Computer Modeling of Immersive and Multi-Channel Arrays

David Kennedy • December 2019Tech Feature • December 14, 2019

Welcome back to the (ever) expanding universe of Immersive Audio! To recap, last month (FRONT of HOUSE, Nov. 2019, page 54) we discussed how and why live “surround sound” has become “immersive audio.” And it’s coming on and coming fast — as an example, The Recording Academy changed its “Surround Sound” Grammy Award category to an “Immersive Audio” category. Also last month, we looked at how immersive systems require considerably more speaker locations than older systems, yet at the same time, immersive arrays don’t have to be as large as arrays in a system with a smaller number of arrays.

This month, we’ll examine how immersive, LCR (Left, Center, Right) and stereo live-sound systems differ in cost, coverage and max. sound levels. Of course, audio imaging can be improved with a greater number of immersive arrays across a stage. But I also wanted to investigate how half-sized line-arrays or point-source arrays can provide the same maximum SPL and distant coverage of large, stereo line-arrays. I also returned to d&b audiotechnik’s Marc Lopez to provide more insights on comparing some additional legacy vs. immersive system design concepts. Rather than showing small/square venues (i.e., a typical cinema shape) — that are fairly easy to cover with immersive sound — we are delving into some sound coverage models of an actual venue that represent a large, fan-shaped church. Distance from P.A. to back row is 100 feet, trim heights are relatively restricted due to ceiling height, the seating capacity is 3,000 people. Not shown are the subwoofers, fill or surround speakers.

‡‡         Comparisons: d&b’s Stereo vs. LCR vs. Soundscape

Figure 1: “Stereo” d&b line-arrays, with a total of 24 modules — up to 111 dBA @ 4k Hz.

Fig. 1 shows Option A, with “stereo” line-arrays, and a total of 24 modules that can provide 111 dBA with a front-to-back level variation of 8.4 dBA. It has the typical pros and cons of a stereo design, including high-impact for center audience areas and standard L/R mix. But it has poor side coverage and localization for all audience areas — except at the center — with each array only covering about half of the venue — so, it’s little more than dual-mono (no side-fill speakers nor comb-filtering are shown).

Figure 2: Pseudo-L/C/R d&b line-arrays, 24 modules — up to 108.6 dBA @ 4k Hz

Fig. 2 shows pseudo-LCR line arrays — actually more of a three exploded array design — with a total of 24 modules — that can provide 108.6 dBA broadband — with a front-to-back level deviation of 7.4 dBA. It has the typical pros and cons of an exploded array design, including better localization for some audience areas and reduced impact on sightlines. But LCR mixing is a hardship for engineers; the Center array does not fully cover the audience area, and it requires “walking the room” to adjust blend between the LCR line-array elements.

The cost of the pseudo-LCR line array compared to “Option A” (the stereo L/R line array in Fig. 1) is roughly the same (multi-channel interaction not shown).

Figure 3: Immersive design using six pair d&b point-source arrays — up to 106.5 dBA @ 4k Hz

Fig. 3 shows a six-pair point-source array Soundscape immersive design — with a total of 12 loudspeakers — that can provide 106.5 dBA broadband — with a front-to-back level deviation of about 5 dBA. It has many more pros than cons. Benefits include more consistent coverage, better localization and time-alignment for all audience areas, and reduced sight-line impact, but it also requires a non-traditional immersive mix from the engineers. Cost compared to Option A: Roughly 50 percent cheaper — but with much less SPL than the other design concepts.

Figure 4: Six d&b line-array immersive design — up to 111.6 dBA @ 4k Hz.

Fig. 4 shows a six line-array Soundscape immersive design — with 48 of d&b audiotechnik’s smallest line array, T-Series modules (smaller than other designs). That can provide 111.6 dBA broadband — with a front-to-back level deviation of 6 dBA. It has many more pros than cons. Benefits include: superior sound coverage, better source localization and time-alignment for all audience areas, improved gain-before-feedback and reduced sightline impact; however, this requires a new immersive mix from the engineers. The cost is roughly the same as Option A — but with higher SPL. Credit to d&b’s Nick Malgieri for the modeling study.

As Marc Lopez explains, “for all line array options, ArrayProcessing can also be considered; it is a unique DSP algorithm for optimizing the configuration of d&b line arrays to achieve consistent tonality and level balance from front-to-back of a venue.” The dB scale for all of these sound coverage models is 3 dB per color. The heat (sound coverage) map shows 4k Hz for coverage purposes, but the max SPL figure is an IEC-60268 broadband signal.

‡‡         Multi-Channel Modeling

Earlier this year, I wrote a series for FRONT of HOUSE on the coverage modeling of loudspeaker arrays. In the first article (March, 2019, page 30), I showed examples where sound levels two times the standard scale/range, over an average of several octaves, or displaying multiple arrays at the same time, is not realistic or true. Rocky Giannetta, CTS-D, added “the most unfortunate trend is the misuse of software programs as marketing tools. The bottom line is that both modeling and measurement software programs are useful tools for professionals who have a good understanding of electro-acoustics.”

I recently discussed modeling for multi-channel speaker array coverage with Michael Garrison of MGA in Fresno, CA (who was the team leader for the design and development of JBL’s CADP2 loudspeaker modeling program decades ago). Garrison notes that “the benefits of well-designed and properly executed immersive loudspeaker systems is dramatic and easily appreciated by average listeners. But those of us designing these systems will need a clear understanding of the requirements to achieve optimal results. We will also need expanded computer modeling routines to aid in determining the proper selection, and placement of the loudspeakers, with a graphical display to model the degree of immersion predicted across the audience.”

Multiple/immersive arrays can improve coverage at the front corners of a venue. So, while extreme edges of a venue do not get the full multi-channel immersive effect — sound coverage is not as bad as the two to three array legacy systems had typically been. Achieving reasonably accurate directional realism and intelligibility of the talent is another related issue for me; this is difficult to accomplish in large venues with just stereo loudspeakers. Sound imaging with poor directional realism (imaging) causes cognitive dissonance — mental stress due to contradictory data. While directional realism has not been a priority for the concert industry, it is important to live-theater sound designers (and some churches) and is a primary benefit of immersive live-sound systems.

I just surveyed several fellow sound consultants and several of the leading loudspeaker array modeling software developers, asking, “How is venue coverage modeling done for multi-array and immersive systems?” Most of them said to “evaluate the coverage of each array — one at a time.” But d&b replied “because Soundscape uses level and delay across multiple loudspeakers to deliver a coherent spatial image, we can design a system using a few simple guidelines: Through the use of wide horizontal coverage arrays, preserving image integrity can be achieved by calculating speaker density with a ‘70 percent’ rule, where lateral speaker distances from one another should not be wider than 70 percent of distance to the nearest listeners. For spaces that require fills or delays, Soundscape has ‘Function Groups’ which intelligently calculate proper delay and imaging for those additional zones to be seamless and phase-aligned with the main zone.” L-Acoustics’ reply follows below.

‡‡         Comparisons: LAcoustics’ Stereo vs. L-ISA Immersive

Figure 5: L-Acoustics’ Soundvision model of classic stereo design; SPL-A map

Now, let’s look at how immersive and stereo live-sound systems differ in coverage, multi-array overlap and max SPL, according to Soundvision 3D modeling software. Scott Sugden of L-Acoustics explained, “Soundvision 3D modeling software is able to quickly and easily model L-ISA Immersive Hyperreal sound design solutions in great detail. Soundvision can create a great number of unique loudspeaker feeds to inform the designer the quantity and quality of the immersive field” [multi-channel coverage].

Figure 6: L-Acoustics’ Soundvision model of classic stereo sound design; SPL-A map.

Fig. 5 shows a classic L-Acoustics “stereo” line-array system modeled in Soundvision software. This venue is not nearly as wide as the example in Fig. 1, so the coverage of a dual-mono system is good. But this example still exhibits the typical pros and cons of a stereo design, including standard L/R mix and high-impact for the center areas only. Note in Fig. 6 that more than half of the venue seating area is shown in yellow, because much of this large seating area has coverage from only one channel-array. Light green area is the stereo sweet spot, (dual-mono signals not as problematic in this zone). In the dark green area, stereo signals still look good, but dual mono signals are greatly degraded. Fig. 7 & 8 show a L-ISA Hyperreal (immersive) 7.1 sound system design in Soundvision.

Figure 7: L-Acoustics Hyperreal sound design for 7.1 system from Bon Iver show, Aug. 2018

Sugden notes that, “starting with the Immersive coverage map, Soundvision will highlight the audience zone with complete coverage from the multichannel L-ISA sound system. The highlighted green area is the shared coverage region from the seven arrays of the Hyperreal sound system design; meaning an object mixed anywhere in the system will be heard at a similar level to any other object at any other position in the system. This is calculated based on the loudspeaker position, acoustic parameters of the given loudspeaker and electronic settings in the particular design. The ‘Unshared’ region in yellow is the audience coverage area that is missing coverage from at least one array of the L-ISA Hyperreal sound design and will require some fill speakers to properly produce the mix for the audience. In this particular design, the dark green in the coverage map represents an area where the time alignment between different speakers will be a problem for listeners.” The coverage maps are from the Santa Barbara Bowl concert sound design for Bon Iver on Aug. 8, 2018.

Figure 8: L-Acoustics’ L-ISA Hyperreal sound design for 7.1 system from Bon Iver show, Aug. 2018

Soundvision generates a radial graph to show the designer seven criteria to understand the immersive design:

1) SPL max. average across all the seating-coverage area,

2) the distribution across the seating,

3) Time, indicating how much of the audience is in an area without disruptions in the temporal information of the multiple signals,

4) A count of the number of sources the mix can be distributed across,

5 & 6) Show the horizontal and vertical fusion between the real object onstage and the amplified signal from the system,

7) Panorama is a count of how much of the audience will experience a wide sound field from the sound system.

At this time, LAcoustics’ Soundvision software is the only modeler that can simulate the multi-array coverage of an immersive design (shown in Figs. 6 & 8). So, other modeling SW developers should consider adding a method to display directional realism with multi-channel coverage.

‡‡         Conclusion

Although the term “immersive sound” allows for surround and overhead loudspeakers, surround speakers are not always needed or included. Immersive live sound is at least, if not more, about superior sound coverage, better source localization and time-alignment across a venue as it is about surround sound. The surround (and overhead) speakers can be added later, as some systems allow for gradual system expansion over time (to reduce initial sticker shock). The need for surround loudspeakers is dictated by budget, content/program, venue space and setup time limitations — for example, if acoustic simulation is desired — or if any sound localization is required around the audience area. Some immersive DSPs, are speaker-type-agnostic, thus are less confined to rigid speaker system geometries than spatial/ immersive algorithm devices that may be based on Ambisonic, Flux-Spat, SpatialSound Wave or Wave Field-based principles. Examples include Astro Spatial’s SARA II and Out Board’s TiMax with Showcontrol.

So, immersive sound is very flexible — providing better coverage and localization with less than 180° of immersion (across a proscenium/stage) — or up to a full 360° of sound immersion — and can be sourced from a turnkey immersive manufacturer — or power users can “roll their own” DIY systems. Stay tuned!

David K. Kennedy operates David Kennedy Associates, consulting on the design of architectural acoustics and live-sound systems, along with contract applications engineering and market research for speaker manufacturers. He has designed hundreds of auditorium sound systems for churches, schools, performing arts centers and AV contractors. Visit him at www.immersive-pa.com.

Leave a Comment:

The Latest News and Gear in Your Inbox - Sign Up Today!