Reference Levels

[ LiB ]

Reference Levels

Have you ever wondered why television commercials are always so much louder than the program? I always have to turn the TV down when the commercial break comes on and then turn it back up again when the program resumes. This is an obvious sign that reference levels are not being adhered to. If they were, the average level of all the program material on the television would be about the same and you would not have to constantly turn the volume up and down to maintain a consistent level.

The Purpose of Reference Levels and Tones

A reference level, or nominal operating level, is a measurement of the average program level of an audio signal. Obviously signals will go above and below this level in the course of a normal program, but the nominal operating level is used as a reference so that signals will be interchangeable between the different pieces of equipment, including broadcast systems, theaters, DVD players, and TVs. In analog systems, this level is measured using a VU meter. A VU meter is not a fast- reacting meter like the peak meters found in many digital systems. Extremely short transient peaks will not register on a VU meter and, therefore, it is important to realize that the actual level is exceeding the VU reading for very short peaks. The interesting thing is that human hearing perceives volume more like a VU meter reads audio signals. Watching a peak meter does not necessarily give an accurate reading of how your ear perceives the volume of that signal. It is because of this fact that VU meters are used to calibrate nominal operating levels.

NOTE

VU METERS

VU meters are used to calibrate volume levels. VU stands for Volume Units . VU meters' slow reaction time makes them less sensitive to transient peaks. As a result, they give a reasonable representation of the overall volume of a signal rather than the instantaneous level.

When mixing, some point of reference must be used to judge the average level of your mix. As discussed in the Calibration section of Chapter 8, speaker systems for film and theatrical presentations are calibrated to specific volume measured by an SPL meter. The digital level used to calibrate this volume is -20dBFS. This level corresponds to 0VU. Working backwards from the sound pressure level, 85dBC = 0VU =-20dBFS. Figure 9.1 shows how reference levels relate to the overall signal.

Figure 9.1. A graphic representation of reference levels and how they relate to dynamic range of a system. The gray areas represent analog noise and distortion artifacts as they increase the further you get away from the reference level.


NOTE

85DBC OR 83DBC?

It should be noted that many mix rooms are calibrated to 83dBC for film mixing and 79dBC for TV. True 85dBC coming from all three front speakers can get pretty loud. Dolby Laboratories suggest 85dBC as the reference, while Tomlinson Holman recommends 83dBC. If you mix primarily loud, bombastic action thrillers, you might consider using 83dBC just to protect your hearing! Smaller mixing rooms can benefit from a lower SPL setting due to the damping effect of having less air in the room.

As normal signals will include transient material exceeding 0VU quite often, any recording or reproduction medium must be capable of passing those peak signals without distortion. This capacity is called headroom . The amount of signal that can be passed prior to distortion above 0VU or the reference level is the amount of headroom available to that system. With analog recording systems, distortion artifacts start to creep in gradually. Tape saturation will typically begin around +12dBVU, depending on how the tape deck is calibrated. As the level gets higher, the distortion will increase until the point at which the system can no longer output a higher level; then total clipping results. With digital systems, there is no gradual increase in distortion. Once a signal reaches the 0dBFS, it enters complete clipping and distortion.

Knowing how much headroom is available in any given recording medium is imperative. With digital systems, the amount of headroom available is determined solely by the reference level that you use to record, in this case -20dBFS. That leaves 20dB of headroom above the reference level before digital clipping occurs. With an analog tape recorder, it's safe to assume that only 12dB above the reference level is available before some amount of distortion will be introduced. This means that when you are recording a digital master referenced to -20dBFS, and this master is going to be transferred to an analog recording medium such as Betcam SP video or VHS, the peak level should not exceed -8dBFS. If the final medium is going to be digital, such as DVD or a Dolby] Digital film print, then the entire 20dB worth of digital headroom is available for you to use. Understand that even though you have all this wonderful headroom at your disposal, the final playback system must be able to reproduce all the dynamic range you are using. If you constantly have peaks that reach 0dBFS, only the hardiest of playback systems will accurately reproduce these without distortion or failure.

On the low end of the scale is the noise floor. The difference between the reference operating level and the point at which program material becomes indistinguishable from the inherent noise of a system is called the signal-to-noise ratio . Digital systems are capable of recording signals at much lower levels before reaching the noise floor. Analog systems typically have a higher noise floor.

Using Reference Levels

The way to establish your reference is by generating a 1kHz sine wave at the prescribed level. Using the Signal Generator Audiosuite plug-in, you can generate a 1kHz sine wave at -20dBFS, as shown in Figure 9.2. This tone can be recorded prior to any mix on to the tape deck, DAT machine, or even digital file, so that anyone receiving your mix will have a reference tone they can use to align other equipment for 0VU. When this tone is used to set the proper recording level, the transfer will maintain the optimal signal-to-noise ratio and headroom. Digital transfers will not suffer from the lack of a reference tone. However, it is always prudent to include a reference tone because you don't always know where your mix will end up.

Figure 9.2. The Signal Generator plug-in set up to generate a 1kHz tone at -20dBFS.


NOTE

DON'T PROCESS REFERENCE TONES

Make sure that reference tones are not being processed by any plug-ins in the chain. Processing the master fader is no exception. When recording reference tones on a Master recording, disable any processing or plug-ins.

Film and Dolby Encoding

When mixing film, always use a reference level of -20dBFS. Pink noise at that level should play back through a single speaker in your system at 85dBC on an SPL meter. Both surround channels should be individually referenced at 82dBC. This compensates for the fact that theaters calibrate the surround channels 3dB lower than the main channels. The reference level on tape remains the same, -20dBFS. Only the speaker volume is reduced by 3dB so that the mix you create for the surround channels sounds right to your ears but is recorded on tape +3dB louder than the other channels. The theater calibration will reduce the output of the surround speakers by 3dB, thereby playing your mix exactly the way you heard it.

The two types of test signals and what they are used for are as follows :

  • Pink noise at -20dBFS RMS is used to calibrate speaker volume to 85dBC.

  • 1kHz sine wave at -20dBFS is used as a recorded reference tone, not for volume calibration.

DVD Authoring

When preparing mixes for DVD authoring, the methods are the same as for film except that the surround channels should not be referenced -3dB from the others. All speaker channels should use the 85dBC reference volume while mixing except for the subwoofer , which should be +10dB louder, as usual. A 1kHz sine wave should be recorded at -20dBFS on all channels as a reference tone.

Broadcast Video

Broadcast video is an interesting topic these days. With HDTV entering the fray, there are many choices for mixes that will be broadcast. On the one hand, commercial television has progressively become louder and louder with respect to the normal reference level of -20dBFS. TV spots are overly compressed and suffer from the same problems that FM broadcast and music mastering have faced. Clients want their audio to be louder than the competition's. This forces audio engineers to maximize the perceived volume of their mix while sacrificing headroom, through the use of digital peak limiters that, in essence, clip off the transient peaks of waveforms, allowing the overall level to be increased without over-modulating the signal. TV and radio broadcasters also employ severe peak limiting in the signal chain so that their transmission signals appear to be louder than competing stations without exceeding FCC modulation limitations. All this processing is detrimental to the overall sound quality. The result is a mix with severely limited dynamic range and a large quantity of harmonic distortion.

However, HDTV brings the promise of renewed fidelity and dynamic range in audio. Hopefully, the trend will be to use this dynamic range for creative purposes instead of trying to maximize level for perceived loudness. Only time will tell. Hopefully, you'll decide to maintain integrity within the audio mix instead of yielding to clients' demands for volume that results in the loss of quality. Clients, however, also pay the bills, and their demands must be met somehow. Education of the client in this area can help.

One thing many clients do not realize is that even though your mix might be limited and compressed and very loud in the studio, once it hits the broadcast chain, the same type of processing will be applied to it again, resulting in even more distortion and loss of fidelity. If properly mixed, a commercial with a reasonable amount of dynamic control can sound better than an overly compressed mix once it has been processed by the broadcast signal chain. Louder is not always better.

Hopefully, you can reach a compromise between the needs of clients to be perceived as louder and the desire for high-fidelity sound. Sticking to the prescribed reference levels for all program material will help end the Volume War.

Internet Media

As most Internet media will be highly data-compressed in order to reduce the bandwidth necessary for transmission, it will require more dynamic control than even a broadcast commercial. Most computer users still listen to audio on small, low-fidelity, multimedia speakers that are connected directly to their computer. It is advisable to at least preview your mix on a similar set of speakers to ensure that it will be heard the way you want it to.

Referencing mixes on small speakers is not unusual, especially in the music-mixing world. Top mix engineers like Bob Clearmountain and Tom Lord-Alge both check their mixes on simple consumer playback systems. This is similar to the use of Auratone speakers in the past. Auratones were designed to mimic the band -limited sound of small AM radio speakers. If you can get your mix to sound good on a tiny set of speakers and still sound incredible on larger, hi-fi speakers, you are doing a great job.

Find out the type and settings used for the data compression that your Internet audio will be processed with. If you are able to, run your mix through this processing to hear what it sounds like. You might be able to compensate for losses with mix alterations. Might as well mix for the final result.

Times are changing, and the age of the multimedia PC is dawning. Computer users and home video enthusiasts are using personal computers to process video and audio for viewing in the home theater. Multimedia that is transmitted via the Internet might find its way onto a higher-quality playback system in the near future. Always make sure you know who the intended audience is and on what type of system they will be listening to your mix. Something you mix to be heard on a two-inch multimedia speaker might, in a year or two, be heard on a 5.1 home theater surround system with a 15-inch subwoofer. You never know.

[ LiB ]


PRO TOOLS R for video, film, and multimedia
PRO TOOLS R for video, film, and multimedia
ISBN: N/A
EAN: N/A
Year: 2003
Pages: 70

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net