August 2019

Forget The Specs
Roger Skoff writes something that could change your mind completely.
Article By Roger Skoff

  Until not very long ago, "spec sheets" were an important part of our audio hobby. You know what I'm talking about – those single-sheet write-ups that showed a picture of a product, gave a description of its features and qualities, and set down in detail such things as frequency response, measured distortion, signal-to-noise ("S/N") ratio, and so on.

People used to pick them up at Hi-Fi Shows as reminders about products that had particularly caught their attention, or they would ask for them, supposedly "to show the Little Lady", as "get-out-of-the-store without-buying-anything" passes, when they'd auditioned something at their local HiFi dealer and either didn't want, or weren't yet ready, to buy them.  There were even guys – like those sports fans who know every bit of statistical information there is about a player or a team but never go to a game – who just collected product "literature" so they could knowledgeably engage in hours-long bull sessions with their Hi-Fi Crazy friends, but really had no intention of ever buying any of the stuff it described.

Nowadays, it seems like more and more manufacturers are just posting their spec sheets on the internet instead of having them printed. And even the magazines – many of which used to publish specifications along with every review – are leaving the 'specs out and going ever more purely subjective.

Is this a loss? Are audiophiles being denied information they need in order to make informed buying decisions? I don't think so.

For one thing, many of the specifications that used to be published were simply meaningless: What possible good does it do a potential buyer to know that a speaker, a system, or a piece of audio gear has a frequency range of whatever it might be (even 10Hz to 40kHz) if no frequency response "curve" is supplied and no numerical information is provided as to how "flat" (within plus or minus how many decibels) that frequency response is claimed to be? If it's +/- 60dB, for example, it's near certain that the frequencies at the extremes, even if measurable, won't be audible. So what good is having them?

And even if the frequency response is clearly stated as being within the (customary) plus or minus 3dB range above or below whatever "reference level" has been selected, that still says mostly nothing: It could either mean what we generally assume to be the case – that it's mostly flat, with a 3dB peak somewhere in the midrange and with 3dB of "roll-off" at both frequency extremes – or it might be perfectly flat (at whatever "reference" level) across most of its range, with either a (very definitely noticeable) 6dB peak or a 6dB "suck-out" somewhere along the way. It could even, for that matter, have a "peak" or "valley" of plus or minus 6dB every half-octave (or at any other interval, regular or irregular) constantly repeating itself and screwing-up your sound all across the frequency range. Just the numbers, without an actual curve or, better yet, a MLSSA display, don't really tell us anything useful.

It's not just frequency response specs that can be meaningless, either: Probably the most meaningless specification of all is wattage. Even though some of the, either more ignorant or more deceptive, mass marketers used to advertise speakers they offered as "having" however many watts, ("Wow! 300 Watts! Yeah!") wattage has to nothing at all to do with sound quality, and really doesn't even have much to do with how loud your music is going to play.

Loudness, as we hear it, is a logarithmic function that, weirdly enough, has very little to do with how many watts are going to the speakers. From whatever level you're listening at (the "reference" level), doubling the wattage to the speakers will produce exactly a three decibel (3dB, about the minimum easily perceived) difference in loudness), and that difference will always be exactly the same whether you double from one watt of amplifier output to two or from one hundred Watts to two hundred.

Given that either a one watt change or a change of a hundred watts can produce exactly the same difference in sound level, it should be obvious that the actual number of watts to the speakers isn't the most important factor in determining loudness. What's much more important is loudspeaker "sensitivity" – usually stated as how loud a speaker will measure at 1 kHz from an on-axis distance of one meter.

My Acoustat 1+1s come in at 82dB/W/m – quite insensitive, as compared to most home speakers, which seem usually to fall into the 92dB@1W/1M range, and some horn-type "pro" speakers, which will measure over 102dB/W/m – a full 20dB louder.  This has nothing at all to do with sound quality, but it does mean that to achieve the same loudness levels as those 102dB/W/m horns will deliver from one watt, the average home speaker will require 10 Watts of amplifier power and my Acoustats will require 100! It's also the reason why speaker power-handling capacity matters (The speakers have to be able to take enough power to play as loudly [or maybe even a little more so] than you want them to), and it's one of the two major reasons why amplifier output power matters: 

1) To provide enough power to drive your speakers loud enough, with enough left over to meet musical peak power demands, and...

2) Depending on how it "clips" when it "runs out of steam", a higher powered amplifier may be less likely to blow your tweeters by playing them too loudly than a lower-powered amp is.

Other than for those reasons, if your system is loud enough for you, your speakers' power-handling capacity or your amplifier's power output really don't matter much at all.

Neither do most of the other "specs" available for your system: The fact of it is that every published specification for every piece of gear is just a measurement of what it will do by itself, NOT in your system or your room. Once you start hooking things together to make a functioning system, you expose those things to potential problems of impedance match (both input and output), characteristic impedance, polarity, hum and noise, and all of the other things that can make a system sound different than the manufacturers of its component parts intended. And once you put that completed system into a listening room, you're stuck with other issues -- of room acoustics, resonances, speaker placement, listening position, and any number of other things that can make the measured response at any given point in your room nothing like what it's indicated to be in the manufacturer's specifications.

So what can you do about it?  Don't just rely on published specifications. Listen to things, preferably before you buy them, preferably in your own home system. (See if your dealer will let you try things at home before you buy them or if you can return them for full credit within some specified period of time if they don't work for you. You'll never know if you don't ask!)

Once you've settled on all the elements of your system, take the time and make the effort to set them up as well as you can, making sure to set up your room, its acoustics, and your listening position in the process. Then, put on some tunes, sit back, close your eyes, and...

 

Enjoy the music!

-- Roger Skoff