The sound level meter

Two, five or more loudspeakers are used in a music or audiovisual installation to create a two- or three-dimensional stereo sound atmosphere. Two loudspeakers can deliver a stereo image, while five can create an almost real surround ambience. It will always be "almost" real, but for us, that is more than enough.

One of the critical aspects (but not the only one) when setting up the installation is the sound pressure level. Beforehand we will have placed the loudspeakers in position, set their delay, etc., aspects that we have already examined in this magazine or will examine shortly.

Setting the sound pressure level is an essential step in ensuring that the sound image, created by dispersing sound through two or more loudspeakers, is coherent. Although our ears are not entirely perfect (and partly thanks to this we can achieve very cohesive sound fields with just 5 loudspeakers), they are very sensitive to differences in pressure, sometimes with variations of just 1 or 2 dB. But setting levels "by ear" is not entirely accurate, so it is advisable to use a completely objective tool, which in this case is the sound level meter. In the following article we show you how to use it, after first explaining why it is worth using. The result is reasonably objective, and you get optimal performance from your investment.

We know that audible sound is produced as a result of compressions and rarefactions of the air. The frequency of a sound wave is defined as the number of cycles per unit of time and is measured in Hz. For human beings, the audible spectrum extends from 20 to 20,000 Hz. But it is sound pressure that defines the loudness of each sound. Variations of just 1 pascal represent, for the human ear, the change from a faint sound to a very loud one. Sound pressure is measured in decibels. And what is a decibel?

The dB is a logarithmic unit of measurement used in various scientific applications. It is a comparative value, so a reference value must always be specified. Talking about 20 dB, for example, without saying anything else is useless, since the point of origin is not expressed. It is like saying that the meeting place is 20 km away but... from where? From Barcelona or from Madrid?

In acoustics, the dB is used to compare sound pressure in air with a reference pressure. This is the minimum pressure level at which our ears are able to hear "something". In the case of sound pressure (in air), this value is 0.00002 pascals (2E-5 Pa) rms. This value is not the same if we are talking about sound intensity (1E-12 W/m2) or sound power (12-12 W).

Decibel means "one tenth of a Bel", and the Bel is the base-10 logarithm of the ratio between two powers or intensities. Since the Bel is a very large unit, the use of its tenth part has become standard. The sound pressure level is therefore 20 times the logarithm of the ratio between the measured pressure and the reference pressure. 0 dB corresponds to the faintest sound our ears can hear.

Using the dB makes sound pressure easier to understand and calculate mathematically, since otherwise (that is, on a linear scale) we would have to work with either very small or excessively large numbers. We have mentioned before that the behavior of the human ear is closer to a logarithmic function than to a linear one.

BEFORE MEASURING



The sound level meter is the tool used to determine the sound pressure level. We will often come across the abbreviation SPL, which stands for Sound Presure Level. There are two types of sound level meters: those that use a needle to indicate the value and digital ones, which display the result numerically on an LCD screen. But in both cases the readings are always in dB and, as sound level meters, they take the 0 dB mentioned above as their reference. Many sound level meters also provide values such as rms pressure, maximum peaks and minimum levels.

Using the dB makes it possible to "convert" a logarithmic scale into a more manageable linear value. But when applied to the human ear (which is logarithmic), this means the resulting value bears little relation to reality (this is because our ears do not respond equally to the same sound pressure level at different frequencies). Our ears (as we have seen in other articles) are more sensitive to midrange frequencies than to bass and upper treble. If we play a disc with a tone sweeping from 20 to 20,000 Hz at the same sound pressure (the sound level meter would always show the same value), we would not hear it linearly: we might not even hear the bass, the midrange would become increasingly clear, and as we approach 20,000 Hz we would progressively lose our hearing again. But the sound level meter would remain fixed at the same value.

That is why a way has been found to match sound pressure levels to the (average) perception of the human ear at each frequency. There are weighting curves called A (low level), B (intermediate level) and C (high level), and the result is expressed in dBA, dBB or dBC respectively (see graph below). The frequency of 1 kHz has been taken as the reference value (that is, the measurement in dB is the same as in dBA, B or C). Thus, a 125 Hz tone that reads 75 dB on the sound level meter would be 61.6 dBA with A weighting or 74.9 dBC with C weighting.

Sound pressure level is really easy to measure (in fact, all we need is the sound level meter), although the value we obtain depends on other parameters such as the distance between the source and the measuring point, the acoustics of the room, the audio being played, etc. All these conditioning factors make it logical for the measurement to be carried out once all the other parameters, such as loudspeaker placement, room decoration, etc., have been set. In fact, it would be the last step as far as setup is concerned. Manufacturers specify the sensitivity value in dB, using 1 W of power as the source and measuring at a distance of 1 m. The value resulting from this measurement is valid for the loudspeaker tested, but it gives us information that is of little use in our case.

The aim is to adjust each and every one of the loudspeakers in our installation so that, from the ideal listening position, they are all perceived in the same way. If one loudspeaker is louder than another, the sound will "lean" toward the louder loudspeaker. The vast majority of A/V receivers have a section where we can change the output level of each channel by +/- 10 dB.



MEASURING



Pink noise is usually used for the measurement. This is noise whose level is characterized by a drop of 3 dB per octave. If we analyzed this noise with a spectrum analyzer, we should get a flat graph, indicating that all the octave bands have the same level; but for this to be so, each time we go up an octave we have to reduce the output sound pressure by 3 dB (halving the power).

Seated at the ideal listening position (on the sofa), we will first have set up our sound level meter. Ideally, we would use B weighting (for medium pressure levels), although it is rarely found on sound level meters, so we will choose C weighting. This applies less attenuation to low frequencies than A weighting. In fact, if only A weighting is available, it is recommended not to use any weighting at all.

A/V receivers usually have a pink noise generator, which moves from channel to channel while allowing the output level to be changed. If not, there are test discs (such as our usual "Digital Video Essentials") that include it.

As we said, we will sit in our favorite spot and, with the sound level meter as close as possible to us and to our head, we will note down the level the meter shows for each of the loudspeakers. I recommend setting the receiver or amplifier volume to the usual listening level or, to be somewhat more precise, setting it so that the sound pressure level from one of the loudspeakers (which we will use as the reference) is 75 dB. In fact, all the loudspeakers should give the same value. To achieve this, we will change all the output values for each channel according to the difference between the reading and the ideal 75 dB. This will mean that on some channels we will have to raise the level by 5 dB, on others by 2 dB, or attenuate by -3 dB.

The subwoofer is also a loudspeaker, and therefore needs to be calibrated. The problem is that this particular channel has a tolerance margin that depends on each person's taste or type of listening. Thanks to its omnidirectional radiation, the problem of sound image balance does not arise even if the subwoofer is at +10 or -10 dB relative to the other channels. Some people set it to 95 dB, while others, after measuring, consider that the resulting deep bass is too prominent and noticeably reduce its level. For music we may even switch it off, while for movies its spectacular impact may pleasantly surprise us.

NOT ALL SOUND LEVEL METERS ARE THE SAME

Just as one loudspeaker costs 100 euros and another 10,000, there are very affordable sound level meters and others that are much more expensive. Is there any difference? Obviously there is, and it is called tolerance.

The sound level meter is a "scientific" measuring tool, and as such it has a tolerance rating. For example, the most expensive (and professional) ones guarantee that measurements are made with a margin of error of only 0.05 dB, whereas the cheapest can show discrepancies of up to 3 dB or more (which is a lot). The sound level meter is based on a microphone that must be able to "hear" all sounds from 20 to 20,000 Hz without attenuation or emphasis, together with a complex processing system (usually digital) that must react in real time to the differences in sound pressure picked up by the microphone.