Metering Surround Sound
 

I'm getting ready to purchase equipment for my newly constructed studio. I'm setting it up to be able to mix 5.1 surround sound. However, I'm a little confused about metering. In the past I've used and liked Dorough meters. However, how does one use two channel meters to judge the loudness of 5 channels? Even if you had 5 separate meters, it would be hard to judge the overall loudness without mixing them down to one meter. Typically with 5.1 the voice and dialog sits mostly in the center channel. So if you patch the center channel into a meter you can judge loudness there. However, it would be 3db lower than if you had the same loudness in two channels like you do with stereo mixes. I could fold down the entire mix to stereo and use that to judge loudness, but then I would be continually switching between stereo and 5.1.

Also which do people prefer the -14db models or the -20db models? It seems like the standard for TV production is -20db, so why would I want the -14db one?

The 0VU = -20dBFS model is the one generally used today for high dynamic range delivery systems such as DVD and theatrical cinema. The 0VU=-14dBFS or -12dBFS models are generally used for lower headroom media such as VHS video. Very high dynamic range media like DVD and other strictly digital delivery may go with reference levels even lower than -20.

Meters will only correspond roughly to the subject sensation of loudness. As such, IMHO meters should be used to monitor technical characteristics of the signal while your ears and calibrated monitor speakers should be used to judge subjectve qualities such as loudness. In a surround sound setup you should be metering each channel individually, watching for clipping etc. The procedure is to set the gain of each channel in the system so that reference level pink noise hits a specific sound pressure level. depending on delivery destination, at the mastering engineer's ears. T. Holman suggests for high headroom destinations like theatrical release and DVD to align each channel with 1kHz, 0VU reference tone to read -20dBFS on the meters. Then without readjusting the system's levels, switch to a -20dBFS pink noise signal and while muting all but one channel at a time, go from one channel to the next in turn and set each monitor speaker's volume control with a sound pressure meter at the listening position. His recommended levels are 79dB SPL on all channels for video release, 81dB SPL on the L,C, and R channels and 78dB on the LS and RS channels for theatrical masters.

You might want to check out Thomlinson Holman's (he's the 'TH' in "THX Sound") book "Sound for Digital Video" where he devotes an entire chapter to mastering and monitoring and how to go about setting up the system. It includes a CD with the necessary sine and pink noise alignment tones recorded at the right level.

I will definitely take your suggestion and purchase the book. I also like your tips on calibration. But, my question was more about setting the levels than the calibration of the system. Is there an SPL target for dialog? And wouldn't meters help you repeat that SPL target? I find visual things much less subjective than auditory. Hence the market for $1000 speaker cables. So to me there is still value in having that visual reference. I guess the trick is not to become too attached to it. I'm also curious if anyone has meters set up to monitor surround sound, if so, how they do it. Is it simply a meter for each channel? That would get quite expensive.

Average dialog will meter similar to the pink noise reference I mentioned - that's why its used. With sine wave reference tone at -20dBFS, the pink noise will be about -12 or -10. So average dialog levels in the 80dB SPL range up to about 85dB SPL for theatrical release. And yep, it do get pricey.