AV 9000 Standard Checklist Item Under Test:
For audio conference systems, confirm microphone input gain settings are such that the "standard talker" (60 dB SPA at 1 m, IEC 60268-16), positioned at each talker position in the room, produces a nominal “0 dB” level at the input of the mixer bus of the audio conference DSP device. If there is local reinforcement ("mix-minus"), AGC and ALC may need to be restricted. Record test results as pass/fail. Record level across analog telephone line. Inspect DSP mixer telephone line levels, both transmit and receive, when normal speech is encountered in the room.
In order to provide optimal signal to noise ratio on microphone channels, while still allowing for adequate headroom, the AV 9000 standard suggests setting all audio levels (microphones, conference transmit, conference receive, etc.) in a DSP mixer to “0 dB”. The reason “0 dB” is in quotations is because sometimes it is a solid yellow, sometimes it is actually 0 dB read on a meter, and sometimes it might be -20 dBFS. As common as “0 dB” (typically 0 dBu) is in today’s DSP mixers, I think dBFS may be more informative and useful to the site file programmer.
We drive the whole “0 dB” thing into all our specialists from a very early age. My son is 2-1/2 years old and he knows that mics get set to 0 dB. So, when some manufacturers decided to use dBFS, instead of dBu on their meters, we heard some funky sounding far ends when the levels were set to 0 dBFS.
Before I get ahead of myself, I’ll explain the difference between then in laymen’s terms. When a meter uses a “dB” or dBu scale, the signal voltage level is expressed relative to 0.7746 V. When a meter uses dBFS, the signal voltage is expressed relative to the maximum voltage a device can handle before clipping occurs. So setting signals to 0 dB in a mixer using dBFS is setting signal levels at clipping. Hence, the funky sounding far ends, and my awful attempt at AV-centric humor (boo).
AV9000 adopted the “set to 0 dB” mantra because it worked for most devices. At 0 dBu, the signal level was at least 60 dB above the noise floor (strong signal to noise ratio), and the mixer still had 12, 15, 20, or 24 dB of headroom before clipping above 0 dBu (good headroom before distortion). Whether I had 12, 15, 20, or 24 dB headroom before clipping didn’t matter too much because I still had 12 dB of headroom or better.
However, did I miss out on 10 dB of signal to noise ratio by using my 0 dB generic setting? If my mixer was capable of handling +24 dBu of signal, and I only need 15 dB of headroom, shouldn’t I really set my levels at +9 dB to really optimize signal to noise and headroom?
If my mixer uses dBFS scales, it would force me to think that through. If the specification calls for 20 dB of headroom, then I can set my levels to -20 dBFS and be optimized. If the spec only calls for 15 dB of headroom, I can set my levels to -15 dBFS and be optimized.
Setting levels to “0 dB” is a proven rule of thumb that will work 99 times out of 100. In a field where many integrators are not even following this simple rule of thumb, it is a big step in the right direction. However, once we elevate the industry to a point where users begin to appreciate the difference between 60 dB and 70 dB signal to noise ratio on their microphones, dBFS might be slightly better for us. And don’t even get me started on how often people think “setting levels to 0 dB” and “unity gain theory” are the same thing!
Of course, knowing where the rails of the mixer are can give us the same performance. However, this requires looking up the maximum voltage accepted on the inputs of the mixer in the manual, as well as some subtraction…that’s a lot of work. With dBFS scales, I can just look at the required headroom and know where to set my levels. It also gives a lot more significance to the clients’ choice of headroom level. Instead of being a just a number included in the design boilerplate, now it dictates where I set my levels. The dBFS camp may get some flak for being different from the “dB” or dBu camp, but count me as a supporter of Team Full-Scale.
Originally appeared in Sound & Communications Magazine 1/1/14