Before we start allow me to introduce myself. I am Leslie Gaston-Bird (AMPS, MPSE) and I am a re-recording engineer and sound engineer currently based in Brighton, England. I have an A.S. in audio technology, B.A. in telecommunications and an M.S. in Recording Arts. I worked for National Public Radio at their headquarters in Washington, D.C. as a Broadcast/Recording Technician from 1991-1995 before moving to Colorado to become the Audio Systems Manager at Colorado Public Radio until 2002. During that time I was also the recording engineer for the Colorado Symphony Orchestra.
In 2002 I transitioned to audio post production, working at Post Modern Company doing soundtrack restoration for films from the Sony/Columbia picture archives. From 2005-2018 I taught at the University of Colorado Denver.
In 2018 I relocated to Brighton, England where I continue to work as a freelance audio engineer. Some of the films I have worked on include short features and documentaries like Doc of the Dead, Infinity Chamber, Klocked and A Feral World.
I also currently serve as the Vice President (Western Region) for the Audio Engineering Society, and will begin a term as Governor-at-Large for the AES in 2019.
OK, back to the problem in hand. So you’ve booked two days in a studio to finish a 5.1 mix. You show up with your hard drive, and open the session. When you go to check the monitor calibration, you find a few level anomalies. What do you do?
The best thing to do is to notify the lead engineer and have them sort things out - after all, you’re paying for the time. But there are lot of “what-ifs” to consider: What if the engineer is with another client? What if the engineer doesn’t want to do it or scoffs at you? What if they’re calibrating to a different specification?
This type of rhetorical question is great for a social media flame-war. However, there is a learning opportunity here: what if you want to do a quick level-calibration as a stop-gap solution while you navigate the studio politics?
The goal of this article is to show you how to do a level calibration that keeps you out of hot water and doesn’t interfere with the existing studio set-up. The understanding throughout this article is that we are operating in conditions beyond our control, and that ultimately the best solution is to address these issues ahead of time..
Overview Of Level Calibration Steps
The Recording Academy's Producers & Engineers Wing “Recommendations For Surround Sound Production” document is the starting place for proper level calibration. In sections 3-5 and 3-6 of that document, engineers are advised to calibrate the main 5 loudspeakers, one at a time, using two octaves of band-limited pink noise (measured with an SPL meter set to C-weighting, slow) so that all speakers are the same level (between 79 and 85 dB SPL), and in a bass managed system to use two octaves of band-limited pink noise for the subwoofer so that it is 10 dB SPL above the main loudspeakers. Ideally, there will be a flat frequency response when all speakers are active. According to the recommendation,
“All speakers must be correctly calibrated so that they are not only equally matched in level, but so that their crossover frequencies are aligned to that of the subwoofer being used in order to ensure a flat frequency response. This is important whether bass management is being used or not.”
Ideally, your loudspeakers will also be time aligned, not just calibrated for level. This helps with image “smearing” and localization.
The Real World
In the real world, many studios can accomplish these specifications, but a few do not for a variety of reasons: cost, diligence with maintenance and installation, expertise, balancing the needs of their usual clientele, etc.
Fortunately, to do a level calibration, you can insert a monitor bus as an Aux send in Pro Tools after your record tracks and before the feed to the studio loudspeaker outputs.
Even if you have several stems (dialog, music, effects, M&E and Composite), the only thing you have to do is insert this track before the output and proceed to adjust the individual speaker feed (not the loudspeakers themselves).
In this “quick and dirty” process, my goal is to make sure that my panning and front-to-rear balance is working for the film I am mixing. Therefore, I am only concerned about making sure the left, center, and right are within 0.5 dB of each other and the front and rear are closely aligned in level as well.
You’ll need to generate pink noise with the Sound Generator plug in, and apply a band limiter. When you set up your inserts, the first one should be the sound generator, followed by the EQ. I recommend EQ3 but any will do, as long as you can perform a steep roll off; 24 dB per octave is what I use. The recommendations call for 2 octaves of pink noise at -20 dBFS for the main channels (I used 500 Hz to 2000 Hz) and 2 octaves for the subwoofer (I used 20 Hz to 80 Hz).
The tool I use to measure sound pressure level (SPL) is an app called Audio Tools by Studio Six Digital (currently $19.99 on the iPhone App Store). Its interface is reminiscent -- perhaps even styled after -- the Radio Shack SPL meter which was commonly recommended by many texts on surround sound calibration. In Figure 2, notice that the weighting is set to “C” and the response is set to “slow”.
You will be sending this signal to each channel’s loudspeaker, adjusting your aux bus of each channel to make the meter read the desired level, which in my case was 78 dB SPL (which, as it turns out, was still a bit too loud for the small room I was in, but more on that later.)
I held the phone in front of me at chest level. Ideally, you might have a tripod and set the device at ear height.
When I did my level calibration, I got the following results:
|Left:||78.0 dB SPL|
|Right:||73.8 dB SPL|
|Center:||73.8 dB SPL|
|Left Surround:||75.0 dB SPL|
|Right Surround:||72.6 dB SPL|
|LFE:||82.0 dB SPL|
I wanted to even out these levels so that I was monitoring at 78 dB SPL. If you do not ensure the levels are even, you run the risk of improperly balancing multichannel music and effects (and any dialog that is not panned to the center channel), and the viewer would experience a left- or right-heavy mix, or some panning elements that don’t work. Front-to-back pans and the relative level between front and rear would also be out of balance.
My host studio I also uses bass management. This means the nearfield monitors are crossed over to the subwoofer, and the LFE channel also goes to the subwoofer. I have two explosions in the film for which I wish to use the LFE channel, so I definitely want to make sure I take advantage of the +10 dB boost. By raising the level on my LFE monitor auxiliary feed so that the measured output is +10 louder than the mail channels, I am hopefully hearing what the audience will hear.
As mentioned before, the reason to do level changes within Pro Tools is you don’t want to get behind the speakers when you’re a guest in someone else’s studio. If you are working in your own studio (on your own “turf”), there’s no need to do this because you should spend time properly calibrating your room.
There are a few ways to achieve the changes you’ll be making to your monitors.
Use individual aux tracks for each loudspeaker (Figure 3)
Use the trim plug-in and unlinking each channel using the “pan linking function” (Figure 4)
One advantage of using individual aux tracks is that you can do a quick, visual check to make sure your levels are where you want them, and that channels don’t accidentally get linked or unlinked using the pan linking function. The advantage of using the pan linking function of the trim pot is that you can keep your track count low.
With the readings I obtained (see Table 1), I needed to keep the left channel the same, but raise the right and center 4.2 dB, the left surround channel 3 dB and the right surround channel +5.4 dB. I did this using aux busses called “5.1 Monitor Bus” (abbreviated 5.1MonBus” in the IO Field in Figure 3)and sending them to a monitor aux called “LG Surround Mon” (abbreviated “LGSurndMON”):
I could accomplish the same goal using Pan Linking on the Monitor Aux by inserting a Trim plug-in and unlinking the channel. Click on the “link” symbol in the upper right-hand corner of the plug in and it will turn from blue to white, and the small squares representing each loudspeaker will turn from green to black with a green border. See Figure 4.
Routing: Step By Step
Your dialog will go to a dialog bus, Foley and SFX (sound effects) to the effects bus, and music will go to the music bus. If you are working in 5.1, then these busses should be in 5.1 as well.
The Composite track will contain dialog, SFX, and music, therefore those busses should be routed to the Composite (Comp) track. For example, the figure below shows the paths for various mix elements:
In Figure 6, you can see that the audio from the composite mix (COMP Surround bus, abbreviated COMPSurndBs) is being routed to the monitor auxes (5.1 Monitor Bus, abbreviated 5.1MonitorBs). Other mixes are routed to bus 9 only so that they have both an input and output assigned; I can check them by temporarily routing them to the 5.1 Monitor Bus.
When working on a long movie, I don’t bounce the stems. Instead, I route those channels to an audio track. This insures that I can “punch in” changes if they have to be made. It also serves as a paranoia check because I can keep an eye on the meters and detect anything that might have been misrouted (for example, dialog showing up on the music track).
The Composite track is what I will be monitoring through the Monitor Aux. Because it is an audio track being recorded, and I am monitoring what is being recorded, I know that my makeshift monitor levels will not interrupt what is going “to tape”.
Room Volume And Monitor Levels
During the mix, I noticed I wasn’t hitting the ideal -24 LKFS specification for broadcast. I use the Waves Loudness Meter as a plug in on my composite track. The problem was that I was simply monitoring too loud. By turning it down, I would have pushed the faders up and hit my target loudness level. Below is a table of recommended room volume and monitor levels. The room I was in was approximately 10 ft x 10 ft x 7 ft, or 700 cubic feet.
|Categories||Room Volume in Cubic Feet||SPL in dB re 20 μN/m2|
|1, 2||> 20,000||85|
|10,000 < 19,999||82|
|5,000 < 9,999||80|
|1,500 < 4,999||78|
You can also read the AES recommended loudness levels for Over the Top Television.
If you mix in the same studio every day, you don’t have to worry about this, but if you have the ability to bring your mix from one environment to another you do need a baseline for understanding how your mixes are translating from room to room. After all, this film will eventually be on the festival circuit, and could play in an outdoor amphitheater, a high school gym (believe it or not, high-profile movie screenings happen in this kind of environment), or a proper cinema.
I’d like to thank Will Worsley and Tanya Auclair at Coda to Coda Studios for hosting this mix and letting me grab some screen shots. They have a welcoming studio with a great, friendly staff and they are doing amazing work. Thanks to their “home away from home”, I was able to work comfortable and have a great-sounding mix in the end.