New Horizons: Live Concerts In Surround?
Despite the obstacles, it can indeed be done

By Michael Miles

Shall we surround them in the future?

About 15 or so years ago, I went to a rock concert at a stadium in Chicago. Having heard this band’s best-recorded music on some excellent small systems, including high-end four-channel setups, I looked forward to this concert with great anticipation. They were known for having elaborate and advanced concert systems. I expected many stacks of loudspeakers and amazing “big, immersive” sound along with a huge video display.

Unfortunately, the outcome proved just the opposite - talk about disappointing! It was barely loud enough, muddy, and distant-sounding. I couldn’t even tell whether the giant loudspeaker stacks on each side of the stage were supposed to create stereo sound. They were just too far away, and their coverage of the audience was inadequate.

The show may have been better for those who were close to the stage, but the rest of us could get superior sound with our home stereos.

Sound reinforcement (in general terms) has improved a lot since then, and good sound for every audience member is now possible in most venues. Although concerts are commonly recorded and produced in 5.1 format for DVD, surround sound for live concert reinforcement is still almost nonexistent. Because progress is inevitable, it seems likely some day concerts will be routinely presented in high-quality surround sound.

But even if the budget is large enough, the laws of physics still create problems for design of a large surround sound system. Fortunately, we can make useful compromises, and as loudspeaker designs improve, we can more readily achieve effective stereo sound for an audience. This is a good basis for surround sound, which can utilize many of the same principles.

START WITH STEREO

Achieving good stereo sound for a large audience requires careful design with close attention to acoustic issues. The basic requirement is for each listener to hear each of the loudspeakers at nearly the same level, nearly the same time, and with matched frequency response. This is a tall order for any size venue, and requires certain compromises.

The biggest problem is the timing of the sound heard from different loudspeakers. The arrival time of the sound from each loudspeaker will be different for every member of the audience.

Due to the precedence (or Haas) effect, the sound source heard first usually determines the perceived direction the sound is coming from. When panning between adjacent loudspeakers, the sound from each loudspeaker (or cluster) needs to arrive nearly simultaneously.

Figure 1: Covering the farther part of the audience with a higher level (loudspeakers shown in yellow) helps to restore the balance for off-center listeners.

To produce reasonably accurate panning, the sound arrivals need to be within about a millisecond (ms) at each listener. For an audience of more than a few people, this won’t happen because 9 ms delay is created by every 10 feet of distance from the sound source. The audible effect of this time error between left and right loudspeakers at a listener depends on both the time difference and the relative level difference. When the levels are about equal, just 5 or 10 ms can cause significant image shifting toward the earlier sound source. Longer time differences, up to about 40 ms or so, will yield significant precedence effect - the earlier sound source will set the direction. Also there will be coloration and loss of clarity. Time differences over 50 ms can be heard as echoes, which of course is a major problem.

DEALING WITH DELAY

Keeping the loudspeakers high enough above the audience will reduce the relative delay, due to the geometry of the layout. This is certainly helpful, but except for the smallest venues, it alone does not solve the problem.

One way to cope with delay is to control the relative amplitude of the loudspeakers in specific audience areas. If the delayed sound has a higher level, this can partially compensate for the delay and help to restore the spatial balance.

In the simplest implementation, left and right loudspeakers are toed in and across. This insures that the opposite side of the audience area is on axis and thus covered with higher output from the loudspeaker than those on the same side and nearer, who are then off axis. It may offer an improvement, but typically suffers from degraded off-axis frequency response, as well as more sound energy being dispersed away from the audience where it’s not needed.

A better approach is to use loudspeakers with a directivity function tailored to the specific audience area to be covered. The loudspeaker further from the listener should operate at a higher level to compensate for its longer delay. This requires a specific loudspeaker directivity function in which the output level is dependent on the angle off axis, while maintaining good frequency response.

A cluster of two or more loudspeakers can be designed to approximate this function. Figure 1 illustrates the principle. Note that the higher-level cluster components (shown in yellow) have higher directivity and are aimed toward the farther part of the audience. The right-side loudspeakers mirror the left. The center loudspeaker helps greatly to expand the effective stereo listening area, and it needs to provide even coverage for the entire audience.

Even with the most optimum directivity control, there is a limitation in how far apart the adjacent-channel loudspeakers can be. If they are more than about 40 feet (12 m) apart, then the relative delay causes serious problems for off-center listeners, including loss of intended localization and a noticeable echo.

The use of a left-center-right system increases the possible width of the front sound stage, as long as channel assignment or panning is kept pair-wise to adjacent channels (left to center, or center to right), so that the same signal is not heard from both left and right.

LONG DISTANCE SURROUND

A well-balanced left-center-right sound system is a beautiful thing, and there are many examples of them producing great stereo sound for an entire audience. But we want more than that; we want surround sound. How can this be achieved?

Sound propagation in air takes about 90 ms for every 100 feet (30.5 m) of distance, and this creates a very bad situation for large venues that may be hundreds of feet across the audience area. Additional loudspeakers at the back or sides cause problems because even a medium-sized facility has enough distance from front to back to produce echoes in both directions, as shown in Figure 2.

Figure 2: Listener at “Position A” hears the rear loudspeaker 50 ms too late, while the listener at “Position B” hears the rear loudspeaker 50 ms too soon.

Listeners in the front will hear the back loudspeakers too late, and listeners in the back will hear the back loudspeakers too soon, compared to the sound propagating across the audience from the front stage. In theater and cinema surround sound systems, these delay problems are largely mitigated through the limited size of the venue, and judicious application of the program material to the surround system. Rarely is a particular component of the mix routed to both front and rear simultaneously; this would cause a problem with clarity and intelligibility. But high-quality surround music needs more precise timing to provide sound from all directions nearly simultaneously for all (or at least most) listeners. One might conclude that there is no way to implement this in a large venue because of the great distances involved.

However, compromises can be made-sometimes with great results. The careful use of electronic delay and a number of well-placed, well-behaved loudspeakers can work wonders.

Whether the goal is to provide additional envelopment, bring the audience “into” the performance, or simply to present sounds from directions other than the front, we need the ability to provide sound that the entire audience can hear specifically from the sides or rear ­ without delay problems. Next time, I’ll present some possible solutions to these issues.

Michael Miles has two decades of product and system design experience, and is the founder of Miles Technology, a provider of professional multichannel audio products. He holds an Electrical Engineering (EE) degree from the University of Michigan, and can be reached at mike@milestech.com

December 2003 Live Sound International