Current Issue

May 2008

Print Article     Email Article

Debunking Spec-ology Part 1

by Karl Winkler

Introduction
This is the first in a series about equipment specifications, how to better understand them, and how some manufacturers can obscure the truth. Why is this important? First, audio is a technical area, and we use specifications in our daily work in order to better understand the equipment we are using. Perhaps more importantly, we make choices between similar, competitive products when choosing components and in the design and installation of sound systems.

Years ago, specifications were not as standardized as they are today, and choosing equipment required more testing, and perhaps more faith. However, as the audio equipment business grew and matured, specifications came to be more uniform in order that customers, consultants, and users could make more valid comparisons and better understand what they were getting. Unfortunately, not all manufacturers are straightforward in their approach to listing specifications. In fact, some are deliberately dodgy so that their gear appears to compete with other equipment, when in reality, it may be far inferior, at least in certain applications. In this segment, we’ll look at speakers and microphones.

Watt Did You Say?
The first and most obvious line of inquiry has to do with speakers. If you’re already familiar with this issue, please forgive me, because it is really basic. However, I’ve overheard this very topic being discussed far too often for me to simply chalk it up to just the lack of understanding among newbies. The issue is this: what does the Watt rating on speakers really mean? More times than I can count I’ve heard someone say something like this: “Speaker ‘A’ is 250 Watts while speaker ‘B’ is only 200 Watts, so speaker ‘A’ gets louder.”

You can remain smug from your vantage point of realizing how off-base a comment like this really is. Of COURSE the Watt rating has nothing to do with how loud a speaker is capable of going. But doesn’t power affect how loud a speaker goes? The simple answer is: yes and no. But in this example, so far, we don’t have enough information to know the answer to that question. But let’s say you were feeding speaker “A” with a 250-Watt amp and speaker “B” with a 200-Watt amp. Is it possible that speaker “A” might produce more SPL than speaker “B”? Sure it’s possible. But only if a number of other criteria are met.

First off, let’s look at those two power ratings. How much more power is 250 Watts than 200 Watts? It’s 50 more Watts, right? But is it 25% more? Not in dB, it isn’t. In fact, it represents about 1dB more power. That’s right ONE dB more. To get 3dB more power than that offered by 200 Watts you would need 400 Watts. And for 6dB more power (i.e., what is required to create the perception of twice the loudness for the listener) you would need 800 Watts. So in other words, the difference between 200 and 250 Watts is almost meaningless. But really, we’re on the wrong track with this, anyway.

The power rating of speakers refers really only to one thing: how much power the speaker can take before overload, meaning either gross distortion or, in some cases, speaker damage. I say “gross distortion” because all speakers have fairly high distortion figures when compared to things like amplifiers and mixing consoles. So how can you look at the specs and see which speaker will be louder with a given amount of amplifier power? The sensitivity rating gives us most of the answer. For instance, Speaker “A” may have a sensitivity of 89dB at 1 Watt, at 1 meter distance. At the same time, speaker “B” may exhibit a sensitivity of 95dB at 1 Watt, at 1 meter distance. Thus, given the same amount of input power, speaker B will produce 6dB more output. That’s like quadrupling the amplifier power.

But let’s say you see a spec like this: speaker X claims “95dB at 1 Watt.” What can you tell from this? Well, unfortunately, not enough. They have neglected to tell you the distance factor for the measurement. To a large degree, this is the most important part. And generally, when something like that is missing from the specification, it is likely a deliberate omission. In other words, they don’t want you to know how this product compares to competitive products and, at the same time, they want it to look as good as possible.

Since sound dissipates according to the inverse square law (SPL is inversely proportional to the square of the distance from the source), the distance factor of the sensitivity measurement is extremely important. For instance, if the measurement was taken at 1 meter (the usual standard) then it can be compared to other speakers measured at 1 meter. However, let’s say it was measured at ½ meter instead. This would mean that at 1 meter, the speaker would only measure 89dB. (This is equivalent to a power reduction of 4x.)

So, one thing to definitely look for in component specifications is a complete set of measurement criteria. You can assume that if a particular criterion is missing, that whoever wrote the specification is hiding something. If in doubt, contact the manufacturer for clarification. If they are not forthcoming, it is usually recommended to avoid such a manufacturer.

Tolerance—and not the Social Kind
Another thing to look for in all specifications is a tolerance range. This is most often expressed as something like “plus or minus XdB.” An example would be a loudspeaker frequency response measurement such as 55 Hz to 18 kHz ±3dB. This means that for this device, there can be as much as a 6dB difference between the peaks in response and the dips. And from the math above, you know that 6dB is significant, and represents four times the power and two to one in perceived loudness. You can definitely hear this, and in fact, depending on where the peaks and dips are, they may be painfully obvious. As an example, say there is a -3dB dip at 3 kHz, but a +3dB peak at 8 kHz. Most likely, the speakers would sound kind of dull and yet scratchy and sibilant at the same time. No thanks.

So while 55 Hz to 18 kHz response initially sounds good, the ±3dB tolerance range gives you a better sense of what’s really going on (and is actually fairly good for a loudspeaker). Then the next step would be to look at the graph to see where the peaks and dips are. But most importantly (as with any sound equipment) listen to it. See if what you hear matches with what you think you will hear based on the graph. This is a great first step in coming to understand how written specifications can be interpreted.

Of Mics and Men
Another type of product where specifications are subject to huge variations based on measurement distance is microphones. First of all, all directional microphones, including cardioid, super-cardioid, hypercardioid, and figure-eight types, exhibit proximity effect. This means that the low-frequency response of these microphones is dependent on their distance from the sound source. Most people who are familiar with this term understand it to mean that when you get really close to the microphone, there is an increase in bass response. Indeed, this is true, and it can be heard on any radio announcer’s voice.

But the opposite is also true: as the microphone gets further away from the sound source, the low-frequency response is diminished. Omnidirectional microphones do not exhibit this characteristic. However, omni mics do have a drop in high-frequency response the further they get from the source due to absorption of short audio wavelengths through air and also from reflections losing HF information. This is why many mic manufacturers offer omni mics with a boosted high-frequency response.

But let’s get back to the issue of bass loss with greater distance when using directional mics. If you examine the on-axis amplitude response (often known as “frequency response”) graph of a given directional microphone, you’ll notice that it appears to have a drooping bass response below, say, 100 Hz. But since the low-frequency response is dependent on the distance between the microphone and the source, you have to first find out at what distance was this microphone measured. The most often used standard is, again, 1 meter. However, it is not universal. In fact, there is a manufacturer making a cardioid microphone that measures their product at 0.3 meters, with the result that the frequency response looks very similar to many other microphones. However, they certainly won’t sound similar. A microphone with a “normal-looking” response at 0.3 meters will sound very thin at 1 meter.

And again, if the measurement distance is not specified, you can assume that they found a distance that made the graph look good, but aren’t reporting in the specifications because it is not a standard.

Peculiar Polar Patterns
Another area that usually requires careful investigation when it comes to microphone specifications is the pickup pattern. Often, the only diagram shown is for response at 1 kHz, while the rest of the spectrum is ignored. And this can be a major issue because the polar pattern uniformity across the frequency band, or lack thereof, plays a huge role in the sound of the microphone and also in how much gain before feedback the microphone provides. For instance, if the response is “textbook,” i.e., very uniform at 1 kHz but exhibits significant lobes at 4 kHz, it is possible that feedback may be quite difficult to avoid at 4 kHz. If you can’t see this on the graph, then you won’t know it’s there until you try to use the microphone.

The best microphones have been designed with considerable effort towards ending up with a uniform polar pattern at all frequencies. In this case, usually the manufacturer will proudly show you a graph indicating this feature. Even still, there are variations in pickup patterns at the different frequencies, so it is a good idea to familiarize yourself with how all the microphones in your collection sound in the real working environment.

To wrap up, let me say this: you will be much better off when you assume that missing criteria such as tolerance ranges and measurement distances are deliberate omissions. The more reputable manufacturers will provide complete specifications so that you can make intelligent choices. And generally, this means the cost for the better components is higher. Caveat Emptor, and when it comes to audio equipment, you get what you pay for, 95% of the time, ±3dB.

Editors note: This is the first of a three-part article. Part one originally appeared in the Summer 2006 issue of Church Sound Magazine. Parts two and three are exclusive to Church Production Magazine Online and will run concurrent with the May and June issues respectively.

Karl Winkler has worked in the professional audio industry for more than 15 years including as a touring mixer, a recording engineer and a technician. He is currently Director of Business Development for Lectrosonics, Inc. in Rio Rancho, NM.