YDCG Design Considerations for Broadcast and Production

DESIGN CONSIDERATIONS FOR BROADCAST AND
PRODUCTION FACILITIES IN THE DIGITAL AUDIO ERA

The evolution of sound recording technology during the past decade has changed the design criteria for the current generation of broadcast and production facilities. Sales growth of stereo television receivers, movie video rentals, home theater and high end automobile sound systems as well as the rapid acceptance of compact discs has raised the standard by which broadcast audio is compared.

The old three inch internal television speaker which filters out hum and buzz, minimized the impact of distortion, excessive compression and equalization, differences in level and masked the effects of poor mic placement is no longer the benchmark. Consequently it can no longer be the scapegoat.

Newer receivers, with improved internal speakers, also offer the added capability of easier connection to the home entertainment system. Many tests have shown the improved audio quality provides the perception of improved picture quality. In audio, the standard is no longer "broadcast quality" but rather "CD quality."

High-definition or advanced television will provide improved picture quality and the capability of "CD quality" sound transmission. However, the benefits of bigger and better pictures will be fully realized only with bigger and better sound. Many of these improvements can be implemented today, by simple electronic upgrades such as replacing the analog cart machine with a CD player, digital cart machine or workstation.

Other improvements can be accomplished with modified production techniques and quality control measures. Greater attention should be placed on maintaining level, loudness and phase uniformity, eliminating over equalization or compression, correct alignment of analog tape machines and Dolby processors and insuring a system interface free of hum and noise.

The extension of the usable lower frequency limits and the expansion of available dynamic range to 90 dB and greater has placed more stringent demands on the monitoring system as well as requires an acoustic environment of equivalent capability. As video production and post-production facilities are constructed or upgraded to accommodate HDTV with stereo/surround sound of "CD quality", attention should be placed on providing an acoustical environment which does not hinder production and in which quality judgments can be accurately made.

Improved Audio Monitoring

At every location where quality judgments are made, the audio monitoring system must be equal to the task. It must be linear and capable of accurately reproducing the lower octaves. The best way to accomplish this is to start with a linear monitor, with dispersion selected to complement the room geometry and use. In most cases, some degree of high frequency rolloff is preferred to a ruler flat response to beyond 20 kHz.

In an audio control room, the physical conflicts of a dedicated center audio monitor for surround mixing, a video monitor at something less than neck straining height and a panoramic window into the studio presents a real physical challenge. Dolby recommends that the center monitor match the left and right and be placed at the same height for optimum imaging.

The main left and right speakers should be and usually are placed symmetrically about the main video monitor. In most cases, this is at or near one or more of the room corners. At low frequencies most loudspeakers radiate over 360 degrees. Some of the low frequency energy will be reflected back into the room, slightly delayed.

Because of the long wavelengths involved, most of this energy will be in phase. Unless the speaker is designed with this type of mounting, the low frequency response will be boosted. Certain frequencies, however, will be canceled, depending upon the exact geometry involved. The worst case occurs when the loudspeaker is equidistant from two adjacent walls as well as the ceiling.

Professional recording studios attempt to circumvent this problem by soffit mounting the audio monitor loudspeakers. This forces the low frequency energy to radiate over a 180 degree angle rather than 360 degrees, boosting the low frequencies more uniformly. Some loudspeakers have internal equalization for this alternate mounting method. If not, external equalization should be incorporated. A 6 dB boost in the low frequencies does not allow for quality monitoring.

In either case, when using audio monitors with extended bass response, it is important that they be decoupled from the structure of the room. This is necessary not only to prevent rattles but also to maintain good stereo imaging. Sound travels faster through most structural materials than it does in air. The structure borne sound will arrive before the direct sound.

Enhanced Sound Isolation

All on-air, production, and post-production facilities have the common requirements for the appropriate degree of sound isolation from the outside. For the single room facility, location is usually the determining factor. Proximity to external stationary as well as transportation noise sources will establish the degree of isolation required.

When choosing a site, be sure to evaluate every possible condition. External noise levels vary greatly from day to evening. Truck traffic, railroad and manufacturing plant schedules may be quite different on weekends than on weekdays. Air traffic patterns can be altered due to weather.

In a multiroom facility, the unavoidable proximity to high level, low frequency sources demand a level of isolation that in most instances exceeds those of any potential external sources. Rooms for different applications require different degrees of isolation. A foley room or narration studio where live sound is being recorded requires a lower noise floor than a post production room where external noise is more a distraction. The distinction is extremely important because the cost of each increment of isolation can increase construction costs geometrically.

The conventional method of specifying the noise level requirements in a room is the NC (Noise Criteria) curve. The data on which the NC curves are based was gathered from surveys of office workers on the effects of environmental noise on their ability to perform work and to communicate speech. The curves were developed for the purpose of providing criteria for reducing complaints to tolerable levels. They have been adapted by the building industry as a simplified method of communicating the allowable noise levels in a room. [2],[3]

NC 15-20 used to be the standard for studios and concert halls. Digital recording and Dolby SR have lowered the standard goal to NC 5-10, and even below. Achieving this low level of noise is extremely costly, each 5 point reduction in noise criteria translating to a minimum 10-15% increase in construction costs. [4]

At NC-20 the allowable noise level at 63 Hz is 51 dB SPL. While less than ideal, for some music and most voice recording this may not be an insurmountable problem if, for example, the use of a high pass filter is tolerable. For Foley work, especially when the sounds being recorded fall in this frequency range and are at or below the background level, a filter won't help. And when several channels are summed, the situation becomes further aggravated.

The human ear is less sensitive to low frequency sound. However, most microphones are designed to be flat. In passing through the electronic chain, any rumble picked up can increase distortion and in extreme cases even cause amplifier clipping.

The degree of sound isolation actually achieved is dependent upon three factors: 1) the mass, stiffness and damping of the enclosing walls, ceiling and floor, 2) the airtight sealing of all penetrations for doors, windows, cables, and air ducts and 3) the physical separation from internal and external noise sources. [5]

The first two are summarized in the STC (sound transmission class) rating system adopted by the building industry as a simplified means of comparing various barrier constructions. Because this rating system was devised primarily for residential and office applications, data for frequency bands below 125 Hz is not always available nor is it reliable.

At low frequencies, wall, floor and ceiling transmission loss become much more complex to predict. The transmission loss is no longer governed only by mass, but such additional factors as damping, stiffness and panel dimensions.

Materials such as concrete and concrete block perform well in the low frequency range by virtue of their combined mass and stiffness. Theoretically, STC increases by 6 dB by each doubling of mass. This is works to good advantage at first, but the benefits quickly diminish.

At 2,243 kg/m3 (140 lbs/ft2), a 152 mm (6 in) thick 3.66 m (12 ft) high wall will weigh approximately 1,250 kg/m (840 lbs/ft). This is not a severe problem on grade, but on the upper floors of an existing high-rise office building could present some interesting structural challenges.

Floating Floors

When increasing STC by adding mass quickly reaches its practical limit, the only other practical method of increasing sound isolation is physical isolation. Floating or "room within a room" construction is usually necessary to prevent vibration and structure-borne sound from entering the room. Ease of access into the room and the law of gravity limits the physical separation of a floated floor from the structural floor.

There are three basic types of floated floors. The lowest cost and least effective is the use of a continuous underlayment. A continuous sheet of neoprene, fiberboard, or proprietary materials is laid down, covered with building paper or polyethylene, and concrete poured on top. Because of the limit static deflection, this type of floated floor is effective only in the mid and high frequency ranges.

The second type employs a neoprene or coated fiberglass fixed mount. These are placed at 305 mm (12 in) to 610 mm (24 in) centers and covered with plywood and polyethylene. Concrete is then poured over this form. A STC of 73 can be achieved with using this method.

For extreme low frequency vibration isolation, spring mounts are required. Although fixed springs have been used, the most common system is the raised slab system. Housed metal springs are placed on .91 to 1.22 m (3 to 4 ft) centers with an integral steel reinforcing grid. Concrete is poured over the mounts and grid. After the concrete has cured sufficiently, about 30 days, the slab is raised by a process of slowly turning jack screws built into the mounts. Isolation down to sub-sonic frequencies can be obtained and STC ratings of 82 and greater are possible.

Inside Sources

Internal noise sources such as lights, dimmers and fans as well as the necessary penetrations of the enclosure such as cables and air-conditioning ductwork must be addressed. Dimmers, ballasts and transformers can be remotely located. When budget permits, DC lights and dimmers eliminate this potential noise source as well as significantly reducing the introduction of hum into electronic circuits.

Cable ducts and conduits provide very efficient transmission paths for sound. These must be sealed airtight, yet must be available for future changes. A good solution, especially when cable trays must be used, are products such as Crouse Hinds' "Thru-Wall Barrier." These devices consist of mounting frames of various sizes and elastomeric sealing blocks that form a tight seal around cables and conduits.

Quiet Air Conditioning

Excluding doors and windows, air-conditioning ductwork represents the most significant penetration of an enclosure. Careful sealing and structural decoupling of these penetrations are crucial. Table 3 illustrates the importance of eliminating even the smallest crack in an otherwise adequate partition. [6],[7]

Air conditioning noise, especially low frequency noise, presents one of the most difficult challenges to providing a low noise floor. Conventional techniques such as fibrous duct lining and passive silencers are effective above 250 Hz. A recent study has suggested that placing the duct lining in a manner that the fibers are normal to the duct axis can nearly double the attenuation at low frequencies over that of the conventional orientation. [8]

Passive silencers of sufficient length to control low frequency noise place severe static pressure restrictions on the air handling units. Fortunately, active noise control systems are becoming a viable solution to solving low frequency attenuation problems.

First patented in 1934, active noise control systems consist of a microphone that detects the noise as it propagates down the duct. A DSP controller processes this signal, determines a canceling waveform and introduces this signal through a loudspeaker. A second microphone located just beyond the speaker provides error correction feedback. Attenuation is 12-20 dB between 40 Hz and 160 Hz. [9]

Independent duct systems for each critical room are necessary. Whether sheet metal, rigid fiberglass, or flexible, the ductwork should be routed and isolated so that it does not pick up any noise along its path or couple sound from one room to another. The ductwork should generate through vibration any noise of its own.

Sheet metal duct, although having higher transmission loss that rigid fiberglass or flexible duct, can generate popping noises when it is pressurized or depressurized when it is not properly braced and damped. Square or rectangular duct is more susceptible to producing aerodynamically generated noise. Round or flat oval sheet metal duct minimizes both of these problems. Flexible duct has very low transmission loss. It can generate crackling noises as it expands and contracts when the fan is turned on and off. Very often it becomes pinched, restricting airflow and generating noise. On the other hand it is very economical and does not transmit vibration as well as the other types.

To introduce the air into the room without adding noise requires a low outlet velocity. To attain NC 15 a outlet velocity of no greater than 250 ft/min is required. Consequently, a large outlet area is required. Dampers within 6m (20 ft) of an outlet should be avoided as they reduce the net area of the duct and therefore increase noise. Grilles and diffusers must be sized so as not to restrict air flow.

Fine Tuning

Attention to detail is important in eliminating other potential internal noise sources. Any device with a fan should be relocated to a non-critical room if possible. Devices which must remain in the room should be treated on an individual basis to minimize their noise contribution. Light fixtures, air grilles and registers, and console and rack panels can be set into vibration at particular frequencies. These offenders can be easily identified with a sweep oscillator and damped with neoprene or foam.

In the studio, script stands or tables should also be investigated for resonance and covered with an absorbent material such as heavy felt or carpet. Chairs should be selected for quietness as well as comfort. Storage cabinets should be built so as not to rattle or resonate and prevent any items stored in them from doing so as well. Again, the best solution is to remove potential problem items from the room if they are not required.

Reverberation and absorption

A reverberant room will be noisier than a non-reverberant room of the same volume. Sound is not absorbed when it strikes a boundary but is merely reflected back.

Unlike music recording studios where room character and moderate reverberation is desirable, for voice or Foley work, the room should be as transparent as possible. Voice and Foley studios for video are usually very small. Because of their low volume, traditional reverberation time calculations are not meaningful. However, any reflected energy should be made diffuse and of essentially uniform frequency content.

Reflection room resonance can be controlled by the addition of absorption with porous materials, diaphragms and resonators, geometric reflection control and diffusers. All of these techniques are to varying degrees non-linear with frequency. Porous absorbers become ineffective below 250-500 Hz, depending upon their thickness. Diaphragmatic and resonant absorbers are by design frequency sensitive. Most reflective surfaces become absorptive, usually with increasing frequency, and at some point can become diaphragmatic as well. Diffusive surfaces not only vary with frequency but also can be dependent upon orientation. They also usually become absorptive and diaphragmatic at certain frequencies.

At frequencies below 250 Hz, conventional techniques require a depth of porous material or a cavity equal to 1/4 the wavelength of the lowest frequency to be absorbed. At 20 Hz this is over 4.27 m (14 ft). Even at 40 Hz it is approximately 2.13 m (7 ft). This consumes much real estate but for a critical listening environment the improvement in low frequency transient response is worthwhile.

Recent work at the U. S. Army Construction Engineering Research Lab has developed low frequency absorbers with normal incidence absorption coefficients approaching 1 with a thickness less than 10% of a wavelength. [10]

These are a few of the areas that should be given attention when constructing or remodeling a production or post-production room for critical audio recording or monitoring. Although fine tuning can sometimes be done after construction, appropriate isolation must be built in. Correcting problems after the fact will always be more costly than designing it right from the beginning.

Is home room noise the limiting factor?

It has long been assumed that home receiving and playback equipment as well as the home listening environment were limiting factors in chain. The CD and the proposed advanced television systems will eliminate the electronic portion of the chain from this consideration.

That leaves the home listening environment in question. Environment acoustics has received much attention over the past decade. Many communities have set criteria for the intrusion of transportation noise sources into the residential environment. Concern for energy efficiency has produced the side benefit of increased isolation for exterior noise.

In a recent AES paper, Cohen and Fielder [4] have re-addressed the question of what is the appropriate background noise level for a recording facility. In a study of 27 home listening rooms, they found the rms average noise level to be NC 17. A substantial portion of the rooms measured had less noise than the average. These results agree with previous surveys referenced in the paper.

These conditions may not exist with kids, pets, kitchen appliances and HVAC units adding their contribution. But during serious viewing and listening situations, the noise levels measured are most likely reliable. More extensive research should be done to confirm this data.

Evaluation of an existing room or proposed environment

Before buying or leasing any facility, the prospective owner should minimize his investment risk by conducting an initial walk-through, preferably together with the architect, designer, and contractor. The following are some of the items that should be assessed:

Control Room Shell of 55-85 m2 (600-900 ft2)

Studio Shell as required (55 m3 / 2,000 ft3 minimum)

Support Space as required

Room for expansion

Clear Height of 4.5 m (15 ft) minimum

Column spacing 7.5 m (25 ft) or greater

Floor loading capability of 730-975 kg/m2 (150-200 lbs/ft2)

Existing mechanical system capacity (heating, ventilation, air-conditioning)

Electrical service- ease of upgrade - power conditioning required

Structural system and exterior wall composition

Ability of roof to accept additional loads for isolation construction and HVAC

Roof Condition

Air traffic paths

Railroads

Automobile, truck, motorcycle

Road condition (potholes)

Soil type

Proximity to TV & radio transmitters

On the basis of this on-site research and other information, the prospective owner should be able to determine the suitability of the existing structure. Some idea of the extent of changes required for transition from the existing structure to the new use should also be developed. Assess the scope of these changes as they affect both the interior and exterior of the building. Examples of some specific considerations would include zoning requirements, building codes, required upgrades to an older building, disabled access, additional parking, earthquake or wind reinforcement, and fire codes.

Some general contractors offer preconstruction services, such as value engineering, that can save time and money, and minimize surprises and headaches later on in the project.

Grounding the multi-room facility

From a technical power and signal interfacing standpoint, a single room facility is relatively straightforward. However, all technical systems take on an added dimension of complexity when designing the multiroom facility. Inattention to detail can result in facilities where individual rooms function internally, but interface between more than one room may be erratic, or even impossible. This would severely limit the usefulness of the facility.

Paramount among considerations is a well planned and scrupulously executed grounding scheme. For AC power, 2 grounding concepts predominate: 1) a true "star ground" with every outlet having a separate ground wire pulled back to the central technical ground point, and 2) a devolved ground, in which each major room has a central ground point, and these points connect to the central technical ground for the entire facility.

The true star ground is by far the more effective method. The third pin ground wires from all outlets should be brought to a bus bar connected to the central technical ground point. All receptacles must be the isolated ground type, usually identifiable by their orange color or green triangle. All technical power should be completely shielded in steel conduit and raceways. Romex should never be used. However, conduits must be connected to ground and must not be mechanically connected to conduits for any other power system, or to anything else metallic such as water or sprinkler systems.

Technical power should never be used for any other function such as copiers, kitchen equipment, air-conditioning, or any non-production equipment. Only when all these conditions exist can an interface system be developed which is consistent and trouble free.

All equipment should have only one connection to the ground system. The most convenient ground connection is the third wire, usually green, in the AC power cord. According to the National Electrical Code (NEC) and other codes this wire must be used for safety. Some prefer to use a separate ground wire to each piece of equipment. This is not usually necessary and is very inconvenient in cases where equipment has to move from room to room.

Obviously, if one leaves the AC cord ground wire intact to avoid ground loops, any other signal connection between equipment must not complete a ground connection. In other words, all signal shields must connect at one end only. Various methods exist for achieving this goal, including complex schemes involving shields cut just about everywhere and bussed grounds on patchbays.

Most of these ground schemes can be made to work as long as the cardinal rule of grounding is applied: Do it consistently. However, any scheme involving ground bussing can create different ground "nodes", with varying impedance to true technical ground.

In a multi-room installation, this can create certain patches or equipment configurations that never seem to work totally hum and buzz free. The simplest shielding method is to carry all shields through any interconnects and patchbay normals, and lift the shield at one end. The most prevalent choice is to connect any shield at its source and lift it at its destination. This means right at the equipment and not at an intermediate connector or patchbay.

Most equipment has some type of captive cabling allowing easy disconnect and establishing a connector "standard" such as XLR to "Elco" for multitrack machines. Thus, all input ground lifting is done in this equipment specific cable and not in any truck or tie lines. This greatly simplifies installation as all bulk wiring is done in one simple way, with all shields intact.

References

1. Westerink, J. H. D. M. and Roufs, J. A. J., "Subjective Image Quality as a Function of Viewing Distance, Resolution, and Picture Size, "Journal of the Society of Motion Picture and Television Engineers, 98:113-119, Feb. 1989.

2. Acustica, Vol. 14, pg. 33, Fig. 14, 1964.

3. Gizzy, Vin, "Acoustic Design, Noise Control," Recording Engineer/Producer, August, 1988.

4. Cohen, Elizabeth A. and Fielder, Louis D., "Determining Noise Criteria for Recording Environments", Journal of the Audio Engineering Society, Volume 40, No.5, pp. 384-401 (1992 May).

5. Retinger, Michael, "Handbook of Architectural Acoustics and Noise Control, (TAB Books, Blue Ridge Summit, PA, 1988.

6. Jones, Robert S., "Noise and Vibration Control In Buildings", (McGraw Hill, New York, 1984

7. Waropay, Vincent M and Roller, H. Stanley, "Design Aid for Office Acoustics", Form & Function, United States Gypsum Company.

8. Wassilieff, Con, "Inverse Anisotropy Ductlinings", Internoise 91, pp. 87-90.

9. Goodman, S., Burlage, K., Dineen, S., Austin, S., and Wise, S., "Using Active Noise Control for Recording Studio HVAC System Silencing", presented at the 93rd Convention of the Audio Engineering Society, San Francisco, 1992, Preprint 3376 (E-4).

10. Jinkyo, Lee and G. W. Swenson, Jr. "Compact Sound-Absorbing Structures for Low Frequencies," "Proceedings of the 1991 National Conference on Noise Control Engineering," D. A. Quinlan and M. G. Prasad, Eds. (Noise Control Foundation, New York, NY, 1991)

Back to Top