guitar.com/music/article/lies-damn-lies-and-specs
David Lamkins picked up his first guitar a long time ago. As best he can recall the year was 1967: the year of the Summer of Love. Four decades later David has conjured up an amalgam of folk, rock and jazz solo guitar music for the occasional intimate Portland audience.
LCW on Bandcamp
location: Portland, OR USA

Facets: business, marketing hype, terminology, @musings info
Loading...

Lies, damn lies and specs

As a guitarist you've probably become used to hearing fanciful "cork-sniffer" descriptions of the tone associated with various products. It goes with the territory; the best you can do is to hold your nose, open your ears (and wallet) and hope that the product delivers on its advertised promises. You may have become so used to subjective descriptions that you're inclined to treat objective specifications as gospel truth.

Unfortunately, you can't assume that numbers and charts mean anything. Unless you're well-versed in the underlying technologies - and by "well-versed" I mean that you have an actual understanding of the physics and mathematics - I'd suggest that you ignore any spec that you can't verify. For most people that means shape, color and dimension.

Here are some common specifications and the ways in which marketers abuse your naïveté:

Power
Measured in watts, this is one of the most common technical terms that you'll encounter as a guitarist. It's also one of the most abused by marketers who have invented all kinds of utterly meaningless adjectives with which they decorate a perfectly good unit of measurement to remove its technical meaning and allow them to pick impressive numbers out of thin air.
In audio, the only power measurement that has a standardized meaning is "continuous RMS watts". Any mention of the modifiers "peak", "instantaneous", "program", or "music" in conjunction with a power measurement should be a red flag that you're being told a lie. I've seen quite a bit of discussion about how there are conversion factors from, for example, peak watts to continuous RMS watts. Not so. The only thing you can be reasonably sure of is that any power rating not expressed in continuous RMS watts is probably some integer multiple of the actual power rating. The problem with all of these made-up numbers is that they obfuscate the actual capabilities of the amplifier in order to make it seem more powerful.
The rationale behind these made-up numbers is that the amplifier is said to be capable of short-term outputs much higher than what the amplifier would be able to produce continuously. Some vendors even quantify this by adding some kind of duration to the the number. But in most cases the duration itself is meaningless. For example, when a vendor says that the peak power is delivered over "two cycles", they neglect to specify a measurement frequency. Two cycles of a 20 Hz test tone lasts a lot longer than two cycles of a 15 KHz test tone. At any rate, a huge difference between peak and RMS power just means that the amp's power supply is not well designed.
Power is directly expressible in terms of heat. The more power, the more heat must be dissipated both in the amplifier and its load. If I read a spec that said an amp could produce X amount of power continuously and Y > X power over some limited number of minutes, I'd infer that the amp has thermal-management problems. It's nice of a vendor to disclose that, but not really useful to me as I don't have any way of relating that specification to my program material except by observing when the amp goes into thermal shutdown.
Another important and often overlooked qualifier of a power measurement is the distortion level. Most amplifiers will deliver increasing amounts of power along with increasing distortion. This may be OK for a guitar amp, but if your spec'ing a PA or studio monitors, you want to hear an accurate reproduction of the program material. Even when vendors provide distortion measurements they're often woefully incomplete. For one thing, there's no way (short of actual measurement) to extrapolate from one measurement of distortion and power to some other value of distortion or power. Another caveat is that distortion often varies with frequency (being typically higher at very low and very high frequencies).
Finally, a power rating - even an honest, fully specified rating - is not an indication of loudness. The sound pressure level (SPL) depends upon the power delivered to the speaker and the efficiency of the speaker (again, in a frequency-dependent manner).
Frequency Response
If you thought that power ratings are shady, wait until you hear about frequency response. I'm sure you've seen a spec that specifies frequency response as something along the lines of "20 Hz to 20,000 Hz". Surely that must be good: it covers the entire range of human hearing. But what does that spec actually mean? In truth: nothing at all.
A more common frequency response spec is written like this: "-10 dB at 40 Hz and 18 KHz". The problem with this, aside from the fact that 10 dB down is way down, is that there's no upper bound on the response devation. Response could be up by 10 dB at 2 KHz, for example; this spec says nothing about that. It's called lying by omission; you're given the response at the extremes and expected to assume that everything's as you'd expect in between the extremes. Caveat emptor...
A frequency response spec that states something like "50 Hz to 15 KHz +1 db, -3 dB" is about as complete (and useful) as you'll see short of an actual graph. It gives you a frequency range and tells you that the response does not deviate more than 4 dB over that range.
Distortion
Here's another spec that you can't depend upon. For starters you have the same kind of issues regarding underqualification of the measurement (e.g.: omission of measured levels and frequencies). Beyond that, the last forty or fifty years of audio equipment design have taught us that gear having immeasurable distortion levels doesn't always sound good, nor does gear having modest distortion levels always sound bad. There are lots of different distortion measurements, the most common being Total Harmonic Distortion (THD) and Intermodulation Distortion (IMD). Neither tells you much of anything when you put numbers side-by-side on a spreadsheet. You still have to use your ears...

But what about guitar amps?

Little of this discussion applies to guitar amps, which intentionally create distortion and have non-neutral frequency response.

In terms of power output, tube amps tend to work on a level playing field. There's a certain design "sweet spot" that designers all tend to hit with respect to power output for a given choice of power tubes. You can assume, for example, that most amps using a pair of 6L6s in the output stage will produce 50 watts give or take a little. Remember though that the perceived loudness is heavily dependent upon the speaker and the frequencies emphasized by the amplifier.

The only area where you have to beware of guitar amplifier power specs is in solid-state amps. For some reason these are subject to the same kind of specsmanship I've described above. There's no rule of thumb for derating solid-state amps. A few manufactuers spec honestly; many don't. Short of taking independent measurements, your best bet for dealing with a solid state amp is to try it at volume and find out whether it holds up for the duration of a gig.

September 05 2008 18:45:18 GMT