Multi-listener equalization in the home theater

With multichannel audio now widely available in consumer media such as DVD, DVD-Audio, Super Audio CD (SACD), as well as HDTV, home entertainment has rapidly transitioned from a single person sitting in the “sweet spot” to a multi-person experience. With this transition, a new set of requirements for audio playback systems is necessary to provide each listener with the same high quality experience, regardless of his seating position within the listening area.

To make sure that each listener is hearing the best sound, it is necessary to minimize the effects of room acoustics that arise from loudspeaker sound interaction with the acoustical environment. This must be achieved at every listening location, and not just one seat, to allow multichannel systems to achieve their true potential.

The rich experience offered by multichannel sound is often lost in consumer installations due to the lack of proper room equalization and system set up. In today’s systems, set up is time-consuming and equalization often relies on improper implementations that compromise sound signals. Based on a novel digital signal processing approach, a new room correction and calibration technology to resolve these problems is now available to not only improve multichannel sound quality, but also automate system set up.

Room-response correction is critical
Even the best-designed loudspeakers can deviate significantly from the desired frequency response due to speaker placement and where the listener is seated in the room. In typical home theaters, there are problems in the low frequency range because of standing waves that are created by the walls, ceiling and floor. These waves can cause peaks or dips in the frequency response to be as high as ±10dB and vary significantly from seat to seat. There are also problems at higher frequencies, such as comb-filtering, that arise from the combination of direct sound from the loudspeakers with sound reflected from nearby surfaces like large projection screens or coffee tables. Finally, the program material itself is mixed in rooms with completely different acoustical characteristics. The mixers hear the direct sound rather than room reflected sound, and this influences the decisions made during content creation. In a home theater, listeners sit farther away from the loudspeakers and the room is more reverberant; thus, changing the perceived tonal balance of the content from what was originally intended.

All of these factors influence the quality of sound and they do so differently for each listening position in the room.

Until now, proposed methods for room correction suffer from two main drawbacks. The first is the inability to correct the frequency response for more than one listener. Even methods that correct for one listener often make the response at other locations worse than it would have been before the corrections. Corrections for a larger listening area rely on some type of spatially averaged filter that assigns equal weight to each frequency and to each listening position. Room problems, however, are not uniform in frequency nor in spatial distribution, so spatial averaging, while computationally efficient, falls short of providing the best solution.

The second drawback is that existing methods rely on a parametric equalization approach that is typically implemented with a bank of infinite impulse response (IIR) filters. Unfortunately, while these filters may perform correction in the frequency domain, they often introduce audible problems (such as smearing or modal ringing) in the time domain. In addition, parametric equalizers have a limited number of frequency bands (at most 10), and these simply do not offer enough precision to correct the problems found throughout the entire frequency range.

With all these pervasive weaknesses in equalization tools, setting up a system has continued to be a largely subjective process that prevents the sound system from achieving its true potential. Until now, the right tools have not been available to solve the problems correctly. Both professional installers and consumers would benefit from tools that automatically equalize the room, adjust levels and perform other set up tasks to create predictably high-quality sound.

A new room correction and calibration technology
A new technology that addresses these problems has been developed as a result of several years of university research. Multiple-Listener Equalization or MultEQ from Audyssey Laboratories, which is currently being integrated into home theater systems, uses a completely new approach to solve the problem of room correction and system calibration.

With powerful digital signal processors (DSP) available today, such as the Aureus Audio DSP from Texas Instruments, more sophisticated algorithms can be used to find the necessary room correction filters for multiple listeners while reducing the set up time from an hour or more to ten minutes or less.

MultEQ filters are designed to be finite impulse response (FIR) and therefore, unlike IIR filters, correct both the frequency and the time domain. Although FIR filters do not suffer from the artifacts introduced by IIR methods, special considerations must be taken in their implementation in order to apply the correction where it is needed most. In a traditional FIR implementation, the frequency resolution of the filter is constant throughout the audio bandwidth. Room acoustical problems in home theaters, however, are more prominent in the lower frequency range. To address this, MultEQ employs dynamic frequency allocation that increases the frequency resolution in the areas with the most need.

Because of the large variations in frequency response at different seats in the room, MultEQ uses an innovative approach based on pattern recognition. Responses from different seats that are similar are clustered in a group and a response from that group is created that represents the characteristics of each response better than a simple average. These representative responses from each cluster are then combined to form one final response for the entire listening area that is weighted according to the severity of the acoustical problems at each seat. This process is entirely automated and takes less than five minutes to complete using TI’s Aureus DSP platform.

In addition to finding the correction filters for each channel, MultEQ also performs a complete calibration of the audio system. It determines how many loudspeakers are connected, if the polarity of each one is correct, what type each loudspeaker is (“large” or “small” satellite or subwoofer), and what the optimum crossover frequency to the subwoofer should be in the bass management system. It also adjusts the individual delays and level trims for each channel, and can set the playback volume to the desired reference level.

New products with automatic room correction
Various manufacturers already have home theater products in development that take advantage of the features of MultEQ. All of these provide a set up process that the installer or consumer can use to automatically configure and equalize a home theater system in very little time. Some products will offer additional options that will allow the installer to add information such as the room dimensions, surface materials and loudspeaker characteristics to allow the proper target curve to be easily determined automatically. One product scheduled for release later this year puts these capabilities and others in a standalone box, giving installers additional support for extremely problematic spaces. Finally, products will be designed to be upgradeable by CD/DVD or network connection to provide access to the latest control features and new algorithm versions.

The new MultEQ equalization technology takes the time and guesswork out of setting up home theaters and other multichannel audio systems. These “set and forget” installations are easy to use, requiring only a few minutes to calibrate. Listeners can then enjoy the best possible sound in their room from their system that has been automatically customized to overcome the acoustical problems of the space. These advances in room correction help provide a listening experience that fully lives up to the name of home theater.

Authors information
Chris Kyriakakis is the CTO and co-founder of Audyssey Laboratories. He is also a professor at the USC Viterbi School of Engineering where he is the director of the Immersive Audio Laboratory that he founded in 1996. Dr. Kyriakakis is a member of the Audio Engineering Society (AES) and the Institute of Electrical and Electronics Engineers (IEEE). Mohsin Imtiaz is the marketing manager of Texas Instruments’ Performance Audio group. He works with worldwide sales to win new business, driving product release to market and ensures TI’s roadmap reflects market needs and positions the company to succeed moving forward. Imtiaz has served as the product marketing manager for cable broadband communications and DSP Product TSR.