Live Sound

The difference, in decibels, between the loudest and the quietest portion of a program is its dynamic range. Dynamic range applies to sound and to sound systems. Every sound system has an inherent noise floor, which is theresidual electronic noise in the system. The dynamic range of a sound system is equal to the difference between the peak output level of the system and the electroacoustic noise floor.

There really is no such thing as "typical." The widest dynamic range you're ever likely to encounter won't be at a concert because there's too much noise around you – it'll be in a quiet recording studio. In an exceptional concert setting, the sound levels at the microphones may range from 40 dB SPL to 130 dB SPL (above the threshold of pain). What is the dynamic range of this "extreme" concert? Subtracting the noise floor from the peak levels:

Dynamic Range...
(Peak Level) – (Noise Floor)
130dB SPL – 40dB SPL
90 dB

This concert has a 90 dB dynamic range at the microphone. In reality, most concerts have 70 dB or less dynamic range.

Ideally, it should be at least as large as what's on stage. The electrical signal level is proportional to the original sound pressure level at the microphone. When the sound level reaches 130dB SPL at the mic, the maximum line level (at the mixing console's output) may reach +24 dBu (12.3 volts). The maximum output level from each power amplifier may peak at 250 watts. When the sound level on stage falls to 40 dB SPL, the minimum line level falls to -66 dBu (388 microvolts), and the power amplifier output level falls to 250 nanowatts (250 billionths ofa watt).

Do the acoustical environment on stage and the electrical signals in the console have the same dynamic range?

Dynamic Range...
(Peak Level) – (Noise Floor)
+24 dBu- (-66 dBu)
90 dB

Yes, this program retains the same dynamic range at the mixing console output as at the mic, but how about at the power amplifier output? The relationship for 250 nanowatts to 250 watts is also 90 dB. What kind of acoustic levels must the speakers generate so the audience can hear the full dynamic range?

When speakers aren't capable of handling the full dynamic range, they're going to distort or burn out on the peaks and/or be incapable of responding to the lowest power levels. You don't want the audience to feel their ears are an inch from the lead vocalist's mouth. Ouch! 130 dB SPL is a very high peak level. 120dB SPL is the threshold of pain,' let's make that the max.

Since the speakers are some distance from the audience, they must generate more than the desired 120dB – say 10dB more, or 130dB SPL. Given 90 dB dynamic range, they'll produce 40 dB SPL during the quietest passages. Since the sound reaching the audience is attenuated 10 dB by air and distance, the 40 dB SPL generated by the speakers during quiet passages drops to 30 dB. That is probably below the ambient noise level in the audience so they may not hear the very quietest parts of the show. In this case, compression of the loudest peaks by 10 dB would allow the average level to be turned up that much louder – rising above the audience noise. We explain this a little further on.

The average electronic line level in the sound system just described is +4 dBu (1.23 volts), corresponding to an average sound level of 110 dB SPLat the microphone. This average level is also called the nominal program level. Remember that the maximum line level was +24 dBu (12.3 volts). This is possible with most Yamaha consoles, but on many other consoles the maximum's only +18 dBu (6.2 volts), which is half the level. The difference between the nominal and the highest (peak) levels in a program is the headroom. Given the levels at the microphone, let's calculate the headroom required for the concert sound system previously described.

Headroom…
(Peak Level) – (Nominal Level)
130dB SPL – 110 dB SPL
20 dB
Similarly, the electrical headroom is 20 dB, as calculated here:
Headroom…
(Peak Level) – (Nominal Level)
+24 dBu – (+4 dBu)
20 dB

Provided the power amplifier is operated just below its clipping level of 250 watts peak, and at nominal levels of 2.5 watts, then it also operates with 20 dE of headroom.

Figure 1 illustrates headroom and dynamic range in a particular sound system, both in acoustical and electrical terms. The S/N Ratio shown in this illustration refers to Signal-To-Noise Ratio. This is the difference behween the nominal level and the noise floor.

Dynamic range is the difference between the loudest and quietest portions of the program. It's the whole gamut ofavailable levels. The S/N Ratio, on the other band, begins at the noise floor and goes to some arbitrarv nominal level. When the S/N ratio (in dB) is added to the headroom (in dB) the result is the dynamic range (in dB) ignoring for now the possibility of a recognizable signal which is below the noise floor.

Headroom describes the ability of the sound system to handle loud program peaks. Given two sound systems that both operate at the same nominal level, the system with the greater headroom will be able to handle louder peaks before distorting or breaking. Headroom requirements change with the nature of the program material and the purpose for which the sound system is operated. A sound system intended for paging in a loud factory environment may need to have a very high nominal sound level (say 110dB) to overcome machinery noise, but it needs only 6dB or so of headroom. A rock music system may need 10 dB of headroom. And a symphony orchestra concert may need more than 20 dB of headroom. This is because the average level of the orchestra may be low – say 90 dB – but on loud peaks a given instrument may momentarily exceed 120 dB SPL. That needs 30 dB of headroom. If the sound system is capable of only 20 dB of headroom, the peaks will be distorted.

Are more power amplifiers and speakers needed for a symphony concert than for a rock concert? Not really. The same amount of equipment, or possibly less, will suffice. You can get the extra 10 dB of headroom by setting the average level 10 dB lower.

Except for small systems, where a 3 dB increase may mean one more amplifier and loudspeaker, it's generally too costly to build in more dynamic range than is absolutely necessary. You can easily spend thousands of dollars per dB SPL of extra capability in a large sound system. It pays to instead find ways to reduce the dynamic range requirements.

What happens when the sound system is inadequate?

When the dynamic range of the program material exceeds the dynamic range capability of the sound system, some combination of the following will result:
a) Program peaks will be distorted due to clipping and/or loudspeaker break-up, or b) Quiet passages will not be heard because they will be below the electrical and/or acoustic noise floor.

Yamaha loudspeakers are conservatively rated, and have meaningful power specifications so you can be sure you're buying enough speaker, but not more than you need. Yamaha power amps are accurately rated at the power they will deliver in "real world" use, so again, you can make sure you buy the right size amp. For example, an 8-ohm Yamaha S11SMTII can handle 200 watts program or 400 watts peak. One channel of a Yamaha P2700 amp will deliver 350 watts into 8 ohms. Therefore, the amp can easily drive the speaker at the rated 200 watt program level while maintaining a 3 dB margin before clipping, yet if the amp does peak at full power, it won't blow the speaker.

Without special processing, for every 2 dB change of input level, the output level also changes by 2 dB. Suppose that for every 2 dB change of input level, the output changed only 1 dB. The dynamic range of the program would be cut in half: 90 dB dynamic range would become 45 dB (Figure 2).



This is exactly what is possible with a signal processor known as a compressor. By setting the compressor for a relatively gentle compression ratio of 2:1, every dB of input level change will result in half a dB of output level change. This is tolerable in all but the most critical musical reproduction.

Sometimes even higher compression ratios are used, but there can be side effects, such as making quiet breath sounds louder, creating a pumping effect in some cases, and increasing the distortion of low frequency signals. By applying compression only above a given threshold, most of the program sounds completely natural. Above the threshold we use compression as necessary to prevent clipping (Figure 3).



Some devices permit compression to be applied above a set threshold, and the compression ratio to be very high These devices are known as limiters because, above the threshold, the output changes very little regardless of changes in input level. Devices that can be set for compression or limiting are known as compressor/limiters. It's a good idea to have at least one available per output channel, and sometimes additional units for "problem" inputs.

The choice of a headroom figure depends on the type of program material, the application, and the available budget for amplifiers. For a musical application where high fidelity is the ultimate consideration, 15dB to 20dB of headroom is desirable. For most sound reinforcement applications. especially with large numbers of amplifiers. where economics play an important role, a 10 dB headroom figure is usually adequate. In these applications, a compressor or limiter can help hold program peaks within the chosen headroom value, and thus avoid clipping problems.

To set up your system for optimum signal-to-noise radio and headroom, see the Sound Information Series sheet titled "Optimizing Gain Structure."