To reduce boredom, the future will be full of surprises. Sometimes, even the surprises can be accommodated without buying new hardware. The secret is headroom. Excess power amplifier may be overkill for one show, yet barely enough for the next. When it comes to the system processing chain (drive rack), headroom takes a different form -- DSP power.

When you build a sound system you want to evaluate every possible requirement and provide the necessary processing to meet these needs. However, over time the expectations of the system users and the needs of the facility may demand new or different processing. It may be the addition of signal delays, the use of multi-channel effects or simply the addition of limiters where, too often, drivers are being lost.

When sound systems relied on modular signal processing, these changes required the addition or replacement of a few components in the drive rack. Now that all of this processing may take place in a single DSP-based system processor, the change may be done in software or require a whole new unit (equivalent to replacing everything in the drive rack, except the amps). To prevent the latter case from blowing the upgrade budget, it is worth considering the flexibility of the processing architecture before buying the system processor.

There is a wide range of DSP-based signal processors available these days. Some offer a predefined signal path that usually emulates the rack of analogue equipment they replace. Then there are the units that must be configured to suit the application. In between, there are processors that appear to be pre-defined but have an "Advanced Mode" that allows the more sophisticated user to alter the signal processing chain more radically.

The predefined-signal-path units were the first to appear. The first units to be widely used were the TOA DACSys processors, but now there are dozens of processors that take this form. They offer a fixed selection of configurations and are quick to setup. They often have a drop-down menu in the setup software or step through selections on the front panel display. The user makes a mode selection, such as two-channel two-way crossover, and all of the signal processing is ready to adjust. This is certainly the fastest way to get a unit passing audio. It can also be the most efficient use of DSP power, as the designers can carefully optimize each configuration. Unfortunately, any configuration that steps outside of the manufacturer's vision of sound reinforcement applications for that unit will not be allowed. This is a gotcha that might limit the extent the unit's processing can be revised to suit new and more complex tasks later in the life of the system.

The other extreme is the processors that allow each processing block within the signal chain to be selected by the user. A broad range of basic signal processing components are offered; dropped into a graphic of the signal flow; and connected together by the user. This requires the processing signal path to be designed rather than selected, but also offers the widest possible range of applications to be served. Originally innovated by the Peavey MediaMatrix System, now this format is also offered by devices such as the BSS Soundweb and IED 3200 processors. The limitation of the processing is now constrained only by the library of basic devices and the total processing available in a particular unit. Even these limitations may diminish as the manufacturer extends the component library or (on some units) DSP resources are purchased. These processors offer modular input and output units, in various forms, that can also be expanded at nominal cost. Networking these units has become the latest trend, where processing and signal routing can be distributed between units and over large distances via the ubiquitous Ethernet or Peak Audio's proprietary Cobranet.

The fixed-format units that offer an advanced-mode fall in between these two extremes. The library of basic processing modules is much smaller, but the layout of the signal path can be redefined. This retains the efficiency of the optimized DSP code but isn't likely to allow a two-in and four-out unit to be expanded to a four-in and six-out unit. I have discovered that even units that appear to be fixed selection of processing might have a software application or secret code that accesses the advanced capabilities. It is always worth asking the manufacturer if they offer this capability.

The cost of signal processors for a sound reinforcement system may be deceptive when you are only looking at the immediate needs of your sound system. If you can envision the system changing over the next few years, the investment in more flexible devices could prove worthwhile. Perhaps the real question is: How many options do you want to leave open?

United Entertainment Media Inc.
460 Park Avenue South, 9th Floor
New York, NY, 10016
Ph. 212-378-0400
FAX 212-378-2160

page construction