On Day-2 of my new blog David Prentice made this comment about LEVELS and this is coincidentally followed by a reader question…
Setting up a ‘direct line’ for questions sounds like a wonderful idea!
These days engineers are often application wise and audio signal path ignorant. Old dog concepts like gain-staging, level-matching and finding the control pot’s sweet-spot were essential to negotiating the limited dynamic range between noise-floor and distortion. Young engineers need “mentors,” more experienced engineers that can explain how things work and show how a little Knowledge can make things sound better.
Good luck on the new enterprise. (DP)
What a great setup for today’s question (ec).
I have a decent home studio setup and would like to get the most from it. I have a Tascam DM-4800 mixer that interfaces to PC via an RME 9652. (The self-rolled PC is running Cubase 4.5.2 on XP64, with plans to upgrade to cubase 6 and win7.)
The RME has 24 channels (3 ports) of lightpipe but the DM-4800 only comes with 8 channels (1 port) of adat stock. You can get more but you have to buy adat expansion cards for it. To overcome this limitation I use another audio interface, the EMU 1820m, as a converter to get 8 more channels out of the RME/Cubase rig to the DM so that I am able to mix 16 channels from the DAW – basically 8 stems. Everything is slaved to the RME clock. I have a UA LA-610 mkII for vocals and guitars, a UAD-2 and various other plug-ins as well.
Q-1: how do I need to set gain structure based upon my signal chain? I’m starting to have some success with placements on the MTV networks for shows like “Real World/Road Rules,” “Married to Rock”, and several others and would like to make my music as professional sounding as possible.
Q-2: How hot would a seasoned engineer record the signal of a sound source using my setup? Should I be recording a synth part at -0.3 dbfs or more in the -12 to -16 dbfs range so that I have more room to use eq, compression and fx?
Q-3: How does recording at full scale affect the mixing process? What I notice, and what concerns me, is that if I have a synth going and then add rhythm guitar the mix bus clip indicators immediately go red on me. Is this the product of recording too hot? Thanks for the help and super-fast reply.
Prophet Speaks Music
A-1: The most important detail is that 0dBFS = full scale = max, it’s the maximum recordable level AND the level at which the analog converters clip. To test both the AD and the DA converters (both in the Tascam DM-4800), plug a test tone into a line input and confirm that the Tascam’s metering is in agreement with Cubase’s metering. (The RME 9652 is ‘just’ the digital interface between the Tascam and Cubase and can not be calibrated.) Record this tone at 0dBFS, then -6dBFS, the -12dBFS and -18dBFS. Then play back and see what happens on the other side. You can copy the recorded track to simulate a session and confirm signal flow and headroom.
A-2: One ‘concern’ is that analog gear can have a max output that exceeds the Tascam’s input capabilities. To test / confirm, you’d want to inject a test tone into the analog gear, get the level to zero VU (0VU) on the analog meter and then see what level that shows up as on the DM’s meters. I’d expect that level to be anywhere from -16dbFS to -20dBFS. This is called NOMINAL level, which on a VU meter / Analog gear is +4dBu.
A-3: Analog gear has a minimum of 14dB to 18dB of headroom above Nominal and often more. The last thing you want to do is overdrive the converters, but IF you wanted to drive the analog gear a little harder, you’d need to insert a pad between the analog gear and the front end of the DM-4800.
I think there are rules being developed for mixing TV commercial audio, and as I understand it, TV programs and Films use -20dBFS as their dialogue reference level – this reserves headroom for the ‘suprise’ of sound effects.
A-4: It sounds like you are recording very hot, ‘across the board’ considering that you say it’s so easy to overload the mix buss. Recording level is program dependent, so you might try recording drums at -6dBFS max and other instruments an additional 6dB lower at -12dBFS.
A-5: If you need more headroom, try looking for a level trim option. EQ plugs on Protools and Adobe Audition have an additional GAIN control that allows overall LEVEL to be reduced to compensate for drastic EQ boosting. Surely Cubase EQ plugs have such an option. PT also has a Level TRIM plug, which, if you have one, can be placed BEFORE the EQ plug. The LEVEL TRIM PLUG can also be used as a “Line Attenuator, or PAD,’ reducing the overall level as a way to keep the faders up in a range where they have more ‘physical’ resolution.
A-6: When mixing, you should avoid overloading the mix buss at all costs. Try this: allow at least 6dB of headroom for the drums – peaks that are no higher than -6dBFS – and 12dB of headroom for instruments and vocals. I submix drums, bass and drum verb / ambience in both the digital and the analog domain. Doing that with other groups – instruments, vocals, etc – allows more overall level control. Please know that pulling the master fader down will not solve an overloaded mix buss.
Let me know if this helps and feel free to provide screen shots.
PS: I wrote this article on AUDIO LEVELS many years ago – It’s due for a rewrite (hopefully soon).
Ellis and I wrote back and forth a few times to clarify details. That dialogue was integrated into this correspondence. Ellis’s first response was “Wow,” and ‘thanks for the quick and awesome response!’
Related Topics: Ask Eddie